AUSTRALIAN COMPETITION TRIBUNAL
Applications by Telstra Corporation Limited and TPG Telecom Limited (No 2) [2023] ACompT 2
IN THE AUSTRALIAN COMPETITION TRIBUNAL
File No: | ||
Re: | Applications by Telstra Corporation Limited and TPG Telecom Limited for review of Australian Competition and Consumer Commission Merger Authorisation Determination MA1000021 | |
Applicants: | ||
Intervenor: |
DETERMINATION
TRIBUNAL: | Justice O’Bryan (President) Dr J Walker (Member) Ms D Eilert (Member) |
DATE: | |
WHERE MADE: | Melbourne |
THE TRIBUNAL DETERMINES AND DIRECTS THAT:
1. The determination of the Australian Competition and Consumer Commission (ACCC) dated 21 December 2022, dismissing the application for merger authorisation MA1000021, be affirmed.
2. Until further direction of the Tribunal, the reasons of the Tribunal in this proceeding dated today are not to be made available to or published to any person save for:
(a) the ACCC, its staff and any other person assisting the ACCC in relation to the proceeding including the ACCC's legal advisers; and
(b) the parties’ legal advisors who, by reason of previous directions of the Tribunal, are permitted to have access to the confidential information of each of the parties to the proceeding.
3. Within 21 days of the date hereof, the parties are to file jointly:
(a) a copy of these reasons that marks, by way of coloured shading (using a different colour for each party), those parts of the reasons that a party seeks to have redacted on the grounds of commercial confidentiality; and
(b) short submissions addressing the basis for the claim of confidentiality on behalf of each party.
THE TRIBUNAL:
The application for authorisation
1 On 23 May 2022, Telstra Corporation Limited (Telstra) and TPG Telecom Limited (TPG) lodged an application with the Australian Competition and Consumer Commission (ACCC) seeking authorisation under s 88(1) of the Competition and Consumer Act 2010 (Cth) (CCA) for Telstra to operate radiocommunications devices under TPG’s spectrum licences pursuant to the terms of an agreement entered into between them dated 17 February 2022 and titled Spectrum Authorisation Agreement – MOCN Area (Spectrum Authorisation Agreement). It is implicit in the application that the authorisation that is sought is unlimited in time, save that it will apply for the duration of the Spectrum Authorisation Agreement.
2 The application for authorisation discloses that the Spectrum Authorisation Agreement is one of three agreements entered into between Telstra and TPG as part of the one commercial transaction (referred to by Telstra and TPG as the Proposed Transaction), which seeks to establish a Multi-Operator Core Network (MOCN) in certain regional and urban fringe areas which comprise around 17% of the Australian population coverage (in the 81.4%-98.8% population coverage area) (Regional Coverage Zone or RCZ). The other two agreements are the MOCN Service Agreement dated 17 February 2022 (MOCN Service Agreement) and the Mobile Site Transition Agreement dated 17 February 2022 (Mobile Site Transition Agreement). The application is clear, however, and Telstra and TPG have acknowledged in this proceeding, that the application for authorisation is confined to conduct comprising Telstra’s use of TPG’s spectrum licences pursuant to the Spectrum Authorisation Agreement (which in these reasons will be referred to as the Proposed Conduct).
3 Under the MOCN Service Agreement, Telstra agrees to use its radio access network to supply TPG with 4G and 5G services in the RCZ. This will deliver TPG an immediate uplift in mobile network coverage, increasing from 96% to 98.8% of the population. Under the Mobile Site Transition Agreement, TPG will authorise Telstra to access, use and occupy space at 169 mobile sites in the RCZ owned or licensed by TPG to enable Telstra to install mobile telecommunications equipment in place of TPG. TPG plans to decommission its remaining 580 mobile sites in the RCZ. Under the Spectrum Authorisation Agreement, TPG will authorise Telstra to operate radiocommunications devices utilising part of TPG’s 4G and 5G spectrum within the RCZ and beyond this zone for the purposes of its radio access network.
4 The overall effect of the Proposed Transaction in the RCZ is that Telstra will augment its radio access network with TPG’s spectrum rights and mobile sites, and will operate the augmented radio access network to supply its own mobile network and to supply services to TPG in the RCZ pursuant to the MOCN Service Agreement. Telstra will also augment its mobile network in the population coverage area beyond the RCZ with TPG’s spectrum rights. Each of Telstra and TPG will continue to operate their own core networks, giving them the ability to differentiate their services on features such as pricing, data and inclusions, and software enabled services.
5 Although Telstra and TPG have applied for authorisation only in respect of Telstra’s use of TPG’s spectrum under the Spectrum Authorisation Agreement, before the ACCC and on this review the applicants have based their application on the likely competitive effects of, and the public benefits and detriments likely to result from, the Proposed Transaction as a whole.
6 The application for authorisation is opposed by Singtel Optus Pty Ltd (Optus) on the basis that the Proposed Conduct, and the Proposed Transaction as a whole, would be likely to substantially lessen competition in the Australian market for mobile telecommunications services and related markets, principally by increasing Telstra’s market power and damaging Optus’s competitive position in those markets, and that any resulting public benefits would not outweigh the anti-competitive detriment.
The commercial and economic context of the Proposed Transaction
7 The Australian market for mobile telecommunications services is presently served by three principal mobile network operators (MNOs): Telstra, Optus and TPG. The MNOs are vertically integrated. They compete at both the retail and wholesale level, with their retail brands competing for consumer and small business customers, and their wholesale arms competing in the provision of wholesale services to resellers of mobile phone services, referred to as mobile virtual network operators (MVNOs), and other telecommunications providers. Since the introduction of competition in mobile telecommunications, the relative positions of the three MNOs has been relatively stable. Historically, Telstra has had the greatest share of retail services, followed by Optus and finally TPG.
8 There are high barriers to entry and expansion in the provision of mobile services. As a result of these barriers, the prospect of new entry, other than by niche providers, is low. The barriers to entry and expansion include: large up-front sunk capital investment; significant economies of scale (including because mobile services involve high fixed costs and low variable costs); scarcity of spectrum (spectrum being an essential input into mobile telecommunications); brand perception (as customer perceptions of network quality and reliability are a key driver of consumer decisions regarding mobile services and these perceptions can be difficult to shift); and first-mover advantages associated with the technology cycle (as changes in market share due to early advantages in technology lifecycles have the potential to endure through the lifecycle of a technology).
9 The three MNOs are currently operating networks that include three generations of mobile technology: 3G, 4G, and 5G. Each subsequent technology generation has brought increased bandwidth and speeds and improved the capabilities of the network. 5G is the newest technology generation to be deployed, and all three MNOs are rolling out their 5G networks currently. 5G technology makes more efficient use of spectrum, delivers faster speeds and provides better reliability and lower latency as compared to 4G technology. MNOs are incentivised to upgrade their networks in order to make use of this more efficient technology and meet evolving consumer needs, but doing so requires large up-front investments. MNOs must also balance repurposing (“re-farming”) their spectrum holdings for newer technology while continuing to operate the older technology simultaneously.
10 The availability of 5G technology is an increasingly critical focus of competition in the supply of mobile telecommunications services. All three MNOs are competing in the supply of retail mobile services on the basis of 5G availability, coverage and speeds and the capabilities enabled by 5G. The provision of 5G is also a basis on which MNOs compete to acquire wholesale customers.
11 Australia’s geography is characterised by a very sparse population density on average, paired with a very urbanised population centred in the capital cities and surrounding major regional centres. Mobile networks therefore need only cover a very small proportion of the total landmass in order to provide mobile coverage to the homes and workplaces of a majority of the population. However, the implication of this degree of urbanisation for mobile coverage is that covering the remainder of the population becomes decreasingly economic, and increasingly uneconomic, as network coverage expands into more remote and less densely populated areas. While providing mobile services in more remote and less densely populated areas is less economic, consumers value mobile coverage in the areas in which they live, work and travel. The extent of geographic coverage is a key component in the attractiveness of mobile services to consumers and to government and enterprise customers.
12 Since the inception of mobile technology in Australia, regional and rural investment has been considered by MNOs to be a challenge, and often not commercially viable. The history of regional mobile telecommunications investment shows that two of the primary ways in which MNOs have sought to make regional investment commercially viable is by obtaining government assistance, and by entering into agreements with their competitors to share network infrastructure.
13 Infrastructure sharing in mobile networks can be broadly classified as either ‘active’ sharing or ‘passive’ sharing. Passive infrastructure sharing may involve the sharing of non-electronic infrastructure such as cell sites, towers and buildings, but does not include the sharing of electronic equipment capable of processing or converting telecommunications signals such as radio equipment, or the sharing of spectrum (which is described as active sharing). Active infrastructure sharing may take a number of forms. One model of active sharing is called ‘roaming’. Roaming involves a host MNO carrying the traffic of another MNO on its behalf. The client MNO is not required to deploy any infrastructure in the relevant area. Other models of active infrastructure sharing are MOCN and multi-operator radio access network (MORAN) arrangements. Both models involve multiple MNO parties sharing active assets in certain coverage areas of their networks. Typical MORAN deployments include the sharing of active base stations, but not spectrum. MOCN deployments typically include the sharing of active base stations, as well as spectrum shared and owned by the operators.
14 TPG’s mobile business has historically been supported by network sharing arrangements with Optus. Optus and TPG have had a joint venture agreement in place since 2004 which they refer to as the eJV. The eJV provides for [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], and Optus provides 3G roaming to TPG in the 80-96% population coverage area. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
15 During 2021 and 2022, TPG engaged in negotiations with each of Optus and Telstra with respect to new network sharing arrangements in respect of the RCZ. Those negotiations culminated in the Proposed Transaction between TPG and Telstra.
16 On 21 December 2022, the ACCC made a determination dismissing the application for authorisation from Telstra and TPG. In its determination, the ACCC concluded:
The ACCC is not satisfied, in all the circumstances, that the Proposed Transaction would not be likely to substantially lessen competition, or would be likely to result in a benefit to the public that would outweigh the public detriment from the Proposed Transaction. Therefore, the ACCC must not make a determination granting authorisation to the Proposed Transaction under section 88(1) of the CCA.
17 It can be seen that the ACCC reached its conclusion based on the likely competitive effects of, and the public benefits and public detriments likely to result from, the Proposed Transaction as a whole rather than those resulting from the Proposed Conduct (the conduct for which authorisation was sought). The correctness of that approach to the application as a matter of law is a disputed issue in the proceeding.
18 The ACCC’s reasons for determination record that it received a large number of submissions from interested persons, including over 90 submissions in response to the ACCC’s initial consultation process and submissions from 44 interested persons in response to the ACCC's Statement of Preliminary Views.1 The categories of interested persons who made submissions, or with whom the ACCC consulted, included TPG (Vodafone) and Optus dealers, MVNOs and neutral host providers (mobile network infrastructure operators that provide access to infrastructure on commercial terms), NBN Co, mobile customers, local governments and industry and community organisations.2 The ACCC summarised the submissions received as follows:3
Overview of submissions
3.30. The views expressed in submissions to the ACCC were mixed. Some interested parties, including Optus, Optus dealers, MVNOs, industry bodies (such as Commpete, the Australian Communications Consumer Action Network, the Internet Association of Australia, and the NSW Farmers Association), investors in digital infrastructure, rural, regional and remote telco providers, and telecommunications consultants expressed concern over the Proposed Transaction.
3.31. Interested parties opposing the authorisation generally submit that the Proposed Transaction will have long-term adverse consequences for competition in the Australian telecommunications markets, noting effects on prices (particularly, an expectation that TPG will increase its prices), future investment incentives, and on the ability for smaller players and neutral hosts to enter the market. Concerns were also raised about Telstra's already dominant position and the effects of spectrum concentration, reduction to wholesale competition, and MNOs being disincentivised from investing in infrastructure.
3.32. Other interested parties such as, customers, TPG/Vodafone dealers, MVNOs, local governments, regional councils, organisations in support of regional businesses and residents, and industry bodies are generally supportive of the Proposed Transaction, believing it would be pro-competitive. These interested parties generally submit that the Proposed Transaction would increase customer choice for regional consumers, provide network improvements and ensure more efficient utilisation of infrastructure.
3.33. Some interested parties suggested that the ACCC could impose conditions of authorisation in order to address their concerns, ranging from imposing an obligation on Telstra to provide high quality wholesale access to any party which requires it, to declaring domestic roaming access in regional Australia, to allowing third parties to use the TPG tower sites that would be decommissioned, to requiring divestment of certain parcels of low-band spectrum to offset increased concentration.
19 On 23 December 2022, each of Telstra and TPG (collectively, the applicants) filed applications in the Tribunal pursuant to s 101 of the CCA for a review of the ACCC’s determination. On 24 January 2023, the Tribunal made directions for the two applications to be determined together. There is no reasons to distinguish between the applications and, in these reasons, the two applications will be referred to as the “application”. The Tribunal also made a direction pursuant to s 109(2) permitting Optus to intervene in this proceeding.
20 It is common ground between the parties that the application for authorisation filed by the applicants is a “merger authorisation” within the meaning of the CCA. Section 4 of the CCA gives the following definition:
merger authorisation means an authorisation that:
(a) is an authorisation for a person to engage in conduct to which section 50 or 50A would or might apply; but
(b) is not an authorisation for a person to engage in conduct to which any provision of Part IV other than section 50 or 50A would or might apply.
21 Section 68(1) of the Radiocommunications Act 1992 (Cth) (Radiocommunications Act) provides that the licensee of a spectrum licence may authorise other persons to operate radiocommunications devices under the licence. Section 68A(1) provides that, for the purposes of s 50 of the CCA (and related provisions), the authorisation under s 68(1) of a person to operate radiocommunications devices under a spectrum licence is taken to be an acquisition by the person of an asset of another person and conduct engaged in by the person.
22 As a result of those provisions, Telstra’s use of TPG’s spectrum licences pursuant to the terms of the Spectrum Authorisation Agreement is taken to be an acquisition of an asset for the purposes of s 50 of the CCA. As noted above, the application for authorisation is confined to that conduct and, accordingly, both limbs of the definition of “merger authorisation” are satisfied.
23 A review by the Tribunal of authorisation determinations made by the ACCC is governed by the provisions of Pt IX of the CCA. As discussed in the Tribunal’s decision in Re Telstra Corporation Limited and TPG Telecom Limited,4 a review of a merger authorisation under Pt IX differs from a review of other authorisations in two material ways:
(a) first, a review of a merger authorisation is required to be completed by the Tribunal within a statutory time period (whereas a review of other authorisations is not subject to any time limit); and
(b) second, a review of a merger authorisation is not a re-hearing of the matter (whereas a review of other authorisations is a re-hearing of the matter) and, correspondingly, restrictions are imposed on the information, documents and evidence to which the Tribunal may have regard in a review of a merger authorisation (whereas no such restrictions are imposed in a review of other authorisations).
24 With respect to the statutory time period for the review, broadly stated a review of a merger authorisation is required to be completed within 90 days. However, under s 102(1AD) of the CCA, the Tribunal may determine in writing that the matter cannot be dealt with properly within the initial period, either because of its complexity or because of other special circumstances, and that an extended period applies for the review, which consists of the initial period and a further specified period of not more than 90 days. On 31 January 2023, the Tribunal made a determination in writing to that effect such that the period of the present review is 180 days (which ends on 21 June 2023).
25 With respect to the information, documents and evidence to which the Tribunal may have regard in this review, and in accordance with s 102(10) of the CCA, the Tribunal has only had regard to:
(a) information that was referred to in the ACCC’s reasons for making its determination;
(b) information furnished, documents produced or evidence given to the ACCC in connection with the making of its determination; and
(c) the model used by the ACCC as referred to at [9.132] and [9.133] of its reasons for determination which the Tribunal required the ACCC to provide pursuant to s 102(6) of the CCA on 17 March 2023.
26 The information, documents and evidence given to the ACCC in connection with the making of its determination was vast in quantity. The parties placed that vast quantity of material before the Tribunal, although in their written and oral submissions the parties referred to a relatively small part of the material.
27 The evidence given to the ACCC included a number of witness statements of executives of each of Telstra, TPG and Optus, and a number of expert reports prepared on behalf of those parties. The Tribunal has found the witness statements to be very helpful in understanding the commercial context in which the Proposed Transaction arose and the options available to TPG if the Proposed Transaction does not proceed. The witnesses who gave statements are as follows:
(a) on behalf of Telstra:
(i) Andrew Richard Penn, who at the time of giving his statement was the CEO of Telstra, and who made a statement dated 12 August 2022;
(ii) Bart-Jan Sweers, Principal, Economic Modelling at Telstra, who made a statement dated 12 August 2022 and a supplementary statement dated 4 November 2022;
(iii) Christopher George Meissner, Network Engineering Executive – Customer Access at Telstra, who made a statement dated 12 August 2022;
(iv) Michael Graeme Ackland, Group Executive, Consumer & Small Business at Telstra, who made a statement dated 15 August 2022; and
(v) Nicolaos Katinakis, Group Executive for Networks & Information Technology at Telstra, who made a statement dated 15 August 2022 and a supplementary statement dated 9 November 2022;
(b) on behalf of TPG:
(i) Inaki Berroeta Aurrecoechea, the CEO and Managing Director of TPG, who made a statement dated 15 August 2022;
(ii) Yago Lopez, General Manager of Technology Strategy and Innovation at TPG, who made a statement dated 8 November 2022;
(iii) Giovanni Paolo Chiarelli, Chief Technology Officer at TPG, who made a statement dated 8 November 2022; and
(iv) Kieren Paul Cooney, Group Executive, Consumer at TPG, who made a statement dated 8 November 2022;
(c) on behalf of Optus:
(i) Yuen Kuan Moon, the CEO of Singapore Telecommunications Limited (Singtel) – the parent company of Optus – who made a statement dated 19 October 2022;
(ii) Paul O'Sullivan, the Chair of the Board of Directors of Optus, who made a statement dated 19 October 2022;
(iii) Kelly Bayer Rosmarin, the CEO of Optus, who made a statement dated 19 October 2022;
(iv) Kanagaratnam Lambotharan, Vice President of Networks at Optus, who made a statement dated 18 October 2022;
(v) Benjamin White, Managing Director of Wholesale & Strategy and Chief Operating Officer of Enterprise & Business at Optus, who made a statement dated 19 October 2022; and
(vi) Steve Turner, Director of Spectrum Strategy and Management at Optus, who made a statement dated 20 October 2022.
28 In the course of its assessment of the application for authorisation, the ACCC also examined a number of those witnesses and other executives of the parties pursuant to its powers under s 155 of the CCA. The persons examined were: Messrs Penn, Sweers and Katinakis of Telstra; Messrs Berroeta and Lopez and Trent Czinner (Group Executive, Legal and External Affairs) of TPG; Ms Bayer Rosmarin, Mr White, Mr Kanagaratnam and Kent Wu Zeyi (Vice President of the Access Networks Strategy, Planning and Quality) of Optus; and Mr Moon of Singtel.
29 The expert reports given to the ACCC were as follows:
(a) on behalf of Telstra:
(i) reports dated 20 May 2022, 25 July 2022 and 10 November 2022 prepared by Richard Feasey with respect to the likely effects on competition of the Proposed Transaction (the reports did not state Mr Feasey’s academic qualifications, but disclose that he was, amongst other things: the Director of Public Policy at Vodafone plc from 2001 to 2013; an Associate at Frontier Economics Ltd, a London based economic consulting firm, from 2013 to 2017; appointed as a Panel Member of the UK Competition and Markets Authority in 2017; and, since 2021, the Inquiry Chair of that Authority);
(ii) reports dated 27 July 2022 and 10 November 2022 prepared by Aetha Consulting Limited (Aetha), which model the capacity of the mobile networks operated by Telstra and Optus in the RCZ;
(iii) report dated 28 July 2022 prepared by Emma Ihaia with respect to the public benefits and public detriments likely to arise from the Proposed Transaction (the report discloses that: Ms Ihaia is an economist with expertise in regulatory and competition economics, with 25 years of experience applying economic analysis to the telecommunications sector; Ms Ihaia has worked with a number of international consultancies including Charles River Associates, where she was a Principal Economist, and Castalia, where she was a Director in the New Zealand and Pacific Practice; Ms Ihaia holds a Bachelor’s degree and a Master’s degree, both in economics, from the University of Auckland);
(iv) statement dated 30 October 2022 by Michael Robert Strople in relation to MOCN based network sharing, including the differences between MOCN and roaming arrangements, based on Mr Strople’s experience working in Canada; and
(v) statement dated 27 October 2022 by Bruce Rodin in relation to MOCN arrangements as they have developed and operated in Canada, based on Mr Rodin’s experience working in Canada;
(b) on behalf of TPG, reports dated 26 July 2022, 2 November 2022 and 17 November 2022 prepared by Dr Jorge Padilla with respect to the likely effects on competition of the Proposed Transaction (the reports disclose that: Dr Padilla is the Senior Managing Director and the Head of Compass Lexecon EMEA, a global economic consultancy and which is part of FTI Consulting, Inc; Dr Padilla has more than 20 years’ experience as an economic consultant and has taught economics for approximately 30 years at CEMFI (Madrid), Boston University, the Barcelona Graduate School of Economics, King’s College (London) and the Toulouse School of Economics; Dr Padilla earned MPhil and DPhil degrees in economics from the University of Oxford);
(c) on behalf of Optus:
(i) a report dated 24 June 2022 prepared by Cambridge Economic Policy Associates Pty Ltd (CEPA), and authored by Chris Doyle and Dr Jonathan Mirrlees-Black, with respect to the likely effects on competition of the Proposed Transaction (the report discloses that Mr Doyle: an economist with over 25 years’ experience advising clients in the communications space; joined CEPA in March 2022 as Head of Telecoms and Senior Advisor; between 2018-2021 was an economist at Ofcom (the communications regulator of the United Kingdom), having previously held senior positions in economic consulting and academia; and obtained a doctorate in economics from Warwick University, specialising in game theory and industrial organisation; and that Dr Jonathan Mirrlees-Black: has over 25 years’ of experience as an economist and finance professional in infrastructure, as an investment analyst and as an advisor to global infrastructure companies, regulators, international organisations, and private equity investors; is the Director of CEPA’s Sydney office; from 2010-15 was Senior Advisor then Head of Research at RARE Infrastructure, a Sydney-based specialist investor in global listed infrastructure; and holds a doctorate in economics from Oxford University);
(ii) a report dated 26 September 2022 prepared by CEPA, and authored by Chris Doyle, with respect to Telstra’s use of TPG’s spectrum under the Spectrum Authorisation Agreement, including efficiency implications and the effects it may have on competition;
(iii) reports dated 28 June 2022 and 26 October 2022 prepared by Greg Houston of HoustonKemp with respect to the likely effects on competition of the Proposed Transaction (the reports disclose that Mr Houston: is a founding partner of the firm of expert economists, HoustonKemp; over a period of more than thirty years has accumulated experience in the economic analysis of markets; holds a BSc (Hons) in economics from the University of Canterbury);
(iv) a report dated 27 June 2022 prepared by Analysys Mason with respect to the technical and commercial characteristics of the Proposed Transaction;
(v) a report dated 24 October 2022 prepared by Analysys Mason with respect to the mobile network cost effects of the Proposed Transaction; and
(vi) a report dated 25 October 2022, 16 November 2022 and 4 December 2022 prepared by AlixPartners, and authored by Matthew Hunt, with respect to the likely effects on competition and efficiency of the Proposed Transaction (the report discloses that Mr Hunt: is an economist and a Managing Director in the Investigations, Disputes and Risk practice in the London office at AlixPartners; has 22 years of experience acting as an expert and economic advisor in the fields of regulation and competition policy; leads AlixPartners’ EMEA economics practice and the economics work in the telecommunications, media and technology sectors; holds a Masters of Economics from the London School of Economics and Political Science, University of London and a Masters of Physics from the University of Oxford.
30 The Tribunal notes that, on 17 March 2023, the Tribunal refused an application made by the applicants seeking directions to enable the Tribunal to receive additional evidence on the review: see Re Telstra/TPG No 1.5
31 In accordance with a direction of the Tribunal made on 31 January 2023, the parties have filed a joint document identifying all findings on factual matters set out in the ACCC’s reasons for determination that are not contested by the parties on this review. On 11 May 2023, the parties filed an updated version of that document (which is referred to herein as the Joint Document of Factual Findings). The Tribunal has adopted those findings for the purposes of this determination.
32 The Tribunal has also received and had regard to:
(a) the concise statements of facts, issues and contentions (SOFIC) filed on behalf of each of the parties to this proceeding;
(b) written submissions filed on behalf of each of the parties in advance of the hearing;
(c) oral submissions advanced on behalf of each of the parties during the hearing, together with a number of aide memoires and further written submissions provided to the Tribunal during the hearing;
(d) a rejoinder filed by Optus and a response to the rejoinder filed by the applicants after the conclusion of the hearing.
33 The Tribunal considers that the authorisation preconditions stated in s 90(7) of the CCA require the Tribunal to assess the likely competitive effects of, and the public benefits and detriments likely to result from, that conduct which is the subject of the application for authorisation, being the Proposed Conduct. That assessment has been undertaken in light of all relevant circumstances, which includes the MOCN Service Agreement and the Mobile Site Transition Agreement. However, the assessment does not involve weighing the likely competitive effects of, and the public benefits and detriments likely to result from, those other agreements. The Tribunal considers that the contrary approach adopted by the ACCC in its determination, and which is supported by the applicants on this review, is erroneous. Against the possibility that the Tribunal’s understanding of its statutory task is incorrect, the Tribunal has also applied the authorisation preconditions stated in s 90(7) to the Proposed Transaction as a whole. Ultimately, it has reached the same determination on both approaches.
34 Having considered the information, documents and evidence before it and the submissions of the parties, the Tribunal is not satisfied that the Proposed Conduct, being Telstra’s use of TPG’s spectrum licences pursuant to the terms of the Spectrum Authorisation Agreement, would not be likely to substantially lessen competition, or would be likely to result in a benefit to the public that would outweigh the public detriment from that use. In relation to the Tribunal’s assessment of the Proposed Transaction, the evaluation required under s 90(7) is more finely balanced. The Tribunal accepts that the Proposed Transaction has pro-competitive benefits in that it will likely improve the competitive position of TPG in comparison to the likely counterfactuals. However, the Proposed Transaction will also deliver to Telstra material competitive advantages which will have the effect of weakening the competitive position of Optus. Overall, and on balance, the Tribunal is not satisfied that the Proposed Transaction would not be likely to substantially lessen competition, or would be likely to result in a benefit to the public that would outweigh the public detriment from the Proposed Transaction.
35 The Tribunal therefore affirms the determination of the ACCC. These are the Tribunal’s reasons for making that determination.
B. THE PROPOSED TRANSACTION, THE APPLICATION FOR AUTHORISATION AND THE PROPOSED UNDERTAKINGS
36 On 21 February 2022, Telstra and TPG entered into three related agreements: the MOCN Service Agreement, the Spectrum Authorisation Agreement, and the Mobile Site Transition Agreement6 (referred to earlier as the Proposed Transaction). On 28 April 2022, Telstra and TPG entered into an agreement varying the terms of the Proposed Transaction agreements. Each of the agreements are subject to a condition precedent that they do not become binding on the parties unless and until the applicants have received a notice in writing from the ACCC that it does not propose to intervene or seek to prevent the implementation of the agreements or the ACCC, or the Tribunal on review, has granted authorisation in respect of the Proposed Transaction.7
37 At the time of entering into the Proposed Transaction, Telstra and TPG issued a joint announcement to the Australian Securities Exchange in which they described the Proposed Transaction and the commercial benefits that Telstra and TPG believed would result from it. The joint announcement is informative and contained the following statements:
Telstra and TPG Telecom Limited today announced a ground-breaking ten-year regional Multi-Operator Core Network (MOCN) commercial agreement, which will provide significant value to Telstra’s wholesale mobile revenues, while providing TPG Telecom group’s subscribers with 4G and 5G services within a defined coverage zone across regional and urban fringe areas.
Under the innovative deal TPG Telecom will gain access to around 3,700 of Telstra’s mobile network assets, increasing TPG Telecom’s current 4G coverage from around 96 per cent to 98.8 per cent of the population.
Telstra will gain access to TPG Telecom’s spectrum across 4G and 5G, which will allow it to grow its network, increase capacity and continue to provide the country’s largest and fastest network.
Under the MOCN arrangement Telstra will share its Radio Access Network (RAN) for 4G and subsequently 5G services in the defined coverage zone, however both carriers will continue to operate their own core network where key differentiating functionality resides. Telstra will also obtain access to and deploy infrastructure on up to 169 TPG Telecom existing mobile sites, improving coverage for TPG and Telstra customers in the zone. The non-exclusive agreement includes the option for TPG Telecom to request two contract extensions of five years each.
Telstra CEO Andrew Penn said the deal provided significant value to shareholders and customers and was a continuation of Telstra’s strategy to maximise the utilisation and monetisation of its assets.
“This additional spectrum will mean that all Telstra customers will continue to experience Australia’s best and fastest network across the country, in combined 4G and 5G speeds. In particular, the spectrum agreement will ensure that regional and rural customers will now experience faster speeds in more locations on their mobiles.”
TPG Telecom CEO Iñaki Berroeta said the landmark network sharing agreement would significantly expand TPG Telecom’s mobile network footprint in regional Australia and enable growth of its customer base in regional and metropolitan areas.
“It represents a material uplift in the capability of our network and will provide significant value for TPG Telecom shareholders over the medium and long term.
“We will be open for business in regional and rural Australia like never before, offering a 4G network that provides 98.8% population coverage and rapidly growing 5G coverage across the nation.
“The agreement demonstrates best-practice asset utilisation and a commitment to rationalising our operations to deliver a better customer experience, while increasing capital efficiency.
Mr Penn said, “With more people moving to regional areas as a result of COVID, congestion in some areas has increased. This additional spectrum will also ensure that Telstra customers will experience significantly reduced congestion at busy times.
“Telstra’s network has always been and will continue to be the best network – the structure of the deal ensures that we will continue to differentiate in network leadership for our customers in coverage and services.
“We can do that because we will maintain our one million square km competitive advantage in mobile coverage where no other operators have invested. Mobile coverage is often talked about as population coverage, however we all know that it’s the square kilometres of coverage when you travel between towns and cities that also matters. It is the fabric of our mobile network.
“This is critical for customers living and working in those areas. It provides security and safety when travelling long distances on major roads and is only available for our customers travelling through or working or living in those areas.”
Mr Penn said the innovative deal would realise more value from Telstra’s network infrastructure for shareholders while making a very significant contribution to Telstra’s wholesale mobile revenues.
“The deal provides TPG Telecom with the opportunity to access some of our network assets within the defined zone. The access is similar to the way Telstra currently provides wholesale services to its MVNOs and Belong in this zone.
“Similar to monetising our passive infrastructure, it allows Telstra to have an innovative way of monetising some of our active mobile infrastructure, in areas where the population coverage is much smaller and more challenging in terms of returns and further investment and where there are already a number of competitors.
“Additional scale from this agreement therefore supports return on invested capital in these areas and makes ongoing investment in the network and innovation more sustainable.”
Mr Berroeta said the agreement was a win for TPG customers who would have access to a significant part of the best regional network in Australia.
“The deal will give TPG Telecom’s consumer, enterprise and wholesale customers seamless access to a national network. This will enable TPG Telecom’s Vodafone, TPG, iiNet, Lebara and felix brands to improve their services for regional Australians.”
Access to this additional coverage will be automatic for all of TPG Telecom group’s customers and will appear to them as being provided by their current TPG Telecom group provider.
TPG Telecom will continue to operate its own 3G, 4G and 5G networks in metropolitan areas reaching around 80 per cent of the population, which includes its network infrastructure sharing arrangement with Optus in those areas.
TPG Telecom will decommission the 725 mobile sites it currently operates within the MOCN coverage area, reducing environmental impact, energy consumption, operating costs and future capex.
38 The remainder of this section provides a description of the three agreements which comprise the Proposed Transaction.
39 Under the MOCN Service Agreement, Telstra agrees to use its radio access network to supply TPG with 4G and 5G services in the RCZ.8 TPG will thereby obtain access to Telstra’s radio access network within the RCZ comprising approximately 3,700 Telstra mobile sites. The MOCN Service will enable TPG to provide mobile network telecommunications services to TPG services in operation (SIOs) in the RCZ. TPG will also be able to provide the following services within the RCZ by relying on services provided by Telstra under the MOCN Service Agreement:
(a) a fixed wireless service using combined 3.6 GHz spectrum;9
(b) Narrow Band Internet of Things (NBIoT) using Telstra 700 MHz spectrum band;10 and
(c) capability on a mobile internet service used as a back-up (Fixed NBN Fallback).11
40 A fixed wireless service is a retail consumer fixed broadband data service for use at a single premises in the RCZ.
41 NBIoT is a service that enables the use of relatively low-power machine communications for uses other than consumer voice or data.
42 Fixed NBN Fallback capability enables TPG to supply its customers who are located at premises in the RCZ and who have a primary fixed broadband service supplied by NBN with a failover mobile broadband data service which will operate during periods of outage of the fixed NBN service.
43 Both TPG and Telstra will continue to operate their own mobile core networks. TPG will remain responsible for enhancements, upgrades, interconnection arrangements, and the acquisition of any goods or services from third parties for the purpose of developing the TPG mobile core network.12
44 The fees payable by TPG to Telstra for the MOCN service will include:13
(a) a fixed annual charge for access, payable in equal quarterly instalments;
(b) charges dependant on the number of services in operation TPG is servicing;
(c) a per GB charge for data consumed by TPG’s use of the MOCN service in the RCZ; and
(d) charges for fixed wireless services in operation, NBIoT services in operation, and Fixed NBN Fallback services in operation.
45 Further details of the relevant fees and charges payable are as follows:14
(a) [REDACTED];
(b) [REDACTED];
(c) [REDACTED];
(d) [REDACTED];
(e) [REDACTED];
(f) [REDACTED];
(g) [REDACTED].
46 The initial term of the MOCN Service Agreement is 10 years, with TPG having two further five-year options to extend the agreement.15 A 36-month “Transition-Out Period” will come into effect on the expiry or termination of the MOCN Service Agreement. During this period, TPG will have the discretion to nominate an earlier date for ceasing use of the services.16 Telstra will be required to continue supplying the services until the end of the Transition-Out Period.17
47 The MOCN Service Agreement contains non-discrimination provisions that require Telstra to ensure that TPG end users and Telstra “comparison customers” receive equal treatment, including in relation to network performance and quality of service.18 The non-discrimination obligation will apply to current services and to the technical upgrade or evolution of the shared radio access network and 4G and 5G standards.19 Telstra “comparison customers” are defined as Telstra customers who are end users on a retail consumer grade plan and who do not have any other type of plan (including in respect of any enterprise grade product or “special service”). There are exceptions to the non-discrimination provisions, including the following:
(a) TPG will not have access to 5G-enabled sites until 6 months after Telstra has activated the sites for 5G.20 The six-month delay will apply on a site-by-site basis, with the effect that TPG’s access to 5G in the RCZ would be “staggered”. For Telstra 5G sites activated prior to the commencement of the MOCN Service Agreement, the six-month period will commence from the date of site activation, rather than the commencement of the MOCN Service Agreement.
(b) As noted in the preceding paragraph, the non-discrimination provisions do not apply in respect of Telstra customers who have an enterprise grade product or “special services”.21
(c) Fixed wireless access will only be supplied to TPG over 3.6 GHz spectrum on a 5G standalone basis, while the fixed wireless service supplied to Telstra comparison customers may use 3.6 GHZ Spectrum on a 5G non-standalone basis and may use other spectrum bands.22
(d) The provisions do not apply to NBIoT capability.23
48 Telstra acknowledges that it is not the exclusive supplier of MOCN services or other network or access services to TPG, and TPG is not restricted from building and operating its own mobile telecommunications network within or outside the RCZ, or procuring other network, access or roaming services from third parties.24 TPG also acknowledges that it is not Telstra's exclusive customer of MOCN services and Telstra is not restricted from selling services equivalent to the MOCN services (or any other network or access service) to any third party.25
49 [REDACTED].26 [REDACTED].27 [REDACTED].28
50 The MOCN Service Agreement stipulates that it does not include any mobile technology generation beyond 5G (including 6G or 7G) and that the addition [REDACTED] [REDACTED] of a technology generation is subject to the “Change Management Process” which is specified in the Agreement.29 Under the Change Management Process, TPG may request a change, including the addition of a technology generation ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]).30 Telstra will be required to act reasonably and negotiate with TPG in good faith regarding any such change.31 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].32
51 There is no obligation on either Telstra or TPG to acquire spectrum at an auction, use any spectrum it acquires at an auction, or automatically include any spectrum it acquires at an auction in the scope of the MOCN services.33 [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]), [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. 34
Spectrum Authorisation Agreement
52 Under the Spectrum Authorisation Agreement, TPG will authorise Telstra to operate radiocommunications devices utilising part of TPG’s 4G and 5G spectrum (including TPG’s spectrum in 700 MHz, 850 MHz, 2.1 GHz and 3.6 GHz bands) within the RCZ for use on the MOCN. Telstra will also be authorised to use certain TPG spectrum beyond the RCZ. Telstra is not authorised to use TPG spectrum in metropolitan areas.35
53 Telstra will be required to pay TPG quarterly spectrum use fees, with discounts calculated to account for any restricted spectrum (being spectrum that is withdrawn, or which TPG and Telstra agree is affected by incumbency or interference issues).36
54 TPG’s spectrum that is the subject of the Spectrum Authorisation Agreement is summarised as follows:
(a) in respect of the 700MHz band, 2 x 10 MHz;
(b) in respect of the 850 MHz band, 2 x 5 MHz;
(c) in respect of the 2100 MHz band, 2 x 5 MHz; and
(d) in respect of the 3600 MHz band, 20-45 MHz.37
55 Telstra and TPG have agreed to cooperate to re-stack their 850 MHz spectrum holdings beyond the outer boundaries of the Regional Coverage Zone, in which Telstra is currently the only provider of services.38
56 The specific areas of Australia in which TPG will authorise Telstra to use its spectrum varies by spectrum band. The spectrum authorisation for the 700 MHz band covers a significant portion of Australia’s landmass, while other bands subject to the spectrum authorisation, such as the 3.6 GHz band, will cover smaller proportions of the country due to the available licences.39
57 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], TPG will authorise Telstra to use its spectrum in the RCZ and areas beyond that zone, [REDACTED] [REDACTED]-[REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]) [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED]).40
58 There will be no restraints on TPG bidding on any new spectrum allocation, or any obligation on TPG to offer to include any new spectrum band within the scope of the Spectrum Authorisation Agreement.41
59 [REDACTED].42
60 On termination or expiry of the MOCN Service Agreement, the parties have certain rights to terminate the Spectrum Authorisation Agreement, but are not obligated to do so. Specifically:
(a) on termination or expiry of the MOCN Service Agreement, Telstra is entitled to terminate the Spectrum Authorisation Agreement [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED];43
(b) [REDACTED]. 44
Mobile Site Transition Agreement
61 The purpose of the Mobile Site Transition Agreement is to set out the principles on which TPG will enable Telstra to access, use and occupy TPG Space at TPG Owned Sites or at TPG Licensed Sites for the purpose of enabling Telstra to install mobile telecommunications equipment in/on the TPG Space in place of TPG.45 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED]- [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].46
62 Telstra and TPG will be required to negotiate in good faith the ability for Telstra to access and deploy infrastructure on the 169 TPG Transitioning Sites which are primarily inside the RCZ.47 In the case of TPG Licensed Sites, access may be via the novation of the existing licence with TPG or the grant of a new licence from the licensor.48
63 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].49 Telstra will [REDACTED] pay TPG [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] under the Mobile Site Transition Agreement, including to access and deploy infrastructure on the TPG Transitioning Sites and to assume TPG’s obligations under the transferred site licences.50
64 The 169 TPG Transitioning Sites are a subset of 749 mobile sites TPG is decommissioning in in the RCZ.51
65 If the MOCN Service Agreement expires or is terminated, TPG can request re-installation of its equipment on facilities at one or more sites.52 Telstra will be required to use commercially reasonable endeavours to facilitate TPG’s access to TPG sites, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].53
The application for authorisation
66 On 23 May 2022, Telstra and TPG lodged a joint application for authorisation with the ACCC. The scope of the application for authorisation differed from that contemplated by the conditions precedent to the Proposed Transaction agreements. The executive summary to the application stated the following (citations omitted):
The commercial agreements
Telstra and TPG have agreed to share active mobile network infrastructure and spectrum in certain parts of regional Australia in response to the challenges of building and maintaining mobile network infrastructure in these regional areas of low population density.
TPG and Telstra will continue to operate their own networks in metropolitan areas where the vast majority (around 81.4%) of the population resides. Furthermore, both TPG and Telstra will continue to operate their own mobile core networks where the key service differentiation and sensitive functions occur.
On 21 February 2022, Telstra Corporation Limited (Telstra) and TPG Telecom Limited (TPG) entered into the following related commercial agreements:
• MOCN Service Agreement dated 17 February 2022 (MOCN Agreement);
• Spectrum Authorisation Agreement – MOCN Area dated 17 February 2022 (Spectrum Authorisation); and
• Mobile Site Transition Agreement dated 17 February 2022 (Site Agreement),
(together, the Agreements) (Proposed Transaction).
The Agreements are in respect of a Multi-Operator Core Network (MOCN) commercial arrangement pursuant to which Telstra will share its Radio Access Network (RAN) with TPG and supply 4G and 5G services in certain regional and urban fringe areas which comprise around 17% of the Australian population coverage (in the 81.4% - 98.8% population coverage area) (17% Regional Coverage Zone). The Agreements are commercially and legally interdependent.
To support the shared use of the MOCN in the 17% Regional Coverage Zone, TPG will authorise certain spectrum it currently owns and is unutilised or underutilised to Telstra in the 17% Regional Coverage Zone to be pooled with Telstra’s spectrum and made available to both Applicants. Telstra will also be authorised to use certain spectrum beyond the 17% Regional Coverage Zone, i.e. only in areas beyond the 98.8% of the Australian population. The initial term of the MOCN Agreement is 10 years and TPG has two options to extend the agreement by 5 years.
Application for Merger Authorisation
Pursuant to s 68A of the Radiocommunications Act 1992 (Cth) (Radiocommunications Act), TPG’s grant of authorisation to Telstra to use TPG spectrum within the meaning of s 68(1) is deemed to be an acquisition within the meaning of s 50 of the Competition and Consumer Act 2010 (Cth) (CCA) and is capable of merger authorisation under Part VII.
Telstra and TPG (together, the Applicants) seek merger authorisation for the authorisation of use of spectrum (under the Spectrum Agreement) which is deemed to be an acquisition within the meaning of s 50, CCA (Authorisation) on the basis that:
• the authorisation of spectrum would not have the effect, and would not be likely to have the effect, of substantial lessening of competition (SLC) in any market; and
• the public benefits associated with the authorisation of spectrum would result, or be likely to result, in a benefit to the public, and the benefit would outweigh any detriment to the public that would result, or be likely to result, from the Proposed Transaction.
To the extent the ACCC considers it needs to have regard to the Proposed Transaction as a whole, this Application has provided information that would enable this assessment.
The Applicants request the ACCC to authorise the use of spectrum as set out in the Spectrum Agreement that is deemed to be an acquisition under s 50, CCA. The evidence contained in this Application is provided in support of the Authorisation.
The Applicants are not seeking authorisation for co-location agreements under the Site Agreement that may, but are unlikely to, involve the transfer of leases or licenses.
67 The following matters can be noted about the above statements.
68 First, Telstra and TPG have entered into three agreements, the MOCN Service Agreement, the Spectrum Authorisation Agreement and the Mobile Site Transition Agreement as part of the one commercial transaction.
69 Second, Telstra and TPG have sought authorisation for the use of spectrum under the Spectrum Authorisation Agreement “which is deemed to be an acquisition within the meaning of s 50”, but not in respect of the entry into and implementation of the MOCN Service Agreement or the Mobile Site Transition Agreement. The applicants stated expressly that they were seeking “merger authorisation” in respect of Telstra’s use of TPG spectrum.
70 Third, Telstra and TPG have sought authorisation on the alternative bases that:
(a) the authorisation of the use of spectrum would not have the effect, and would not be likely to have the effect, of substantial lessening of competition in any market; and
(b) the public benefits associated with the authorisation of the use of spectrum would result, or be likely to result, in a benefit to the public, and the benefit would outweigh any detriment to the public that would result, or be likely to result, from the “Proposed Transaction”.
71 Fourth, Telstra and TPG provided information in relation to the Proposed Transaction as a whole to enable the ACCC to assess that information “to the extent the ACCC considers it needs to have regard to the Proposed Transaction as a whole”.
72 It is apparent from the foregoing that the applicants made a deliberate decision to seek authorisation only for Telstra’s use of TPG’s spectrum under the Spectrum Authorisation Agreement, and not to seek authorisation for the conduct comprising entering into and giving effect to the MOCN Service Agreement and the Mobile Site Transition Agreement. By confining the scope of the authorisation in that manner, the authorisation sought by the applicants satisfies the definition of a “merger authorisation” in s 4 of the CCA. As discussed below, one advantage of a merger authorisation in comparison to other authorisations is that a merger authorisation is subject to statutory time limits imposed on the decision-makers (the ACCC and the Tribunal on review).
73 As noted above, the ACCC refused to grant authorisation for the reason that the ACCC was “not satisfied, in all the circumstances, that the Proposed Transaction would not be likely to substantially lessen competition, or would be likely to result in a benefit to the public that would outweigh the public detriment from the Proposed Transaction”. It is apparent that, in reaching its determination, the ACCC assessed the likely effects of the Proposed Transaction as a whole rather than the likely effects of the conduct for which authorisation was sought, being Telstra’s use of TPG’s spectrum under the Spectrum Authorisation Agreement (defined earlier as the Proposed Conduct).
74 At the commencement of this proceeding, the Tribunal sought confirmation from the parties that:
(a) the application for authorisation was only in respect of the Proposed Conduct, namely Telstra’s use of TPG’s spectrum under the Spectrum Authorisation Agreement; and
(b) as such, the application was in respect of a “merger authorisation” as defined in s 4 of the CCA.
75 All parties gave that confirmation. The SOFICs filed on behalf of the parties also proceed on that basis. The proceeding has therefore been conducted on that basis.
76 Telstra’s SOFIC (adopted by TPG) nevertheless advanced the contention that:
When referring to the Proposed Conduct below, it encompasses the effects, benefits and detriments of the Proposed Transaction, because the Relevant Agreements are interlinked and would not exist without the Spectrum Agreement…
77 On the basis of that contention, the applicants’ SOFICs and written submissions proceeded on the basis that, because the three agreements comprising the Proposed Transaction are interdependent, the Tribunal’s statutory task under s 90(7) is to consider the likely effects of the Proposed Transaction as a whole.
78 For the reasons explained below, the Tribunal disagrees with that characterisation of its statutory task. It can be accepted that the Proposed Transaction agreements have been entered into by the parties as part of the one transaction. However, that premise does not lead to the conclusion that the effects, benefits and detriments of the Proposed Conduct, being Telstra’s use of TPG’s spectrum, encompasses the effects, benefits and detriments of the Proposed Transaction. It was open to the applicants to seek authorisation of the conduct comprising entering into and implementing the Proposed Transaction agreements, but the applicants elected to seek authorisation only in respect of Telstra’s use of TPG’s spectrum under the Spectrum Authorisation Agreement. The Tribunal considers that s 90(7) of the CCA requires it to assess only the effects of the conduct for which authorisation has been sought. The effects of the MOCN Service Agreement and the Mobile Site Transition Agreement cannot be characterised as an effect of Telstra’s use of TPG’s spectrum under the Spectrum Authorisation Agreement for the purposes of s 90(7).
79 The Tribunal has therefore applied the tests stated in s 90(7) to the Proposed Conduct. Against the possibility that the Tribunal’s understanding of its statutory task is incorrect, the Tribunal has also applied the tests stated in s 90(7) to the Proposed Transaction as a whole.
80 In connection with their application for review, the applicants have offered to give the ACCC a joint undertaking pursuant to s 87B of the CCA in relation to the Proposed Transaction, and TPG has offered to give the ACCC a separate undertaking relating to its mobile site infrastructure.
81 The Tribunal is empowered by s 102(1) to accept the undertakings on this review, if it considers it appropriate to do so. Section 102(1) stipulates that, for the purposes of the review, the Tribunal may perform all of the functions and exercise all of the powers of the ACCC. It is implicit in that section that the functions and powers of the ACCC able to be exercised by the Tribunal on the review are the functions and powers able to be exercised by the ACCC in making its determination on the authorisation application. Those powers of the ACCC include accepting an undertaking given by a person under s 87B in connection with a merger authorisation. Relevantly, s 87B(1A) provides that the ACCC may accept a written undertaking given by a person for the purposes of s 87B in connection with a merger authorisation. That provision is reinforced by s 88(4) which provides that the ACCC may grant a merger authorisation on the condition that a person must give, and comply with, an undertaking to the ACCC under section 87B.
82 The proposed joint undertaking contains two substantive promises.
83 The first substantive promise is contained in cl 4. It provides as follows:
4.1. During any period in which the Spectrum Authorisation is in effect, the Undertaking Signatories undertake that they will:
(a) implement and give full force and effect to the MOCN Agreement and Site Agreement, in each case in accordance with their terms;
(b) not make any material amendment to the MOCN Agreement or Site Agreement except:
(i) strictly to the extent any such amendment is contemplated by, and then made in accordance with, the existing terms of those agreements, including a Change made in accordance with the Change Management Process in Part B of Schedule 6 of the MOCN Agreement or a change to the Site Transition Plan agreed in accordance with clause 4.1 of the Site Agreement; or
(ii) otherwise only if the amendment does not take effect until after the ACCC has confirmed in writing that it does not object to the amendment being made.
4.2. If the MOCN Agreement is terminated for any reason, TPG undertakes to exercise its right to terminate the Spectrum Authorisation in accordance with clause 11.1(e) of the Spectrum Authorisation.
84 It can be seen that the effect of the undertaking in cl 4 is twofold. First, it requires the applicants to implement and to continue to give full force and effect to the MOCN Service Agreement and the Site Agreement and to not materially amend those agreements except in accordance with their terms or otherwise with the prior consent of the ACCC. Second, it requires the applicants to terminate the Spectrum Authorisation Agreement if the MOCN Service Agreement is terminated at any time.
85 The second substantive promise is contained in cl 5. It provides as follows:
The Undertaking Signatories commit to cease giving effect to the Agreements, except to the extent necessary to give effect to the Transition-Out Period, unless by 8 years from the date that the original authorisation of the Spectrum Transaction takes effect:
(a) the Undertaking Signatories have received, either unconditionally or on terms and conditions that are acceptable to both of the Undertaking Signatories acting reasonably, a notice in writing from the ACCC stating, or stating to the effect, that the ACCC does not propose to intervene or seek to prevent the parties from continuing to give effect to the Agreements; or
(b) the ACCC or Australian Competition Tribunal has made a final determination to grant authorisation pursuant to Part VII, Division 1 of the Act, the effect of which is to grant the Undertaking Signatories authorisation to continue giving effect to the Agreements:
(i) on an unconditional basis; or
(ii) subject to conditions which are, in the reasonable opinion of the Undertaking Signatories, acceptable.
86 The undertaking proffered in cl 5 was also proffered by the applicants to the ACCC, but was not accepted by the ACCC. The effect of the undertaking in cl 5 is that the applicants would be required to cease giving effect to the Proposed Transaction agreements (save for the transition-out arrangements) unless within 8 years of authorisation the applicants have received informal approval of the Agreements from the ACCC, or formal authorisation from the ACCC or the Tribunal.
87 The TPG undertaking contains one substantive promise contained in cl 3 as follows:
3.1 TPG commits to refrain from:
(a) terminating any licence or lease (a Licence) pursuant to which TPG is granted access to one or more of the 300 mobile sites (Retained Site) set out in Annexure A, each of which is within the Regional Coverage Zone; or
(b) taking any action that causes TPG to commit a breach of any Licence that would entitle the licensor or landlord of a Retained Site (Licensor) to terminate the Licence,
unless
(c) the Licensor is in breach of the Licence and, in consequence, TPG is entitled to terminate the Licence;
(d) the ACCC consents in writing to TPG terminating a Licence or taking any action that would entitle the Licensor to terminate the Licence; or
(e) a period of eight years has elapsed since the original authorisation of the Proposed Transaction took effect.
3.2 For the avoidance of doubt, nothing in this Undertaking:
(a) requires TPG to renew a Licence that expires;
(b) prevents a Licensor from terminating a Licence; or
(c) prevents TPG from terminating a Licence for reasons outside of its control or where required by law.
88 As noted above, TPG currently has access to 749 mobile sites in the RCZ. If the Proposed Transaction proceeds, TPG plans to transition 169 of those sites to Telstra (pursuant to the Mobile Site Transition Agreement) and also plans to decommission the remaining 580 sites to which it has access in the RCZ. TPG offered the undertaking, which relates to 300 of the 580 sites proposed to be decommissioned by TPG, to address a concern stated by the ACCC that, following the decommissioning of sites, the threat of future network expansion by TPG in the RCZ will be diminished.
C. STATUTORY FRAMEWORK FOR THE TRIBUNAL’S REVIEW
89 The statutory framework governing the grant of authorisations by the ACCC under Pt VII of the CCA, and the review by the Tribunal of ACCC authorisation determinations under Pt IX of the CCA, were amended in material ways by the Competition and Consumer Amendment (Competition Policy Review) Act 2017 (Cth) (2017 Amendment Act). Those amendments, including the relevant legislative history, were outlined in Re Telstra/TPG (No 1).54
90 As stated in the Explanatory Memorandum accompanying the Competition and Consumer Amendment (Competition Policy Review) Bill 2017 (2017 Explanatory Memorandum), the amendments largely implemented the recommendations of the Competition Policy Review chaired by Prof Ian Harper (referred to as the Harper Review). While many aspects of the pre-existing statutory framework governing authorisations were maintained, two significant changes were made to the authorisation regime in order to simplify it:
(a) first, a single authorisation application could be made for a single business arrangement or transaction (with the application specifying the conduct that was the subject of the application for authorisation); and
(b) second, the ACCC was empowered to grant authorisation on the basis that the conduct would not be likely to substantially lessen competition (in addition to the pre-existing basis that the likely benefits of the conduct would outweigh the likely detriments).55
91 To a large extent, there was no dispute between the parties as to the relevant statutory requirements governing the grant of authorisation by the ACCC under Pt VII of the CCA and the review by the Tribunal of ACCC authorisation determinations under Pt IX of the CCA. The following description of those requirements is, accordingly, uncontroversial. As already foreshadowed, however, there is one significant legal controversy that arises on this review. It concerns the proper application of the statutory preconditions for authorisation in s 90(7) of the CCA in respect of the conduct for which authorisation has been sought by the applicants. The ACCC determination, and the submissions of the applicants on this review, proceed on the basis that the task of the Tribunal is to assess whether the Proposed Transaction as a whole, rather than the Proposed Conduct (which comprises only part of the Proposed Transaction), satisfies the statutory preconditions for authorisation. For the reasons explained below, the Tribunal considers that such an approach is erroneous.
92 Consistently with s 42(1) of the CCA, the following discussion of the statutory framework for the Tribunal’s review has been written by the presidential member of the Tribunal.
93 Following its amendment by the 2017 Amendment Act, s 88 of the CCA, within Div 1 of Pt VII, relevantly provides as follows:
Granting an authorisation
(1) Subject to this Part, the Commission may, on an application by a person, grant an authorisation to a person to engage in conduct, specified in the authorisation, to which one or more provisions of Part IV specified in the authorisation would or might apply.
Note: For an extended meaning of engaging in conduct, see subsection 4(2).
Effect of an authorisation
(2) While the authorisation remains in force, the provisions of Part IV specified in the authorisation do not apply in relation to the conduct to the extent that it is engaged in by:
(a) the applicant; and
(b) any other person named or referred to in the application as a person who is engaged in, or who is proposed to be engaged in, the conduct; and
(c) any particular persons or classes of persons, as specified in the authorisation, who become engaged in the conduct.
Conditions
(3) The Commission may specify conditions in the authorisation. Subsection (2) does not apply if any of the conditions are not complied with.
(4) Without limiting subsection (3), the Commission may grant a merger authorisation on the condition that a person must give, and comply with, an undertaking to the Commission under section 87B.
Single authorisation may deal with several types of conduct
(5) The Commission may grant a single authorisation for all the conduct specified in an application for authorisation, or may grant separate authorisations for any of the conduct.
…
94 Section 90(1) stipulates that, in respect of an application for an authorisation, the ACCC shall either make a determination in writing granting such authorisation as it considers appropriate or make a determination in writing dismissing the application. Following amendments made by the 2017 Amendment Act, s 90(7) relevantly provides as follows:
The Commission must not make a determination granting an authorisation under section 88 in relation to conduct unless:
(a) the Commission is satisfied in all the circumstances that the conduct would not have the effect, or would not be likely to have the effect, of substantially lessening competition; or
(b) the Commission is satisfied in all the circumstances that:
(i) the conduct would result, or be likely to result, in a benefit to the public; and
(ii) the benefit would outweigh the detriment to the public that would result, or be likely to result, from the conduct;
…
95 The expressions “to engage in conduct” or “engaged in conduct” in s 88, and the word “conduct” in s 90, are defined by s 4(2) of the CCA as follows:
(2) In this Act:
(a) a reference to engaging in conduct shall be read as a reference to doing or refusing to do any act, including the making of, or the giving effect to a provision of, a contract or arrangement, the arriving at, or the giving effect to a provision of, an understanding or the engaging in of a concerted practice;
(b) a reference to conduct, when that expression is used as a noun otherwise than as mentioned in paragraph (a), shall be read as a reference to the doing of or the refusing to do any act, including the making of, or the giving effect to a provision of, a contract or arrangement, the arriving at, or the giving effect to a provision of, an understanding or the engaging in of a concerted practice;
…
96 The 2017 Explanatory Memorandum confirms that the purpose of the amendments made by the 2017 Amendment Act was to significantly simplify the authorisation provisions by removing separate provisions applicable to specific types of authorisations, and instead including a single provision under which conduct may be authorised (s 88) and a single test for authorisation (s 90).56
97 The amended authorisation regime in Pt VII of the CCA serves two important purposes. First, it enables market participants to obtain legal certainty that their proposed business conduct will not contravene the prohibitions in Pt IV of the CCA by satisfying the ACCC that the proposed conduct would not have the effect or would not be likely to have the effect of substantially lessening competition. Second, in circumstances where proposed business conduct may contravene the prohibitions in Pt IV of the CCA, it enables market participants to gain exemption from the prohibitions by satisfying the ACCC that the conduct would result, or be likely to result, in a benefit to the public and the benefit would outweigh the detriment to the public that would result, or be likely to result, from the conduct. As observed by the Harper Review:
Competition is desirable not for its own sake but because, in most circumstances, it improves the welfare of Australians by increasing choice, diversity and efficiency in the supply of goods and services. In other words, competition is a means to an end. In some circumstances, arrangements that lessen competition may nonetheless produce public benefits that outweigh the detriment resulting from the lessening of competition.57
98 It can be seen that s 88(1) empowers the ACCC to grant an authorisation to a person to engage in conduct, specified in the authorisation, to which one or more provisions of Pt IV specified in the authorisation would or might apply. The conduct may involve entering into and implementing more than one contract, being contracts to which one or more provisions of Pt IV may apply (including s 50). The effect of s 88(2) is that, while the authorisation remains in force, those provisions of Pt IV do not apply to the conduct (specified in the authorisation) to the extent it is engaged in by the applicant for authorisation, any other person named or referred to in the application as a person who is engaged in or who is proposed to be engaged in the conduct, and any other particular persons or classes of persons, as specified in the authorisation, who become engaged in the conduct. The plain meaning of s 88(2) is confirmed by the 2017 Explanatory Memorandum: an authorisation provides protection for the conduct, and in respect of the statutory prohibitions in Pt IV, specified in that authorisation – but the protection does not extend to conduct not specified in the authorisation or to provisions of Pt IV that may also apply to the conduct but which have not been specified in the authorisation.58
99 The statutory preconditions for the grant of authorisation specified in s 90(7) are directed to the conduct that is the subject of the application for authorisation under s 88. Section 90(7) prohibits the grant of authorisation unless the ACCC is satisfied that the conduct would not have the effect, or would not be likely to have the effect, of substantially lessening competition or that the conduct would result, or be likely to result, in a net benefit to the public. The word “satisfied” in the context of an administrative decision is not amenable to the application of an evidentiary burden of proof, such as balance of probabilities.59 The absence of an evidentiary burden of proof, however, does not mean that there is an absence of a legal standard of satisfaction. In respect of s 90(7), satisfaction requires that the ACCC reach an affirmative belief that the conduct would not have the effect, or would not be likely to have the effect, of substantially lessening competition or that the conduct would result, or be likely to result, in a net benefit to the public. The meaning of the competition test for the grant of authorisation stated in s 90(7)(a), and the net public benefit test for the grant of authorisation stated in s 90(7)(b), are discussed further below.
100 As discussed in Re Telstra/TPG (No 1),60 significant changes were also made to the authorisation regime governing merger transactions. The 2017 Amendment Act repealed the formal merger clearance and authorisation provisions contained in Div 3 of Pt VII. Merger transactions are now subject to the general authorisation process in s 88, with the decision-maker at first instance being the ACCC, and are subject to the same statutory preconditions for authorisation stated in s 90(7). However, the 2017 Amendment Act enacted a limited number of specific requirements for “merger authorisations” which differ from other authorisations, including that:
(a) if the ACCC does not determine an application for a merger authorisation within a 90 day period (which may be extended under s 90(12)), the ACCC is taken to have refused the application (s 90(10B));
(b) a review by the Tribunal of an ACCC determination in relation to a merger authorisation must be completed within 90 days (s 102(1AC), which may be extended in certain circumstances; and
(c) a review by the Tribunal of an ACCC determination in relation to a merger authorisation is not a re-hearing (s 101(2)) and restrictions are imposed on the information, documents and evidence to which the Tribunal may have regard (s 102(8) to (10)).
101 As explained in the 2017 Explanatory Memorandum, shorter timeframes are set for determining merger authorisations compared with non-merger authorisations because of the commercial sensitivity of merger authorisations.61 The same rationale applies in respect of a review by the Tribunal of a merger authorisation. The 2017 Explanatory Memorandum included the following explanation:
The limitations on the information that may be considered by the Tribunal appropriately balance the interests of all parties to a review of a merger authorisation matter. In particular, they are intended to ensure that applicants for merger authorisation provide the Commission with all relevant material at the time of the application, and do not delay production of that material until later in the process or until Tribunal review. The limitations also facilitate the Tribunal conducting its review expeditiously, given the time sensitive nature of merger transactions.62
102 As set out earlier in these reasons, the expression “merger authorisation” is defined in s 4 of the CCA as follows:
merger authorisation means an authorisation that:
(a) is an authorisation for a person to engage in conduct to which section 50 or 50A would or might apply; but
(b) is not an authorisation for a person to engage in conduct to which any provision of Part IV other than section 50 or 50A would or might apply.
103 It can be seen that the expression “merger authorisation” is defined as a specific category of authorisation by reference to the conduct that is the subject of the application. It is confined to an application for authorisation in respect of conduct to which ss 50 or 50A would or might apply but not any other conduct to which any other provision of Pt IV would or might apply. Thus, if a person applies for authorisation of a business transaction that involves conduct to which s 50 would or might apply (being an acquisition of shares or assets) and that also involves conduct to which other provisions of Pt IV would or might apply (such as an accompanying service agreement or joint venture agreement), the authorisation is not a “merger authorisation” within the meaning of the CCA.
104 Significantly, though, whether a particular transaction is or is not a “merger authorisation”, the application for authorisation is made under s 88(1), the effect of authorisation is as stated in s 88(2), and the preconditions for the grant of authorisation are as stated in s 90(7).
105 In the present matter, the applicants sought authorisation only for Telstra’s use of TPG’s spectrum under the Spectrum Authorisation Agreement, and did not seek authorisation for the conduct comprising entering into and giving effect to the MOCN Service Agreement and the Mobile Site Transition Agreement. By confining the scope of the authorisation in that manner, the authorisation sought by the applicants satisfied the definition of a “merger authorisation” in s 4 of the CCA. This has the effect that the application for authorisation is subject to, amongst other things, statutory time limits.
The review of the ACCC’s determination
106 Section 101 of the CCA, within Div 1 of Pt IX, provides that a person dissatisfied with a determination by the ACCC under Div 1 of Pt VII may apply to the Tribunal for a review of the determination. On a review of a determination of the ACCC in relation to an application for an authorisation, the Tribunal may make a determination affirming, setting aside or varying the determination and, for the purposes of the review, may perform all the functions and exercise all the powers of the ACCC: s 102(1). A determination by the Tribunal affirming, setting aside or varying a determination of the ACCC is taken to be a determination of the ACCC: s 102(2).
107 For the reasons explained in Re Telstra/TPG (No 1),63 in conducting a review of a merger authorisation determination by the ACCC, the Tribunal is not conducting a re-hearing (see s 101(2)). Rather, the Tribunal must conduct a review of the ACCC’s determination, making its own decision with respect to the application of the applicable statutory criteria in s 90(7) having regard to only that material that is enumerated in s 102(10).
108 Although the Tribunal’s review is not a re-hearing in the sense of conducting a de novo review, the nature of the task being performed by the Tribunal on the review of a merger authorisation determination are largely the same as on a review of non-merger authorisations. The Tribunal must “make its own findings of fact and reach its own decision as to whether authorisation should be granted or not and, if so, any conditions to which it is to be subject”.64 That function is not performed by considering “whether the ACCC was right or wrong in the conclusion it reached or whether it could have better formulated its determination”.65 Further, the role of the Tribunal in conducting the review is not confined by the issues raised by the parties to the review and the Tribunal must determine itself whether the statutory test for authorisation is satisfied. However, as observed by the Tribunal (Deane J, Mr J Shipton and Mr J Walker) in Re Herald & Weekly Times Ltd (Media Council of Australia),66 “[t]he published reasons for determination of the Commission may, in an appropriate case, prove a convenient reference point for defining the matters which are truly in dispute between all or any of the Commission, the applicants, and other parties represented, or interested, in the proceedings”. Further, where the parties agree with factual findings made by the ACCC in its determination, ordinarily the Tribunal need not itself examine the facts in detail. As explained by the Tribunal (von Doussa J, Dr B Aldrich, Prof D Round) in Re 7-Eleven Stores Pty Ltd:67
In curial proceedings based on the adversarial system, the role of a court is to determine issues identified by the parties, usually in pleadings. Proceedings before the Tribunal are not adversarial in nature, and the role of the Tribunal is not merely to resolve issues in dispute between the parties. It is an administrative tribunal with a much wider role. It is required to determine whether anti-competitive conduct or anti-competitive provisions in a contract, arrangement or understanding that would otherwise be unlawful, should, in the public interest, be authorised because the public benefit outweighs the detriment constituted by any lessening of competition. Determinations of the Tribunal are likely to impact on the commercial interests of many people who are not participants in the proceedings before the Tribunal.
Notwithstanding the positions taken by the parties in this case, the Tribunal in the exercise of its statutory functions, must consider each of the issues arising under [the applicable statutory provisions] which precede a consideration of the terms and duration of the further authorisation granted by the determinations under review. On these essential steps, the Tribunal must reach its own conclusions. It must make its own assessment of both benefit and detriment.
However, where the applicants and other parties participating in proceedings before the Tribunal agree with findings on factual matters set out in the Commission's published reasons for determination, the Tribunal would ordinarily be justified in treating those findings as common ground which significantly limits the areas of primary fact which the Tribunal is itself required to examine in detail; see Re Herald & Weekly Times Ltd (Media Council of Australia (No 1)) (1978) ATPR ¶40-058 at 17,601; (1978) 17 ALR 281 at 296 where the Tribunal (Deane J, President, Shipton and Walker, Members) observed that fairness and common sense combine to require that the Tribunal determine an application for review within the context of matters which can properly be seen to be in issue between the parties or which the Tribunal itself raises or indicates that it regards as being at large.
109 In the present proceeding, the Tribunal directed the parties to confer and file a joint statement identifying all findings on factual matters set out in the ACCC’s reasons for determination that are not contested by the parties on this review. The parties filed a joint statement on 11 March 2023. At the hearing of the review, the Tribunal informed the parties that, based on the SOFIC and submissions filed by the parties, there appeared to be many more factual matters set out in the ACCC’s reasons for determination that were not only not being contested by the parties but were supported by the parties. In response to the Tribunal’s request, on 11 May 2023 the parties filed an amended joint statement.
110 The Tribunal records its appreciation of the efforts made by the parties to agree a large body of background factual matters that are recorded in the ACCC’s reasons for determination. That enabled the Tribunal to focus its attention on the issues that are truly in dispute between the parties and the application of the statutory conditions for authorisation to the relevant facts. It has also greatly confined the material that the Tribunal has needed to consider in order to reach its decision. The cooperative approach of the parties is particularly important in the context of a review of a merger authorisation in which the Tribunal is required to make a decision within a confined time period.
Statutory preconditions for authorisation
Section 90(7)(a) – the competition test
111 The precondition for the grant of authorisation stated in s 90(7)(a) is that the ACCC is satisfied in all the circumstances that the conduct (the subject of the application for authorisation) would not have the effect, or would not be likely to have the effect, of substantially lessening competition. As noted in the 2017 Explanatory Memorandum, the competition test is a new basis for granting authorisation, following the recommendation of the Harper Review.68 The competition test for the grant of authorisation adopts the language of the competition based prohibitions in s 45, 46, 47 and 50 of the CCA, save that it is expressed in the negative and is focussed only on the effect of the conduct (and not its purpose). It follows that the established legal principles concerning the meaning of the words “likely effect of substantially lessening competition” are directly applicable to the competition test for authorisation. The summary of those principles that follows largely draws upon the discussion of the Full Federal Court in Australian Competition and Consumer Commission v Pacific National Pty Ltd.69
112 The meaning of the word “competition” in the CCA is well established, if somewhat difficult to state in a short form. The meaning given to the word has not altered in any material way since the first decision of the (then named) Trade Practices Tribunal in Re Queensland Co-operative Milling Association Ltd.70 As the Tribunal (Woodward J, Mr Shipton and Dr Brunt) there observed, competition (in a business or economic sense) is a rich concept containing a number of ideas.71 It remains helpful to repeat the oft-cited statements of the Tribunal in QCMA:72
Competition may be valued for many reasons as serving economic, social and political goals. But in identifying the existence of competition in particular industries or markets, we must focus upon its economic role as a device for controlling the disposition of society’s resources. Thus we think of competition as a mechanism for discovery of market information and for enforcement of business decisions in the light of this information. It is a mechanism, first, for firms discovering the kinds of goods and services the community wants and the manner in which these may be supplied in the cheapest possible way. Prices and profits are the signals which register the play of these forces of demand and supply. At the same time, competition is a mechanism of enforcement: firms disregard these signals at their peril, being fully aware that there are other firms, either currently in existence or as yet unborn, which would be only too willing to encroach upon their market share and ultimately supplant them.
This does not mean that we view competition as a series of passive, mechanical responses to “impersonal market forces”. There is, of course, a creative role for firms in devising the new product, the new technology, the more effective service or improved cost efficiency. And there are opportunities and rewards as well as punishments. Competition is a dynamic process; but that process is generated by market pressure from alternative sources of supply and the desire to keep ahead.
…
Competition expresses itself as rivalrous market behaviour. In the course of these proceedings, two rather different emphases were placed upon the most useful form such rivalry can take. On the one hand it was put to us that price competition is the most valuable and desirable form of competition. On the other hand it was said that if there is rivalry in other dimensions of business conduct — in service, in technology, in quality and consistency of product — an absence of price competition need not be of great concern.
In our view effective competition requires both that prices should be flexible, reflecting the forces of demand and supply, and that there should be independent rivalry in all dimensions of the price-product-service packages offered to consumers and customers.
Competition is a process rather than a situation. Nevertheless, whether firms compete is very much a matter of the structure of the markets in which they operate. The elements of market structure which we would stress as needing to be scanned in any case are these:—
(1) the number and size distribution of independent sellers, especially the degree of market concentration;
(2) the height of barriers to entry, that is the ease with which new firms may enter and secure a viable market;
(3) the extent to which the products of the industry are characterized by extreme product differentiation and sales promotion;
(4) the character of “vertical relationships” with customers and with suppliers and the extent of vertical integration; and
(5) the nature of any formal, stable and fundamental arrangements between firms which restrict their ability to function as independent entities.
Of all these elements of market structure, no doubt the most important is (2), the condition of entry. For it is the ease with which firms may enter which establishes the possibilities of market concentration over time; and it is the threat of the entry of a new firm or a new plant into a market which operates as the ultimate regulator of competitive conduct.
113 As also explained in QCMA, it follows that the identification of markets must be the essential first step in an assessment of present competition and likely competitive effects. The Tribunal explained:73
A market is the area of close competition between firms or, putting it a little differently, the field of rivalry between them (if there is no close competition there is of course a monopolistic market). Within the bounds of a market there is substitution — substitution between one product and another, and between one source of supply and another, in response to changing prices. So a market is the field of actual and potential transactions between buyers and sellers amongst whom there can be strong substitution, at least in the long run, if given a sufficient price incentive. Let us suppose that the price of one supplier goes up. Then on the demand side buyers may switch their patronage from this firm’s product to another, or from this geographic source of supply to another. As well, on the supply side, sellers can adjust their production plans, substituting one product for another in their output mix, or substituting one geographic source of supply for another. Whether such substitution is feasible or likely depends ultimately on customer attitudes, technology, distance, and cost and price incentives.
114 In Australian Competition and Consumer Commission v BlueScope Steel Limited (No 5)74, I observed that competition is best described by reference to its aim, mechanism and effect:
(a) The basic aim of business competition is to win sales – competitors strive to replace each other in the supply of products (whether goods or services) sought by customers.
(b) The key mechanism of competition is through substitution – to supply products to customers in place of another competitor’s supply. Substitution occurs on the demand side, whereby customers substitute one product or source of supply for another, and on the supply side, whereby suppliers adjust their production mix to substitute one product for another or one area of supply for another. Competitors strive to bring about substitution in a number of ways: through lowering their costs of production to enable them to profitably lower their prices; through improving the quality of their product and thereby increasing the value of the product to customers; and through inventing new products to meet the needs and wants of customers in new or better ways.
(c) As to effect, competition enhances the welfare of Australians by creating incentives and pressure for suppliers to reduce their costs of production and their prices (which, in the language of economics, is referred to as an improvement in productive efficiency), to commit resources to the production of goods and services most wanted by customers and to improve the quality of those products (which, in the language of economics, is referred to as an improvement in allocative efficiency) and to invest in innovation with the object of inventing new products to meet the needs and wants of customers (which, in the language of economics, is referred to as an improvement in dynamic efficiency).
115 The competition test for the grant of authorisation is concerned with the potential for the proposed conduct to substantially lessen competition. The concept of lessening competition is expanded by s 4G as follows:
For the purposes of this Act, references to the lessening of competition shall be read as including references to preventing or hindering competition.
116 As observed in Pacific National,75 the word “substantially” is imprecise; however, the courts have consistently said that, in each of sections 45, 47 and 50, the word does not connote a large or weighty lessening of competition, but one that is “real or of substance” and thereby meaningful and relevant to the competitive process.
117 The competition test for the grant of authorisation, like the competition based prohibitions in ss 45, 46, 47 and 50, uses the conditional (or hypothetical) future tense: the test requires an assessment of whether the proposed conduct (the subject of the application for authorisation) would have the effect or would be likely to have the effect of substantially lessening competition. The ACCC (and the Tribunal on review) must not grant authorisation unless it is satisfied that the proposed conduct would not have, or would not be likely to have, that effect. It is well settled that the test requires a comparison between the nature and extent of competition in any market potentially affected by the proposed conduct in the future with the proposed conduct being undertaken and in the future without the proposed conduct being undertaken.76 As also explained in Pacific National,77 the word “likely” does not mean more probable than not, but requires an assessment of what could reasonably be expected to be the consequences of the proposed conduct; it encompasses real commercial likelihoods, but not mere possibilities.
118 It follows that the competition test for the grant of authorisation requires the Tribunal to assess the likely effects of the proposed conduct on competition in all relevant markets. The conduct under consideration may have a range of potential effects on competition, both positive and negative, with such effects having different degrees of likelihood (from mere possibilities to near certainties). The Tribunal must not grant authorisation under the competition test in s 90(7)(a), however, unless it is satisfied that the likely effect of the proposed conduct, considered in totality, is not to substantially lessen competition in any market.
119 In Universal Music Australia Pty Ltd v Australian Competition and Consumer Commission78, the Full Federal Court reiterated four matters concerning the statutory test of substantially lessening competition which are relevant to the present application: first, competition is a process and the effect upon competition is not to be equated with the effect upon individual competitors, although the latter may be relevant to the former; second, competition is a means to the end of protecting the interests of consumers rather than competitors in the market; third, short term effects readily corrected by market processes are unlikely to be substantial; and fourth, the lessening of competition must be adjudged to be of such seriousness as to adversely affect competition in the market place, particularly with consumers in mind.79
Section 90(7)(b) – the net public benefit test
120 The precondition for the grant of authorisation stated in s 90(7)(b) is that the ACCC is satisfied in all the circumstances that the conduct (the subject of the application for authorisation) would result, or be likely to result, in a benefit to the public and the benefit would outweigh the detriment to the public that would result, or be likely to result, from the conduct. As observed by the Tribunal in Application by Port of Newcastle Operations Pty Limited (No 2),80 the terminology used and underlying concepts in that precondition are based on the law prior to its amendment by the 2017 Amendment Act. In particular, and as observed by the 2017 Explanatory Memorandum, the precondition for the grant of authorisation stated in s 90(7)(b) is consistent with the tests previously contained in s 90.81 That test required consideration of the benefits and detriments likely to result from the conduct, and involved a comparison of the future with, and without, the conduct for which authorisation is sought.82 As the statutory precondition in s 90(7)(b) requires the ACCC (or the Tribunal on review) to be satisfied that the public benefits likely to result from the conduct outweigh the public detriments likely to result from the conduct, it is convenient to refer to the test by the shorthand “net public benefit”.
121 A benefit to the public includes “anything of value to the community generally, any contribution to the aims pursued by society including as one of its principal elements (in the context of trade practices legislation) the achievement of the economic goals of efficiency and progress”.83 The relevant “public” is the Australian public.84 Similarly, a detriment to the public includes “any impairment to the community generally, any harm or damage to the aims pursued by the society including as one of its principal elements the achievement of the goal of economic efficiency”.85
122 In Qantas Airways,86 the Tribunal concluded that in assessing whether an increase in economic efficiency constitutes a public benefit for the purposes of the CCA, it was appropriate to apply a total welfare or surplus standard of economic efficiency. Accordingly, cost savings (productive efficiency gains) will constitute a public benefit even if the efficiency gain is captured in the first instance by the (private) parties to the proposed conduct. The Tribunal explained:87
187 We consider that the phrase “benefit to the public” is to be given a broad definition which, in addition to group interests, takes into account (with appropriate weighting) individual interests to the extent that such interests are considered by society to be worthy of inclusion and measurement. This broad approach to public benefit promotes the achievement of both static and dynamic efficiencies.
188 Given the above reasoning, we have formed the view that the “public versus private” dichotomy used by the parties in relation to cost savings is of fairly limited assistance when examining the benefits relied upon for the purposes of s 90. Rather, the enquiry should be directed towards the extent to which the benefit has an impact on members of the community, that is society. Does it fall into the category of “anything of value to the community generally”? If it does, what weight should be given to that benefit, having regard to its nature, characterisation and the identity of the beneficiaries of it?
189 It follows that cost savings achieved by a firm in the course of providing goods or services to members of the public are a public benefit which can and should be taken into account for the purposes of s 90 of the Act, where they result in pass through which reduces prices to final consumers, or in other benefits, for example, by way of dividends to a range of shareholders or being returned to the firm for future investment. However, the weight that should be accorded to such cost savings may vary depending upon who takes advantage of them and the time period over which the benefits are received.
123 In relation to public detriments, the Tribunal stated in Medicines Australia that:88
Although “detriment” covers a wider field than anti-competitive effects in many cases the important detriments will have that character. The relevant detriment will flow from the anti-competitive effect of the conduct to which authorisation is sought. This does not exclude consideration of other detriments which may be incidental to and therefore detract from a claimed public benefit. To that extent such detriment will be relevant in weighing the public benefit.
124 As the Tribunal explained in QCMA,89 in assessing public benefits (and detriments) it is necessary to go beyond labels or catchphrases. The Tribunal observed that its appraisal of claimed benefits must depend upon its appreciation of the competitive functioning of the industry and predictions about market behaviour and performance.90 The validity of claimed benefits will rarely be self-evident and “[a] claimed benefit may in fact be judged to be a detriment when viewed in terms of its contribution to a socially useful competitive process”.91
125 The necessity for authorisation applicants to quantify public benefits claimed to arise from proposed conduct was discussed by the Tribunal in Qantas Airways.92 Citing Howard Smith,93 the Tribunal said that the CCA does not require an applicant to quantify, in precise terms, the benefits claimed to arise if authorisation is granted but there must be a factual basis for concluding that the public benefits are likely to result.94 The Tribunal gave the following additional guidance with respect to the quantification of public benefits:
(a) an accurate, objective quantification of public benefits is difficult, in part because benefits have to be estimated for some period in the future and so their magnitude becomes a matter not only of empirical estimation based on assumptions but also one of statistical likelihood;
(b) the nature of public benefits should be defined with some precision, a degree of precision which lies somewhere between quantification in numerical terms at one end of the spectrum and general statements about possible or likely benefits at the other end of the spectrum;
(c) any estimates involved in benefit analysis should be robust and commercially realistic, in the sense of being both significant and tangible;
(d) appropriate weighting will be given to future benefits not achievable in any other less anti‑competitive way, and so the options for achieving the claimed benefits should be explored and presented;
(e) the Tribunal is not assisted by fanciful and speculative modelling of benefits where the underlying assumptions are not clearly spelled out, where the estimates have not been subject to rigorous sensitivity analysis, and where the estimating process is not wholly transparent;
(f) while detailed quantification of benefits is the best option, quantification is not required by the CCA and benefits should be quantified only to the extent that the exercise enlightens the Tribunal more than the alternative of qualitative explanation; and
(g) where benefits cannot be quantified in monetary terms, they can still be claimed in qualitative terms.95
126 A public benefit must, though, rise above the ephemeral and the trivial.96 As observed by the Full Federal Court in Australian Competition and Consumer Commission v Australian Competition Tribunal,97 the need for a benefit to be non-ephemeral can be deduced from the word “benefit” itself and Parliament is unlikely to have intended the Tribunal to concern itself with trifles.98
127 The satisfaction of the statutory precondition to the grant of authorisation does not oblige the ACCC, or the Tribunal on review, to grant authorisation.99 Nevertheless, as the Tribunal (O’Bryan J, Dr J Walker and Ms D Eilert) observed in Flexigroup (No 2), if the ACCC or the Tribunal on review were to be satisfied that the conduct is likely to result in a net public benefit, ordinarily authorisation would be granted.100
128 In Medicines Australia, the Tribunal considered whether authorisation should be granted in circumstances where the public benefits were assessed as insubstantial. The Tribunal said:101
Similarly, where the anti-competitive detriment is low to non-existent the ACCC may be entitled to say, as a matter of discretion, that it would only authorise the conduct if the public benefit to be derived from it, beyond that necessary to outweigh the anti-competitive detriment, or satisfy the per se conduct test is substantial. That is to say that the ACCC can require, in the proper exercise of its discretion, that the conduct yields some substantial measure of public benefit if it is to attract the ACCC’s official sanction. The Tribunal is in a similar position.
129 It is important to observe that the above statements in Medicines Australia concern the exercise of the discretion to grant authorisation once the statutory precondition is satisfied; the statements do not concern the content of the statutory precondition (which, as noted above, does not require the benefits to be substantial).
The conduct the subject of the authorisation
130 As noted above, there is one significant legal controversy that arises on this review. It concerns the proper application of the statutory preconditions for authorisation in s 90(7) of the CCA in respect of the conduct for which authorisation has been sought by the applicants. Section 90(7) requires the ACCC (and the Tribunal on review) to assess whether the conduct the subject of the application for authorisation would not be likely to have the effect of substantially lessening competition or would be likely to result in a net public benefit.
131 Each of the Proposed Transaction agreements is subject to a condition precedent that the agreements be authorised or otherwise approved by the ACCC. Despite that condition, it is common ground that the applicants did not seek authorisation to enter into and give effect to the MOCN Service Agreement or the Mobile Site Transition Agreement. The applicants only sought authorisation in respect of Telstra’s use of TPG’s spectrum licences pursuant to the Spectrum Authorisation Agreement (the Proposed Conduct).
132 In their application for authorisation, the applicants adopted a somewhat inconsistent approach to the application of s 90(7). As set out earlier in these reasons, the applicants propounded their application for authorisation on the following basis (emphasis added):
Telstra and TPG (together, the Applicants) seek merger authorisation for the authorisation of use of spectrum (under the Spectrum Agreement) which is deemed to be an acquisition within the meaning of s 50, CCA (Authorisation) on the basis that:
• the authorisation of spectrum would not have the effect, and would not be likely to have the effect, of substantial lessening of competition (SLC) in any market; and
• the public benefits associated with the authorisation of spectrum would result, or be likely to result, in a benefit to the public, and the benefit would outweigh any detriment to the public that would result, or be likely to result, from the Proposed Transaction.
133 It can be seen that the applicants sought authorisation for the Proposed Conduct (Telstra’s use of TPG’s spectrum licences), and stated the competition test by reference to the effect of the Proposed Conduct, but then inconsistently stated the net public benefit test by reference partly to the Proposed Conduct and partly to the effect of the Proposed Transaction. The submissions advanced in the application principally consider the issues by reference to the Proposed Transaction.
134 In its determination and its accompanying reasons, the ACCC identified the conduct the subject of the authorisation application as the contractual authorisation of Telstra to operate radiocommunications devices under TPG’s spectrum licences pursuant to the Spectrum Authorisation Agreement (the Proposed Conduct). However, in undertaking the statutory assessment required by s 90(7), the ACCC did not limit its analysis to the Proposed Conduct. Rather, it applied the preconditions in s 90(7) to the Proposed Transaction as a whole. The ACCC therefore examined the competitive effects, benefits and detriments that would likely arise in the future from all three agreements, both separately and together, which agreements the ACCC described as “interrelated” components that together implement a single commercial arrangement. The ACCC compared the likely effects, benefits and detriments of the Proposed Transaction as a whole with a future in which none of the three agreements comprising the Proposed Transaction were implemented. Although not stated expressly in the ACCC’s reasons for determination, it is implicit that the ACCC considered that the Spectrum Authorisation Agreement would, if authorised, only be implemented together with the MOCN Service Agreement and the Mobile Site Transition Agreement in the form in which they were put before the ACCC and, as a consequence, the ACCC was required to consider the effects of the Proposed Transaction as a whole notwithstanding that authorisation was sought for only one element of that Proposed Transaction. The legal basis of that approach was not explained in the ACCC’s reasons for determination.
Submissions of the applicants and the ACCC
135 Early in this proceeding (and prior to Optus’s intervention), the Tribunal raised with the applicants and the ACCC the question whether the determination of the ACCC, the subject of the application for review, was a merger authorisation within the meaning of s 4 of the CCA. The applicants and the ACCC submitted that the application was in respect of a merger authorisation. They noted that, while the applicants had entered into three interrelated agreement comprising the Proposed Transaction, authorisation had been sought for only one aspect of the Proposed Transaction being the authorisation given by TPG to Telstra under s 68(1) of the Radiocommunications Act pursuant to the Spectrum Authorisation Agreement. They confirmed that authorisation had not been sought more generally in respect of the Proposed Transaction. They further submitted that:
(a) pursuant to s 68A of the Radiocommunications Act, TPG’s grant of authorisation to Telstra to use TPG spectrum is deemed to be an acquisition within the meaning of s 50 of the CCA and therefore falls within the first limb of the definition of “merger authorisation” in s 4 of the CCA; and
(b) the second limb of the definition of “merger authorisation” is satisfied as the applicants did not seek authorisation for them to engage in conduct to which any provision of Pt IV of the CCA other than ss 50 or 50A would or might apply.
136 The applicants and the ACCC added that:
Under the Spectrum Authorisation Agreement, Telstra’s use of TPG spectrum in the 17% Regional Coverage Zone is limited to Telstra operating radiocommunications devices utilising that spectrum for use in the MOCN (in accordance with the terms of the MOCN Service Agreement). The ACCC therefore considered the Proposed Transaction as a whole in the context of the other Relevant Agreements, as the benefits flowing from the spectrum authorisation are transaction-specific, in that they relate to the use of the spectrum only as part of the implementation of the MOCN.
137 Following that response from the applicants and the ACCC, the Tribunal invited the parties (including Optus) to file submissions addressing the proper application of the statutory test for authorisation in s 90(7) on the basis that the application for authorisation is confined to conduct comprising the use by Telstra of TPG spectrum, and is thereby an application for a merger authorisation. In particular, the Tribunal invited submissions as to whether the statutory test is to be applied only to the conduct in respect of which authorisation is sought (the use by Telstra of TPG spectrum), or whether and on what basis the whole of the Proposed Transaction is relevant to the statutory test.
138 The applicants and the ACCC submitted that, in the circumstances of the present matter, s 90(7) requires the Tribunal to consider the effects arising from the MOCN Service Agreement and the Mobile Site Transition Agreement, as well as from the Spectrum Authorisation Agreement. That is because, having regard to the legally and commercially interdependent nature of the Proposed Transaction, if the Spectrum Authorisation Agreement is implemented, so too will the MOCN Service Agreement and the Mobile Site Transition Agreement. Conversely, if the Spectrum Authorisation Agreement does not proceed, nor will the MOCN Service Agreement or the Mobile Site Transition Agreement. In support of that submission, the applicants and the ACCC pointed to various features of the Proposed Transaction, including the following:
(a) The agreements were negotiated together by Telstra and TPG and entered into simultaneously. In Telstra’s submission, the evidence before the Tribunal reveals that the agreements were always regarded by Telstra and TPG as part of a single transaction. – that is, a “package of benefits”. The commercial rationale of the agreements was assessed, and they were entered into, on that express basis.
(b) The intended benefits to each of Telstra and TPG arising from the agreements operate only if the Proposed Transaction proceeds as a whole. TPG gains the benefit of access to Telstra’s network in the RCZ, and so is able to immediately deliver improved regional coverage to its customers, while Telstra acquires the use of TPG’s spectrum and certain TPG mobile sites, which supports in part the provision of services by Telstra to TPG. On a standalone basis, however, the agreements are of limited benefit to Telstra and TPG. For example, Telstra submits that without the authorised pooling of spectrum, the implementation of the MOCN (which results in TPG customers sharing use of the same RAN) would worsen Telstra’ existing congestion concerns in regional areas associated with its limited spectrum holdings.
(c) Each of the three agreements is subject to the same condition precedent, which requires one of several events to occur, including the grant of regulatory authorisation for the Proposed Transaction in one of the stated forms. Once the condition precedent is satisfied, the operative parts of each agreement will commence. This is said to reflect the interdependent, “all or nothing” nature of the Proposed Transaction.
(d) Once operative, the continued existence of one agreement depends on the existence of the others. In the event that the MOCN Service Agreement expires or is terminated, the Mobile Site Transition Agreement automatically expires and each of the applicants has the right to terminate the Spectrum Authorisation Agreement.
(e) Each of the agreements can have no sensible, practical operation without implementation of the others. For example, Telstra submits that the authority granted under the Spectrum Authorisation Agreement provides that, for the spectrum within the “Coverage Area”, which is defined as the geographic area to which the MOCN Service Agreement applies, Telstra must use that spectrum for the MOCN in accordance with the MOCN Service Agreement (unless it is agreed that it is not technically capable of such use). Accordingly, the Spectrum Authorisation Agreement cannot sensibly operate without the implementation of the MOCN Service Agreement.
139 The applicants and the ACCC submitted that, in the circumstances, the orthodox application of the “future with” and “future without” test pursuant to s 90(7) requires the Tribunal to assess the effects of the Proposed Transaction as a whole. They argued that this approach is supported by the statutory language of s 90(7), which calls for competition effects and the resulting benefits and detriments to be considered “in all the circumstances”.
140 The applicants further submitted that, to the extent the Tribunal had any concern that the applicants could in the future terminate the MOCN Service Agreement and/or Mobile Site Transition Agreement while maintaining the Spectrum Authorisation Agreement, it would be open to the Tribunal to condition authorisation on the applicants giving an undertaking pursuant to s 87B of the CCA requiring the parties to implement and to continue to give full force and effect to the MOCN Service Agreement and/or Mobile Site Transition Agreement, and to not materially amend those agreements except in accordance with the terms or with the prior consent of the ACCC. Such an undertaking could require the parties to undertake to terminate the Spectrum Agreement if the MOCN Service Agreement is terminated at any time. As noted above, the applicants subsequently proffered an undertaking to that effect (see cl 4 of the joint undertaking set out earlier in these reasons). The ACCC submitted that a condition of the kind embodied in cl 4 of the proposed joint undertaking would provide further assurance that the applicants would proceed with the whole of the Proposed Transaction. However, the ACCC contended that the condition proposed need not be secured by way of a s 87B undertaking. Rather, it could be imposed by the Tribunal as a condition if the Tribunal decided to set aside the ACCC’s determination and grant authorisation.
141 At the conclusion of the hearing, Telstra provided the Tribunal with a further written submission addressing the proper construction of s 90(7) and its application in the present case. In that submission, Telstra submitted that the analysis required by s 90(7) is not a strict “but for” test of causation. In that regard, Telstra placed reliance on the reasoning of Jagot J in Australian Competition and Consumer Commission v NSW Ports Operations Hold Co Pty Ltd102 in which her Honour observed, in the context of the competition-based prohibitions in Pt IV of the CCA, that:
(a) the effect of a contractual provision is one of objective fact;103 and
(b) the likely effect of conduct or a provision should not be equated with the “but for” test of causation – an event may not have occurred but for conduct or a provision but may not be the effect of the conduct or provision.104
142 Telstra’s submission continued:
… the counterfactual analysis in s 90(7) does involve a form of but for analysis, but one which asks whether, as a matter of objective fact, the specified conduct makes the relevant effects, benefits and detriments likely. That analysis must accord with the statutory purpose, in that a causal relationship which will not further that purpose ought to be excluded. However, where, as a matter of fact, effects and benefits are likely to occur in a future with the conduct but are unlikely to occur in a future without the conduct, and were those effects and benefits are pertinent to the state of competition in those alternate futures, those are effects and benefits which the Tribunal ought to take into account under s 90(7).
143 Optus’s submissions have altered during the course of the proceeding. Optus’s written submissions early in the proceeding, filed at the invitation of the Tribunal, were largely consistent with those of the applicants and the ACCC. However, Optus’s principal submissions filed in advance of the hearing introduced a caveat to the above analysis. It pointed to a distinction between assessing the effects of the Proposed Conduct in light of the Proposed Transaction (which comprise the relevant circumstances in which the Proposed Conduct will be undertaken if authorised), and assessing the effects of the Proposed Transaction. In oral submissions at the hearing, Optus advanced the submission that s 90(7) requires the Tribunal to assess the competitive effects, and the benefits and detriments resulting from, the Proposed Conduct, being the conduct that is the subject of the application for authorisation, and not the effects and results of the other Proposed Transaction agreements. Optus drew an analogy with the approach adopted by Beach J at first instance in respect of the application of s 45 of the CCA to the provisions of a commercial agreement (the Terminal Services Subcontract, or TSS) in Australian Competition and Consumer Commission v Pacific National Pty Limited (No 2).105 In that case, the ACCC argued that, in the future without the TSS, other commercial arrangements would not have occurred or would have otherwise changed with different competitive effects to those which in fact occurred. Justice Beach rejected that approach to the application of s 45, concluding that the section required consideration of the competitive effects of the provisions of the TSS, not the potential competitive effects of other commercial arrangements that the parties may have entered into if the TSS had not been entered into.106
144 The Tribunal largely accepts the applicants’ factual submissions concerning the interrelationship between the Proposed Transaction agreements. Specifically, the Tribunal accepts that the Proposed Transaction agreements were entered into at the same time as part of a single commercial transaction, and also accepts that each of the agreements is conditional on all of the agreements being authorised (or otherwise approved by the ACCC). The Tribunal also accepts that s 90(7) requires it to assess the likely competitive effects of, and the public benefits and detriments likely to result from, the Proposed Conduct in light of all relevant circumstances which includes the Proposed Transaction as a whole. However, the Tribunal does not agree that this assessment includes the likely competitive effects of, and the public benefits and detriments likely to result from, the MOCN Service Agreement or the Mobile Site Transition Agreement for which no authorisation has been sought. That conclusion is compelled by the plain language of the statutory preconditions for authorisation stated in s 90(7) when considered in its statutory context and having regard to the statutory purpose. The Tribunal rejects the submissions of the applicants and the ACCC to the contrary.
145 It is clear that the statutory preconditions for authorisation in s 90(7) are directed to the conduct that is the subject of the application for authorisation. The statutory preconditions require the ACCC, and the Tribunal on review, to assess the likely competitive effects of, and the public benefits and detriments likely to result from, that conduct. Both the competition test in s 90(7)(a) and the net public benefit test in s 90(7)(b) require a comparison of the future with, and without, the conduct for which authorisation is sought in order to assess the likely competitive effects of, and the public benefits and detriments likely to result from, that conduct. Nevertheless, the statutory test is directed to the effects of the conduct for which authorisation is sought, not the effects of other conduct that is coincident with, but not causally related to, the conduct for which authorisation is sought.
146 That conclusion is not only supported by the plain language of s 90(7), but also by the statutory context and purpose of the authorisation regime. The authorisation regime enables a person to obtain a statutory exemption from the prohibitions against anti-competitive conduct in Pt IV of the CCA. The applicant must specify the conduct for which authorisation is sought in the authorisation application and the ACCC (and the Tribunal on review) is empowered to grant authorisation in respect of that conduct: s 88(1). The authorisation, if granted, exempts that conduct, and not any other conduct, from the prohibitions in Pt IV of the CCA: s 88(2). Consistently with that focus on the conduct that is the subject of the application for authorisation, s 90(7) states the preconditions for authorisation by reference to the likely competitive effects of, and the public benefits and detriments likely to result from, that conduct.
147 It would be inconsistent with that statutory regime for the ACCC (and the Tribunal on review) to take into account, for the purposes of applying s 90(7), the competitive effects and public benefits and detriments resulting from other coincident conduct that is not the subject of the application. The problem with taking into account such other conduct under s 90(7) is readily demonstrated. The competitive effects and public benefits and detriments resulting from other coincident conduct might produce a conclusion that the authorisation preconditions are satisfied (for example, because the other coincident conduct is likely to result in significant public benefits). The result would be that the conduct the subject of the application would be authorised, but the authorisation would not extend to the other coincident conduct that had been taken into account in assessing the net public benefit. The effect of the authorisation would be to exempt the conduct, the subject of the application, from the prohibitions in Pt IV. The exemption would apply regardless of whether the applicants engaged in the other coincident conduct or at some future point in time ceased to engage in that other coincident conduct or varied that other coincident conduct. In other words, an inconsistent position could be reached that such other coincident conduct is assessed for the purposes of s 90(7) to provide a basis for the authorisation exemption, but the conduct would not be part of the authorised conduct. As a result, the applicant would be free to engage or not engage in that conduct without effecting the scope of the authorisation.
148 In support of their contrary contention, the applicants and the ACCC submitted that the orthodox application of the “future with” and “future without” test pursuant to s 90(7) requires the Tribunal to assess the effects of the Proposed Transaction as a whole. That submission involves a misunderstanding, or misapplication, of the “future with” and “future without” test and distorts the causal connection between the specified conduct and its effects which is required by s 90(7).
149 The reference to the “future with” and “future without” test is a reference to the well-established principle, applicable to the competition-based prohibitions in Pt IV of the CCA and the preconditions for authorisation in Pt VII of the CCA, that the relevant statutory provisions are to be applied on a forward-looking basis. In respect of those provisions of Pt IV which prohibit conduct that would have the effect, or would be likely to have the effect of substantially lessen competition, the necessary enquiry has been described as comparing the nature and extent of competition that would be likely to exist in the market in the future with the conduct occurring and without the conduct occurring. That description of the statutory test conveniently explains that the test is not a “before and after” analysis but a forward-looking exercise.107
150 But the “future with” and “future without” phraseology is not a substitute for the statutory language and cannot be applied in a manner that overlooks the need for a causal connection between the impugned conduct and its competitive effects. As is clear from the statutory text, and is apparent from the earliest cases, the enquiry must remain focused on the effects of the impugned conduct. In Dandy Power Equipment Pty Ltd v Mercury Marine Pty Ltd,108 Smithers J explained:109
To apply the concept of substantially lessening competition in a market, it is necessary to assess the nature and extent of the market, the probable nature and extent of competition which would exist therein but for the conduct in question, the way the market operates and the nature and extent of the contemplated lessening. To my mind one must look at the relevant significant portion of the market, ask oneself how and to what extent there would have been competition therein but for the conduct, assess what is left and determine whether what has been lost in relation to what would have been, is seen to be a substantial lessening of competition.
151 This was also the point being made by Beach J at first instance in Pacific National Pty Limited (No 2).110 His Honour observed:111
Section 45(2)(a) is not just about a “but for” factual inquiry. It also involves a normative competition causation inquiry focusing on the impugned provisions and their likely effect.
152 It was also the point being made by Jagot J in NSW Port Operations when her Honour observed that the likely effect of conduct or a provision should not be equated with the “but for” test of causation and that an event may not have occurred but for conduct or a provision but may not be the effect of the conduct or provision.112 In other words, the fact that certain conduct is likely to be coincident with the impugned conduct (in the sense that if the latter occurs the former is also likely to occur) is not a sufficient basis on which to conclude that the coincident conduct is an effect of the impugned conduct.
153 In respect of the test for authorisation in Pt VII of the CCA, the Tribunal has long adopted the same description of the forward-looking enquiry.113 The relevant enquiry is not a “before and after” test, but a “future with” and “future without” test.114 The Tribunal has always emphasised, though, that the statutory test requires a causal relationship between the conduct for which authorisation is sought and the resulting public benefits or detriments. In QCMA, the Tribunal emphasised that “that there must be established a causal relationship between the acquisition and the claimed benefit”.115 In Qantas Airways, the Tribunal explained that the statutory assessment requires that the benefit or detriment be “such that it will, in a tangible and commercially practical way, be a consequence of the relevant agreements if carried into effect”.116 In Re Medicines, the Tribunal observed: 117
The range of public benefits which may be considered is limited, in the context of authorisation, by the requirement that the benefit be the result or the likely result of the conduct which is the subject of authorisation: Re QCMA (1976) 8 ALR 481; 25 FLR 169. Thus the public benefit which may be considered under s 90 is confined to the extent that it must be related to classes of conduct amenable to authorisation and causally related to the conduct authorised.
154 In no sense of the statutory language in s 90(7) can entering into and implementing the MOCN Service Agreement and the Mobile Site Transition Agreement be characterised as an effect of, or a result of, Telstra’s use of TPG’s spectrum under the Spectrum Authorisation Agreement (or, stated more broadly, entering into and implementing the Spectrum Authorisation Agreement). The three agreements were entered into as part of the one commercial transaction. One agreement is not the effect or result of the other; rather, they are coincident agreements. Far less can the commercial and economic effects of the MOCN Service Agreement and the Mobile Site Transition Agreement be characterised as commercial and economic effects of Telstra’s use of TPG’s spectrum under the Spectrum Authorisation Agreement. The submission of the applicants and the ACCC to the contrary is founded solely on the submission that, in a future in which the Spectrum Authorisation Agreement is implemented, the applicants will also implement the MOCN Service Agreement and the Mobile Site Transition Agreement. While that may be accepted as a matter of likelihood, the submission ignores the causal nexus between the conduct for which authorisation is sought and relevant commercial and economic effects that is required by s 90(7).
155 No clear explanation was given by the applicants for their decision to limit the scope of the application for authorisation to Telstra’s use of spectrum under the Spectrum Authorisation Agreement, as opposed to the entirety of the Proposed Transaction. In the course of submissions, the applicants argued that if the Spectrum Authorisation Agreement was considered on its own, that would necessarily lead to a grant of authorisation because that agreement has no anti-competitive effects or other detriments. That submission begs the question why the application was then confined to that agreement. There is at least a hint in some of the material before the Tribunal that the applicants limited the scope of the application for authorisation to Telstra’s use of spectrum under the Spectrum Authorisation Agreement because they wished to benefit from the statutory time limits that are applicable to merger authorisation applications under Pts VII and IX of the CCA. As discussed earlier, by limiting the application for authorisation in that manner, the application satisfied the definition of merger authorisation in s 4 of the CCA. However, the Tribunal was also taken to some material that suggests that the applicants doubted whether a single application for authorisation could be made in respect of a business transaction in respect of which s 50 might apply to one or more elements and s 45 might apply to other elements. For the reasons expressed earlier, the Tribunal considers that s 88(1) permits a single application for authorisation in those circumstances.
156 The Tribunal considers that it was open to the applicants to apply for authorisation to enter into and give effect to the Proposed Transaction agreements. The application would not have satisfied the definition of a merger authorisation in s 4 of the CCA, but that is because the Proposed Transaction is not confined to a merger transaction within s 50 of the CCA. Other elements comprise a limited form of joint or collaborative venture between the parties which would ordinarily fall to be considered under s 45 of the CCA. If such an application had been made, the ACCC (and the Tribunal on review) would have been required to apply the statutory preconditions for authorisation in s 90(7) to the entirety of the Proposed Transaction. If authorisation were then granted, the applicants would have gained a statutory exemption to engage in that conduct. It would also follow that the applicants could not vary any part of the Proposed Transaction, or proceed with only part of the Proposed Transaction, without the risk of losing the statutory exemption.
157 Instead, the applicants confined the application to Telstra’s use of TPG’s spectrum under the Spectrum Authorisation Agreement, but then advanced submissions in favour of authorisation based on propounded effects of the Proposed Transaction as a whole, which included propounded effect of the MOCN Service Agreement and the Mobile Site Transition Agreement. The applicants ask the Tribunal to apply the authorisation preconditions to the effects of the Proposed Transaction as a whole in circumstances where that conduct is not the subject of the application for authorisation. If authorisation were to be granted on the basis sought by the applicants, exemption would be afforded to the Spectrum Authorisation Agreement but not the other two agreements. The applicants would be free to vary – and to terminate – those other agreements without affecting the authorisation granted, demonstrating the flaw in the applicants’ approach. While the applicants have, on this review, sought to address that problem by offering an undertaking under s 87B to maintain and not vary the other agreement while the Spectrum Authorisation Agreement is in force, the need for the undertaking only serves to confirm that the application for authorisation was made on a flawed basis and the applicants’ submissions concerning the construction of s 90(7) cannot be accepted.
158 There is a further incongruity between the manner in which the applicants have framed their application for authorisation and the authorisation regime. By limiting their application to Telstra’s use of TPG’s spectrum under the Spectrum Authorisation Agreement, and thereby bringing their application within the definition of “merger authorisation”, the application became subject to statutory time limits before the ACCC and now before the Tribunal on review. Further, the review before the Tribunal is not a re-hearing and restrictions are imposed on the information, documents and evidence to which the Tribunal may have regard. As noted earlier, these statutory time limits and procedural restrictions are imposed in respect of merger authorisations because merger transactions are time-sensitive. Determinations of merger authorisations are intended to be made expeditiously and this requires procedural restrictions on the conduct of the Tribunal’s review. Despite the confined manner in which the applicants framed their application for authorisation, the applicants have sought authorisation on the basis of the competitive effects of, and the public benefits and detriments resulting from, the Proposed Transaction as a whole. But the Proposed Transaction as a whole is not merely a merger (an acquisition of shares or assets for the purposes of s 50); it involves a long term joint or collaborative arrangement between Telstra and TPG pursuant to the MOCN Service Agreement and Mobile Sites Agreement. As such, the Proposed Transaction is not time-sensitive in the same manner as a merger transaction. If authorisation had been sought for the Proposed Transaction as a whole, the authorisation would not have been a merger authorisation under the CCA and the procedural limitations applicable to merger authorisations, including statutory time limits, would not have applied. On the review, the Tribunal would have been required to conduct a re-hearing which would have afforded a far greater opportunity to consider in detail the evidence adduced by the parties in respect of this long term transaction. Instead, by reason of the manner in which the application was framed by the applicants, the Tribunal’s review has been limited to an assessment of the material before the ACCC and the review has been conducted under considerable time constraints.
159 For those reasons, the Tribunal has applied the authorisation preconditions stated in s 90(7) to the Proposed Conduct and has assessed the likely competitive effects of and public benefits and detriments likely to result from, that conduct which is the subject of the application for authorisation. That assessment has been undertaken in light of all relevant circumstances, which includes the MOCN Service Agreement and the Mobile Site Transition Agreement. But the assessment does not involve weighing the likely competitive effects of, and public benefits and detriments likely to result, from those other agreements. Against the possibility that the Tribunal’s understanding of its statutory task is incorrect, the Tribunal has also applied the authorisation preconditions stated in s 90(7) to the Proposed Transaction as a whole. Ultimately, it has reached the same determination on both approaches.
160 The parties addressed submissions to the time horizon over which an assessment of the likely competitive effects of, and the benefits and detriments resulting from, the relevant conduct is to be assessed under s 90(7). The parties largely adopted a common position, submitting that the timeframe for analysis must relate to the effects being assessed. In the present case, the effects arise from, alternatively, the Spectrum Authorisation Agreement or the Proposed Transaction agreements collectively, which have a relatively long duration, being a minimum of 10 years and with options that may extend the duration to 20 years.
161 In Application by New South Wales Minerals Council (No 3),118 the Tribunal considered the appropriate time horizon in which to assess paragraph (a) of the declaration criteria in s 44CA of the CCA, namely whether access (or increased access) to the service, on reasonable terms and conditions, as a result of a declaration of the service would promote a material increase in competition in at least one market (whether or not in Australia), other than the market for the service. The Tribunal observed that any evaluation of markets, competition and the expected behaviour of economic actors depends upon the time horizon specified for the evaluation.119 In the context of an application for declaration of a service under Pt IIIA of the CCA, the Tribunal concluded that the relevant time horizon was across the medium term, meaning that that the effects of declaration should be assessed having regard to the present market conditions, opportunities and environment, and forecasting how those conditions, opportunities and environment may evolve and change into the medium term with and without declaration.120 The Tribunal further observed:121
151 … The object of Part IIIA is economic: to promote the economically efficient operation of, use of and investment in the infrastructure by which services are provided, thereby promoting effective competition in dependent markets. The infrastructure liable to declaration under Part IIIA is typically long-lived. Most significantly, changes in the terms of access to the infrastructure may not have immediate effects on competition in dependent markets and may only have effects in the medium term. This is because firms in dependent markets may have sunk costs. Provided the terms of access do not result in dependent market firms’ marginal costs exceeding their marginal revenues, the terms may have no immediate effect on existing firms’ consumption or production decisions. However, over the medium term, the terms of access may decrease the existing firms’ incentives to undertake further investment or otherwise increase consumption and production in the dependent market and may deter entry by new firms.
152 What constitutes the medium term in a given case may vary depending on the characteristics of the industries that are the subject of consideration. However, we consider that the assessment of the medium term should be guided by one practical consideration: over what time period is it feasible to make reasonable predictions about the conditions, opportunities or environment for competition in relevant dependent markets with and without declaration? …
162 The Tribunal considers that analogous considerations are relevant to the assessment of the authorisation preconditions in s 90(7). In assessing the likely competitive effects of, and the benefits and detriments resulting from, alternatively, the Spectrum Authorisation Agreement or the Proposed Transaction agreements collectively, the Tribunal is concerned with the medium term and not with the short term. In the context of commercial arrangements that have an expected duration of 10 to 20 years, the Tribunal is concerned with the impact of the arrangements on the supply of services in the relevant markets immediately and over the ensuing 5 to 10 years. While longer term considerations are also relevant, predictions about the development of markets and competition beyond 10 years, particularly markets which experience high levels of technological innovation and accompanying investment, become increasingly speculative. For that reason, the Tribunal places little weight on submissions and evidence that purport to predict the expected behaviour of market participants beyond a 10 year timeframe. Specifically, the Tribunal regards any predictions about TPG’s market strategy at the expiry of the Proposed Transaction agreements to be wholly speculative.
163 While each of the applicants filed separate SOFICs and submissions, in substance they advanced their case collectively. Each addressed different aspects of the application and adopted the submissions made by the other. For that reason, the Tribunal considers it appropriate to address the submissions made by each of Telstra and TPG collectively as the applicants’ submissions.
164 On the application for review, the applicants largely advanced their submissions on the basis that the relevant enquiry is whether the Proposed Transaction as a whole, rather than the Proposed Conduct (which comprises only part of the Proposed Transaction), satisfies the statutory preconditions for authorisation in s 90(7) of the CCA. In doing so, the applicants did not distinguish between the likely effects of, and the likely benefits and detriments resulting from, the Spectrum Authorisation Agreement and those resulting from the MOCN Service Agreement and/or the Mobile Site Transition Agreement. At the conclusion of the hearing, Telstra provided the Tribunal with a short written submission regarding the application of s 90(7) to the Proposed Conduct only.
165 The Tribunal summarises below the key contentions advanced by the applicants in respect of the effects of, and benefits and detriments resulting from firstly, the Proposed Transaction (being the principal basis on which the applicants advanced their case), and secondly, the Proposed Conduct.
The future without the Proposed Transaction – likely counterfactuals
166 In assessing the future without the conduct sought to be authorised (the Spectrum Authorisation Agreement), the applicants assumed that no part of the Proposed Transaction would be entered into.
167 The applicants contend that there is only one commercially realistic counterfactual if the Proposed Transaction were not entered into. Telstra would continue to proceed with its current strategy to achieve 95% 5G coverage across Australia by FY25 and TPG would undertake a targeted, incremental expansion of its mobile network in the 80%+ population coverage zone with the effect that TPG would hold a similar network and market position to that which it presently occupies.
168 It would remain commercially irrational for TPG to expand its own network across a material portion of the RCZ, without the market share or existing sites to justify that expansion, and it would not be able to offer 5G services in regional and rural Australia for the foreseeable future. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
169 Optus would continue to focus its network investment in metropolitan areas, maintaining a targeted investment strategy in regional Australia directed to disrupting consumer perceptions regarding Telstra’s coverage and supporting existing customers. If Optus rolls out its 5G network beyond metropolitan areas, this would remain subject to the risk of future capital restrictions and would not be completed until [REDACTED] [REDACTED] [REDACTED] [REDACTED] (being the time by which Optus originally planned to complete its 5G rollout, with [REDACTED] population coverage).
170 Over several years, Optus and TPG will continue to lose market share to Telstra. There would be no meaningful 5G alternative to Telstra for customers who seek coverage in the RCZ until [REDACTED] [REDACTED] [REDACTED] [REDACTED]. With diminishing market share and dwindling capital reserves available for future investment and innovation, Optus and TPG would be less able to compete with Telstra in wholesale and retail markets over time.
171 The applicants contend that there is no commercially realistic prospect of a comparable active network sharing arrangement between TPG and Optus if the Proposed Transaction does not proceed. This is for several reasons.
(a) [REDACTED].
(b) Active network sharing between TPG and Optus in regional areas is not feasible for at least 3 to 5 years (with a MOCN not feasible for at least 5 years), because of limited spectrum and equipment synergies between TPG and Optus. To the extent that these limitations could be overcome, they would necessitate prohibitively expensive equipment replacements and upgrades. There are, therefore, real impediments to reaching any active sharing arrangement.
(c) [REDACTED].
(d) The commercial relationship between TPG and Optus has deteriorated since the Proposed Transaction was announced and there is no real commercial chance that TPG and Optus will agree upon a roaming or network sharing agreement in parts of the 80%+ coverage area.
(e) If the Proposed Transaction did not proceed because authorisation of the Proposed Conduct was declined, TPG’s only alternative partner in an active sharing arrangement would be Optus. Neither Telstra nor TPG would be likely to enter into any alternative arrangement with each other: both would anticipate a similar, unfavourable regulatory response if they sought to pursue an arrangement on similar terms, and a substantially different arrangement would not be commercially viable. Optus’s bargaining position in subsequent negotiations for such an arrangement would be strengthened in the knowledge that it is the only potential partner to TPG. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED], [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED]).
172 Even if an agreement is reached between Optus and TPG, the applicants contend that the terms of any such agreement are uncertain. As a result, the competitive effects of any such agreement cannot be meaningfully assessed and cannot materially bear on the Tribunal’s assessment of the likely effect of the Proposed Transaction. On any view, however, such an agreement would be less competitive than the Proposed Transaction, as it would (among other things) have smaller coverage and would be limited to 4G for the foreseeable future.
Competitive effects of the Proposed Transaction
173 The applicants contend that the relevant markets within which to assess the likely competitive effects of the Proposed Transaction (and the Proposed Conduct) comprise:
(a) a national market for the supply of retail mobile services to consumers; and
(b) a national market for the supply of wholesale mobile services (being any form of service that an MNO supplies to another party, including other MNOs and MVNOs, to access its mobile network).
174 The applicants submit that it is unnecessary to consider effects in other secondary markets.
Increased competition in the national retail market
175 The applicants contend that TPG has significantly fewer subscribers and less regional coverage than Telstra and Optus and that it cannot overcome its poor infrastructure position in the RCZ unilaterally. TPG’s poor coverage has adversely affected retail mobile services competition. The applicants say that, under the Proposed Transaction, TPG would immediately and substantially increase the quality and extent of its 4G coverage by gaining access to Telstra’s RAN in the RCZ (pursuant to the MOCN Service Agreement). TPG would move from a current population coverage of 96% (which is limited to coverage provided by TPG’s 725 sites in the RCZ, and augmented to [REDACTED] by roaming on Optus’s 3G network) to future population coverage of 98.8% (provided by approximately 3,700 of Telstra’s sites in the RCZ). TPG would also receive immediate access to existing 5G at Telstra sites in the RCZ, and access to new 5G sites six months after the site is activated for Telstra customers. Moreover, relative to a roaming arrangement, the MOCN Service Agreement would provide a superior quality of service in the RCZ due to technical differences in the operation of a MOCN arrangement as distinct from a roaming arrangement.
176 The applicants contend that the immediate and substantial increase in TPG’s coverage and service quality in the RCZ would enable it to attract new customers, to better retain existing customers and therefore to grow its market share. The improvement in TPG’s competitive position will occur in circumstances where TPG’s consistently inferior coverage position in regional Australia (relative to Telstra and Optus) has acted as a barrier to acquiring and retaining retail and enterprise mobile customers, and has resulted in a low or declining market share. Further, if the Proposed Transaction were to proceed, TPG’s pricing would decrease on a quality-adjusted basis, where TPG would improve its service offering [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] (supported by increasing market share).
177 The applicants also emphasise that, under the MOCN Service Agreement, TPG would retain control of its mobile core network, which would allow it to continue to develop and improve its products and services, and to acquire other network or access services from other third parties. TPG would also be able to provide its customers with fixed wireless, NBIoT and Fixed NBN Fallback services in the RCZ for the first time.
178 The applicants contend that the improvements to TPG’s competitive position by reason of the Proposed Transaction will disrupt the competitive status quo and constitute a competitive threat to both Telstra and Optus.
179 The applicants also contend that another competitive benefit of the Proposed Transaction is improved service quality for Telstra, because the pooled spectrum will assist in addressing congestion in regional areas where around [REDACTED] of Telstra’s sites are congested.
Increased competition in the national wholesale market
180 In respect of the national wholesale market, the applicants contend that the Proposed Transaction is itself the outcome of a competitive process between Telstra and Optus. They argue that if the Proposed Transaction does not proceed, competition in the wholesale market would be lessened. The competitive process which resulted in the Proposed Transaction evidences a new competitive dynamic between the MNOs that had not previously existed, and which would not exist in a future without the Proposed Transaction. The applicants further submit that if the Proposed Transaction does not proceed, there is no real chance that Telstra would offer TPG any similar alternative wholesale deal. Telstra would have no ability or incentive to do so both because of the finely balanced economics of the current MOCN arrangement, and for fear of being again opposed by the ACCC. Moreover, the Proposed Transaction preserves future competition for the supply of wholesale infrastructure services to TPG either by Optus or new network suppliers, or through new technologies. TPG remains free to contract with Optus or others for roaming or access services in any part of Australia during the term of the relevant agreements, as well as to reopen competition for its entire network sharing arrangements at each of the 10, 15 and 20 year exit points under the agreements.
181 The applicants further submit that the Proposed Transaction would strengthen TPG’s ability to supply wholesale mobile services to MVNOs by reason of its improved national coverage and service quality. TPG has been unable to provide such services previously due to its inferior coverage position. This would provide current or future MVNOs with a further meaningful choice of MNO with the ability to offer regional 5G coverage, thus reducing barriers to entry and expansion. In turn, this would likely increase the number of MVNOs and, consequently, customer choice, putting downward pressure on pricing throughout the relevant markets.
182 The applicants contend that the benefits that Telstra obtains from the Proposed Transaction do not cause competitive detriment, including by cementing any position of market power in any market.
183 First, the applicants contend that Telstra does not have market power in either of the retail or the wholesale mobile markets. The applicants argue that the markets are competitive and fragmented and that Telstra’s competitive position in those markets is due to the significant investments it has made in its mobile network.
184 Second, the applicants contend that the wholesale revenue that Telstra will earn under the MOCN Service Agreement will not lessen competition. The applicants argue that the payments reflect fair compensation to Telstra for a service that Telstra will provide to TPG at considerable cost. Further, any scale benefits to Telstra from carrying TPG’s traffic on its network and earning associated additional wholesale revenue is not competitively significant. Telstra submits that the decision to proceed with the Proposed Transaction was finely balanced because the risks and rewards of the deal to Telstra were equally apparent – Telstra faces the potential risk of losing market share to TPG and a decline in ARPU. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED], [REDACTED]. [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
185 Third, the applicants contend that Telstra’s ability to access TPG spectrum does not lessen competition for the following reasons:
(a) The additional spectrum is for the applicants’ joint use as part of the MOCN. While access to additional spectrum means that an MNO needs to build fewer sites (and therefore incur less capex) to deliver a given level of network capacity, in the Proposed Transaction the pooled spectrum will mitigate the cost to Telstra of carrying TPG’s traffic. Telstra submitted that the capital savings from pooling spectrum are expected to be less than the additional capex Telstra will need to incur to accommodate the TPG traffic and the technology upgrades required to utilise the TPG spectrum.
(b) The spectrum provides only a modest service quality benefit to Telstra by potentially improving speed for the worst [REDACTED] of users in the RCZ by up to [REDACTED]. While Telstra expected TPG’s spectrum would assist to alleviate congestion, and indeed is necessary to facilitate the MOCN, the service quality improvement for a small subset of regional customers was not sufficient to form [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. The immediate benefit [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]; [REDACTED] [REDACTED] “[REDACTED] [REDACTED]” [REDACTED] [REDACTED].
(c) While it is common ground that the MOCN would have more total spectrum in the RCZ than Optus, Optus overstates the extent. Further, any analysis of relative network capacity must go beyond aggregate spectrum holdings and factor in physical infrastructure and demand on the network as the relevant measure. Aetha’s modelling – which takes into account spectrum holdings, physical infrastructure and customer demand – shows that access to TPG’s spectrum will not give the MOCN any practical capacity advantage over Optus.
(d) Optus’s internal papers confirm that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
186 Fourth, the applicants contend that Telstra’s access to up to 169 of TPG’s mobile sites in the RCZ would not lessen competition. The access would underpin the continuity of coverage under the MOCN Service Agreement in areas where TPG has sites but Telstra does not, and would decrease coverage gaps between Telstra and Optus.
Incentives to invest in infrastructure
187 The applicants further contend that the Proposed Transaction would not be likely to affect the incentives of any of the three MNOs to invest in infrastructure in regional Australia in a manner, or to an extent, that would substantially lessen competition.
188 In relation to TPG, if the Proposed Transaction does not occur, it is not likely that TPG would have sufficient incentive or ability to invest materially in its own mobile infrastructure in regional Australia, beyond a modest targeted build on a small number of sites. In contrast, the Proposed Transaction preserves TPG’s opportunity and incentive to invest in its network: the MOCN is non-exclusive; TPG retains significant independence and flexibility to invest in its network outside of the MOCN area or to build out its own network incrementally within the MOCN area; and TPG’s stronger market position will provide it with an improved capital base and incentives to invest. It follows that the Proposed Transaction will not meaningfully affect competition by reducing TPG’s incentives to invest in regional infrastructure relative to any realistic counterfactual.
189 In relation to Optus, the applicants contend that the Proposed Transaction would be unlikely to reduce its incentives to invest in regional Australia in the foreseeable future. In support, the applicants rely on the following matters:
(a) As a matter of orthodox economic principle, and consistently with its past behaviour, if Optus is threatened by increased competition in relation to the quality of services provided by TPG, it will redouble its efforts to close the quality gap.
(b) Whether Optus is willing to invest capital in 5G infrastructure in the RCZ depends on the risk it faces as a consequence of not investing, such as loss of customers, falling revenues, and the devaluation of its existing investment. It is inherently unlikely that Optus would take this risk. Optus would also risk losing its spectrum advantage in metropolitan areas if metropolitan customers who value regional coverage consider Optus’s regional offering to be inadequate. Moreover, any failure to upgrade to 5G will also jeopardise Optus’s ability to roll out later generations of technology, including 6G.
(c) When the Proposed Transaction was announced, Optus’s initial response was that the deal posed a greater risk to Telstra’s market share than to Optus’s market share, and Optus proposed prioritising its 5G rollout in order to increase its competitiveness. The applicants argue that subsequent statements by Optus that the Proposed Transaction would [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] are part of a legal and regulatory strategy to oppose the Proposed Transaction and should be treated with scepticism.
(d) Similarly, the applicants contend that no weight should be given to Optus’s financial modelling prepared in July 2022 which indicated that, if the Proposed Transaction went ahead, the most favourable option available to Optus would be to [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. They argue that no weight should be given to the model because it was prepared, at least in part, to persuade the ACCC to decline approval for the Proposed Transaction. In any event, the model itself shows that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. The modelling was founded on questionable assumptions to which it is highly sensitive and, if reasonable adjustments are made, the modelling indicates that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. Moreover, the modelling considers only [REDACTED], [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]). [REDACTED] “[REDACTED] [REDACTED] [REDACTED]” [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
(e) The Proposed Transaction would not disturb the factors that currently influence Optus’s investment program, which is focused on [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. The applicants argue that Optus’s commercial and competitive strategy has never been to match Telstra’s regional and rural coverage, but rather to differentiate on speed and customer experience in metropolitan areas, to pursue a targeted build where it makes commercial sense, and to maintain a coverage differential with TPG. In this regard, the Proposed Transaction, which concerns only 17% of the national markets in rural and regional areas, would not affect the core means by which Optus currently competes with Telstra and TPG and would likely continue to compete.
(f) To the extent that Optus could [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
(g) More generally, any decision by Optus and Singtel to reduce future investment in its network in the RCZ in a future with the Proposed Transaction in the face of enhanced competition from TPG, and the disruption of the enduring network relativities between the three MNOs, would not constitute a relevant lessening of competition within the meaning of the CCA. Although the Proposed Transaction would harm the interests of a competitor, it would not harm the competitive process. The applicants submit that, in effect, Optus’s opposition to the Proposed Transaction is a “plea for less competition” in order to earn a better financial return.
190 In relation to Telstra, the applicants contend that neither the Proposed Transaction, nor any reduced regional 5G investment by Optus, would reduce Telstra’s regional 5G investment. In support, the applicants rely on the following matters. First, Telstra has publicly committed to its 5G strategy, which is to achieve 95% coverage by FY25. [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. Fourth, expected increases in data consumption will require Telstra to continue investing in network capacity in regional areas, to avoid congestion. Fifth, Telstra must invest to upgrade to subsequent generations of mobile technology to match mobile handsets as older generations become obsolete.
191 The applicants contend that the risk of coordinated effects from the Proposed Transaction is low, having regard to the ability of Telstra and TPG to differentiate their product offerings under the MOCN and limitations on access by Telstra to TPG customer information. They contend that the risk of coordinated effects would be higher if Optus and TPG were to enter into a network sharing arrangement, especially if such an arrangement provides for roaming only.
Effects on secondary or sub-markets
192 The applicants contend that the only relevant markets in which the Tribunal needs consider the competitive effects of the Proposed Transaction are the national retail and wholesale mobile services markets. Even if other secondary or sub-markets are considered, the Proposed Transaction would not be likely to have the effect of substantially lessening competition in those markets.
(a) The Proposed Transaction would not have any material effect on demand in future spectrum auctions or the secondary market for the acquisition of spectrum.
(b) As to the supply of mobile services to “enterprise customers” (i.e. government and corporate or industrial customers) as a customer segment within the national retail mobile services market, the Proposed Transaction would enable TPG to compete more effectively with Optus and Telstra for enterprise customers who value coverage in the RCZ.
(c) As to the supply of NBIoT services, the improved scale and customer reach which the Proposed Transaction would provide to TPG, as well as its independent control of its core mobile network and its metropolitan infrastructure, would increase TPG’s ability and incentive to invest in new or nascent NBIoT applications or services.
(d) As to any national market for passive mobile network infrastructure services, without the Proposed Transaction, TPG would not be a material acquirer of tower access or any other passive infrastructure in the RCZ. With the Proposed Transaction, TPG would continue to acquire those services in the 81.4% population coverage areas, and it would have the increased incentive and ability to do so, having regard to its improved competitive position and associated revenue benefits and capital savings.
(e) As to the supply of retail fixed wireless services, the Proposed Transaction would enhance competition in the national market for retail fixed broadband services by enabling TPG to supply 5G fixed wireless services to customers in the RCZ and by providing Telstra with an improved capacity to supply such services, in both cases in competition with NBN Co Limited (NBN Co).
Competitive effects of the Proposed Conduct
193 The applicants contend that, if the s 90(7)(a) analysis is limited to the Proposed Conduct only, the Tribunal should be satisfied that the Proposed Conduct would not have the effect, or would not be likely to have the effect, of substantially lessening competition in the relevant markets.
194 In a future with the Proposed Conduct, the applicants contend that Telstra will obtain modest, temporary benefits in respect of its service quality in the RCZ. These benefits would not endure, nor would they have any real impact on Telstra’s competitive position in the national markets relative to Optus and TPG. The relevant benefits, which arise from access to additional spectrum, include: modest and temporary congestion relief for the worst-affected users in the RCZ; a modest increase in service quality (including higher speeds) for Telstra customers in the RCZ; reduced site densification costs and the ability to redeploy cost savings to improve 4G or 5G coverage; and the reduced costs of providing services to TPG under the MOCN Service Agreement.
195 The applicants contend that the above benefits for Telstra are modest, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED]) [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. To the extent that the spectrum provides speed, capacity or capex benefits, these apply only, or principally, within the RCZ and therefore have limited application to competition within the national market. The modest nature of the benefit for Telstra means that the Spectrum Authorisation Agreement (considered alone) would not substantially affect the competitive position of Telstra in the national retail or wholesale mobile services markets, relative to Optus or TPG.
196 By contrast, in a future without the Proposed Conduct, Telstra would not receive these benefits, and would likely not be able to provide the MOCN services to TPG at all.
197 TPG, in a future with the Proposed Conduct, would be able to monetise and use its spectrum in the RCZ (including to serve its own customers). Without the Proposed Conduct, TPG’s spectrum in the RCZ would be underutilised due to TPG’s limited infrastructure in the RCZ and its limited ability to unilaterally expand that infrastructure, and TPG would not receive spectrum payments from Telstra. Even if TPG entered into an alternative arrangement with Optus, Optus is unlikely to seek access to TPG’s spectrum.
198 Optus, in a future with the Proposed Conduct, would continue to have a spectrum advantage over Telstra in terms of spectrum capacity per customer, particularly in contiguous low-band spectrum, which is most valuable in regional areas. Optus’s network would also continue to have greater capacity than the combined MOCN to accommodate usage growth at the same rate of network densification. Optus’s spectrum holdings would otherwise be unaffected, including its mid-band holdings in metropolitan areas which provide Optus with a speed advantage in those areas. In a future without the Proposed Conduct, Optus is unlikely to acquire access to any TPG spectrum that would be meaningful to it – [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED]) [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. Optus sees little value in the TPG spectrum in the RCZ. Optus may also be able to acquire future spectrum across certain bands suited to 5G deployment that are scheduled to become available in the next decade. Optus therefore would not face any competitive disadvantage based on its spectrum holdings relative to Telstra in the future with the Spectrum Authorisation Agreement relative to the future without it.
199 The applicants submit that the Proposed Transaction, would be likely to result in benefits to the public that outweigh any potential detriment. Many of the benefits identified by the applicants are, in effect, alleged pro-competitive features of the Proposed Transaction that are relied on in support of the applicants’ case pursuant to s 90(7)(a) of the CCA. The benefits identified were said to comprise the following.
200 First, the applicants contend that there are a range of public benefits that will result from increased competition in wholesale and retail markets: a reduction in quality-adjusted pricing by increased pricing pressure from TPG; improvements in the service quality of Telstra and TPG in regional areas; an increase in the incentives for, and ability of, TPG to invest in its core network and its network in metropolitan areas, as well as to innovate in its service and product offerings.
201 Second, the applicants contend that there are economic and consumer benefits of reduced mobile congestion in regional areas. Spectrum sharing will help Telstra significantly reduce congestion in regional areas which will result in improved service quality for affected customers.
202 Third, the applicants contend that Telstra pooling spectrum with TPG and gaining access to 169 TPG sites in the RCZ will enable Telstra to avoid the costs of building new mobile sites in the RCZ, thereby generating productive efficiencies. Further, the more efficient use of mobile infrastructure in rural and regional Australia through the pooling of spectrum and the sharing the Telstra’s RAN will avoid duplication in investment between Telstra and TPG and reduce capital expenditure costs, which cost savings are likely to be substantially passed onto consumers and/or redeployed by Telstra to bring forward its 5G rollout or to extend coverage. TPG will avoid the costs of maintaining and upgrading at least 550 existing sites that it will decommission.
203 Fourth, the Proposed Transaction will bring forward Telstra’s 5G rollout in regional areas by freeing up capital that otherwise would be directed to providing infill coverage to address congestion issues.
204 Fifth, the applicants contend that there are environmental benefits associated with the reduced need for site duplication in the RCZ.
Joint undertaking and TPG undertaking
205 The applicants contend that, if the Tribunal is not able to be satisfied of one or other of the matters in s 90(7) of the CCA in the absence of conditions, imposing a condition that Telstra and TPG give the undertakings proposed would be sufficient to resolve any potential competition concerns and to prevent or reverse any harm that arises or may arise within the next eight years. The joint undertaking would require the Proposed Transaction to be authorised or approved within 8 years, which would remove any competition concerns with respect to any impact on Optus and its network investment in regional areas.
206 As noted earlier, Optus contends that s 90(7) requires the Tribunal to assess the competitive effects, and the benefits and detriments resulting from, the Proposed Conduct, being the conduct that is the subject of the application for authorisation, and not the effects and results of the other Proposed Transaction agreements. Optus therefore advanced its case before the Tribunal on alternative bases: as its primary contention, on the basis that the relevant inquiry is whether the Proposed Conduct satisfies the statutory preconditions for authorisation in s 90(7) of the CCA; and as its secondary or alternative contention, on the basis that the relevant inquiry encompasses the effects of the Proposed Transaction as a whole. Optus contends that the test for authorisation set out in s 90(7) of the Act is not satisfied, irrespective of whether it is assessed by reference to the Proposed Conduct or the Proposed Transaction as a whole.
The future without the Proposed Transaction – likely counterfactuals
207 In assessing the future without the conduct sought to be authorised (the Spectrum Authorisation Agreement), Optus assumed that no part of the Proposed Transaction would be entered into by Telstra and TPG (in the same manner as the applicants).
208 Optus contends that, in any counterfactual, Telstra would remain the dominant provider of mobile services, with all the advantages it presently enjoys including the highest number of subscribers, sites, and significant spectrum holdings. Unlike Optus and TPG, Telstra does not face additional capital expenses associated with replacing existing Huawei equipment. It would, however, face a competitive challenge to its network leadership from Optus and would defend its network leadership by investing in infrastructure including in regional areas. Telstra would not face congestion issues in the RCZ because it is able to address congestion issues through mobile site investment. Further, Telstra’s network capacity will experience a significant uplift as it closes down 3G services and migrates to 4G and 5G, which will result in additional low-band spectrum becoming available from July 2024.
209 Optus contends that there are two counterfactuals with a real commercial likelihood in a future without the Proposed Transaction.
210 In the first counterfactual, Optus and TPG would continue to invest in 5G technology, including in the RCZ. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. 5G technology would become progressively available to customers as the rollout progresses, with [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], the target date for completion of a 5G rollout across the [REDACTED] coverage as set out in Optus’s original business case underpinning Optus’s decision to commence investment in 5G technology in 2020 (5G Business Case) in July 2020. Contrary to the applicants’ contention, Optus submits that its investment strategy would [REDACTED] [REDACTED] [REDACTED] “[REDACTED]” [REDACTED] [REDACTED] [REDACTED] “[REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]”. [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. In this counterfactual, TPG would also invest, including in 5G technology and in the RCZ, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. Telstra would remain the dominant provider of mobile services. However, with continued investment by Optus and TPG, Telstra would continue to respond competitively.
211 In the second counterfactual, Optus and TPG would enter into a network sharing agreement in respect of 4G and 5G services, including in the RCZ. Optus contends that there is a real commercial likelihood of such an agreement being made for several reasons, which include the following.
(a) [REDACTED].
(b) The commercial benefits arising from a network sharing deal for both parties would be of a kind and scale such that it would be economically rational for Optus and TPG to enter into an arrangement. Those benefits, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], would include: improving scale and unit costs; generating a higher return on investment (for Optus in particular) and enabling cost savings to be deployed to build more sites and close the coverage gap with Telstra; attaining higher network speeds and performance through spectrum pooling; allowing Optus to provide wider 5G coverage through access to TPG’s low band spectrum; allowing Optus to deploy and sell fixed wireless services through access to TPG’s mid-band spectrum; and, for TPG, extending its coverage footprint from 96% up to 98.5% and providing access to a 4G and, in time, a 5G network.
(c) There are no insurmountable technical barriers to the implementation of active sharing between Optus and TPG.
(d) Any potential loss of market share by Optus to TPG caused by a network sharing deal would be outweighed by the cost reductions gained, such that Optus would have a real commercial incentive to enter into an arrangement with TPG.
(e) Any distrust or deterioration in the commercial relationship between Optus and TPG could be overcome given the economic benefits of a network sharing arrangement.
(f) [REDACTED].
212 Optus further contends that, contrary to the applicants’ submission, the Tribunal need not be satisfied of a likelihood of a network sharing agreement on particular identified terms. Any such agreement with TPG would likely improve Optus’s competitive position vis-à-vis Telstra and would bring greater balance to the benefits of scale experienced by market participants. The fact that such an agreement would constitute a competitive threat to Telstra is borne out by Telstra’s own efforts to forestall it by entering into the Proposed Transaction with TPG.
Competitive effects of the Proposed Conduct/Proposed Transaction
Relevant markets
213 Optus contends that the relevant markets in which to assess competitive effects include the national markets for the wholesale and retail supply of mobile services. Optus submits, however, that the extent of services in the RCZ is of particular importance to people who live or work in that area, and so particular focus on the extent of competition in the provision of services to that area is required.
214 Optus also contends that there are other relevant markets. First, demand from enterprise and government customers is distinct because they have larger data requirements, require services in multiple locations, and require priority traffic and higher quality of service. Second there are regional markets for the acquisition of spectrum. Third, there are upstream markets (such as the market for the supply of services to install mobile infrastructure) and downstream markets (such as markets for the supply of a variety of services using fixed wireless and mobile technology).
Effects in national markets for wholesale and retail supply of mobile services
215 Optus contends that, in a future with the Proposed Conduct and/or the Proposed Transaction, Telstra’s dominant competitive position in the national wholesale and retail mobile markets would be strengthened and entrenched. Telstra is dominant because of its spectrum holdings, superior network coverage, scale and non-reliance on Huawei equipment. The Proposed Conduct and/or the Proposed Transaction strengthens Telstra’s position in each of these dimensions.
216 First, Telstra would obtain additional spectrum under the Spectrum Authorisation Agreement for use in the RCZ and beyond it. Although the spectrum will be pooled for use with TPG, Telstra will benefit to a greater extent where it services a much larger number of the SIOs in the RCZ compared to TPG. The additional spectrum will increase Telstra’s access to total spectrum holdings in both low-band and mid-band spectrum, as well as to contiguous spectrum blocks in key bands, delivering greater improvements in capacity and speed. This will widen Telstra’s existing advantage over Optus in terms of total spectrum. Additional spectrum will also enable Telstra to increase service quality in the RCZ and in more remote areas without incurring the substantially greater costs of site densification, which is the only way to improve capacity without spectrum. This will further strengthen Telstra’s scale and cost advantages relative to Optus, which already exist. Telstra’s acquisition of TPG’s spectrum will also remove that spectrum from the market, where TPG would otherwise likely seek to monetise it. Thus, if the Proposed Conduct is considered alone for the purpose of s 90(7), authorisation ought to be declined.
217 Second, Telstra would obtain additional revenue from the use of its network by TPG customers where, under the MOCN Service Agreement, TPG will make wholesale payments to Telstra for that use. This revenue will defray the costs of Telstra’s 5G rollout, and will strengthen its scale advantage by allowing it to amortise to a greater extent its network improvement costs and to reduce its average costs of providing services. The receipt of wholesale payments will also lessen the incentive for Telstra to compete for customers on price and quality, in circumstances where Telstra will receive revenue in respect of those customers it loses to TPG, and which TPG wins from Optus.
218 Third, Telstra would be able to access and deploy infrastructure on up to 169 existing TPG mobile sites, primarily within the RCZ, enabling Telstra to increase the quality and coverage of its services in areas that would be within range of those sites, eliminate gaps compared to TPG, and decrease gaps compared to Optus.
219 The Proposed Transaction also removes two competitive threats to Telstra: the first is a potential network sharing arrangement between Optus and TPG in the RCZ, and the second is mobile roaming or neutral host network schemes mandated by the relevant regulators.
220 By contrast, Optus contends that its ability to compete against Telstra in a future with the Proposed Conduct and/or the Proposed Transaction would be reduced. The effect of the spectrum advantage that Telstra will acquire by reason of the Proposed Conduct would confer a scale and cost advantage with which Optus would be unable to compete. That advantage is only further compounded by the Proposed Transaction as a whole. Optus’s competitive position would also be weakened where both TPG and Telstra would obtain customers at the expense of Optus as a result of the Proposed Transaction. TPG would “leapfrog” Optus in terms of both coverage and quality of services in the RCZ, and Optus will have no competitive advantage or selling point over either Telstra or TPG.
221 TPG would not, however, be a strong source of infrastructure competition in regional and rural areas if the Proposed Transaction proceeds. Optus contends that TPG would be unlikely to make infrastructure investments in the RCZ and would decommission sites that are not transferred to Telstra. TPG would also lose autonomy over aspects of its network and would have limited to no ability to differentiate itself based on geographic coverage, quality of network, the timing of technology upgrades, and its RAN investment strategy. While TPG would retain control of its core network, this would allow it to differentiate only through pricing or the packaging of data and inclusions. TPG would be dependent on Telstra’s infrastructure investments. That dependency is likely to persist at any of the exit points contemplated under the relevant agreements, which make such an exit commercially unrealistic.
222 Optus also challenges the applicants’ contention that TPG could, as a matter of commercial reality, re-establish its own network in the RCZ prior to, and on, exit from the Proposed Transaction. If it does not do so, without any existing 5G regional network, TPG would face significant difficulty in rolling out 6G. Accordingly, any short term benefits to TPG, such as certainty of network access in regional areas and increased utilisation of spectrum, will not ameliorate the long term structural deficit that entry into the Proposed Transaction would create.
223 Optus further contends that, in a future with the Proposed Transaction, it would be forced to reconsider its own 5G investment strategy, and would [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]-[REDACTED]-[REDACTED] [REDACTED]). [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]; [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. This was recognised in Optus’s 5G Business Case in 2020. Those disincentives will be compounded by the Proposed Transaction, which will strengthen Telstra’s competitive position and its ability to defeat Optus’s investments. Optus will also [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. That the Proposed Transaction affects only 17% of the national market is irrelevant where infrastructure competition in the region that services this part of the market is a nationally significant dimension of competition.
224 Optus contends that, contrary to the applicants’ submission, the anticipated harm to Optus from the Proposed Transaction is not merely harm to it as a competitor. Rather, it constitutes harm to competition. Optus has historically been a crucial source of competition to Telstra. However, if the Proposed Transaction were to proceed, Optus would be unable to compete with Telstra in the same manner as it has to date, and TPG would be unable to step into its shoes. Consequently, the competitive pressure that Optus’s investment generates and which motivates Telstra to invest in regional network infrastructure would be removed. In practical terms, this would likely lead to a reduction in the scale of Telstra’s 5G rollout, as well as to deferral or reduction of any subsequent investments in technology in those areas (including the construction of a 6G network). Telstra would be even further ahead than Optus in its 5G rollout and, necessarily, any future rollout of a new technology generation (which would leverage the rollout of the earlier generation). That relative disadvantage, and Telstra’s dominance, in regional Australia will be “baked in” for a long period of time. Even if it is accepted that the Proposed Transaction creates short term efficiencies for both Telstra and TPG, these cannot negate the long term impacts on competition in the relevant markets.
225 Overall, Telstra’s strengthened competitive position will raise strategic barriers to entry or expansion in the supply of wholesale or retail mobile services, and will lead to reduced competition on price and quality.
226 Optus also contends the Proposed Transaction heightens the risk of coordinated effects between Telstra and TPG, insofar as it will reduce product and service differentiation between them, allow for easier monitoring of and retaliation against deviations from mutually beneficial behaviour, and facilitate a degree of information sharing between them.
Effects in other relevant markets
227 Optus contends that the Proposed Conduct and/or the Proposed Transaction would have the result, or would likely have the result, of substantially lessening competition in other relevant markets.
(a) In the market for the acquisition of spectrum, the Proposed Conduct and/or the Proposed Transaction would likely have the effect of reducing demand, and therefore competition for, spectrum. First, the Proposed Transaction would result in a concentration of control by Telstra of TPG’s substantial spectrum holdings that would otherwise be monetised by TPG. Second, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. Third, potential entrants to the market may be deterred by the extent of Telstra’s spectrum holdings.
(b) In the market for enterprise and government customers, in a future with the Proposed Transaction, Telstra’s dominance would continue, particularly due to its relative scale advantages. TPG’s ability to compete for these customers would be diminished, where it would have less control over the development of its network in the RCZ and delayed access to 5G services. Similarly, Optus’s weakened position will limit its competitiveness in these markets. By contrast, in the counterfactual with an Optus/TPG network sharing agreement, Optus and TPG would be able to offer the higher quality service and prioritised network traffic typical for services in these markets. Optus and TPG would therefore be able to compete more strongly with Telstra for enterprise and government customers.
(c) In the market for fixed wireless access and NBIoT, Telstra’s strengthened market position as a result of the Proposed Transaction would enhance its ability to offer and develop these services. Optus, however, would [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
(d) In the upstream and downstream markets for the supply of services to install or maintain mobile infrastructure, and services for fixed wireless and mobile technology respectively, the suppression of dynamic competition as a result of the Proposed Transaction will negatively affect competition in these markets.
Net public benefit
228 Optus contends that none of the public benefits identified by the applicants are attributable to the Proposed Conduct alone. Therefore, the net public benefit test in s 90(7)(b) of the CCA cannot be satisfied if the analysis is properly confined to the Proposed Conduct and not the wider Proposed Transaction.
229 As to the public benefits which are said to arise from the Proposed Transaction, Optus contends that they either would not arise, would arise in an Optus/TPG deal counterfactual, would not be meaningful or would not result in cost savings that would be passed through to consumers. For example, the anticipated reduction in congestion across Telstra’s network would arise in any event, in circumstances where Telstra would take steps to address congestion even if the Proposed Transaction does not take place. Optus contends that it is also unlikely that the reduced network costs expected by the applicants under the Proposed Transaction would be passed through to consumers, in circumstances where the evidence suggests that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], and competition on pricing and quality will diminish by reason of [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
230 Conversely, Optus contends that the Proposed Transaction will occasion substantial public detriments that will outweigh any public benefits, including competitive harms, reduced network diversity and resilience in regional areas, and higher spectrum concentration impacting long term industry structure. Any immediate improvements in service quality arising from the Proposed Transaction will be short-lived because of reduced competitive pressure on Telstra to invest in coverage, speed, service innovation and technology upgrades, leading to poorer consumer outcomes in the future. As a result of the de-duplication of Telstra and TPG sites in the RCZ and [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], more regional areas would only be serviced by Telstra’s network. Any outages on that network will leave relevant areas with no or limited communications, and a loss of network resilience.
Joint undertaking and TPG undertaking
231 With respect to the joint undertaking, Optus contends that it does not address the relevant competitive harms arising from the Proposed Transaction, which would occur within the period contemplated by the undertaking (being 8 years, or by 2031) and would persist long afterwards even if the Proposed Transaction were then terminated. That is because Optus and TPG need to make investment decisions about 5G technology in the near term. If the Proposed Transaction occurs, Optus’s ability and incentive to invest will rapidly diminish, with the result that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. By the time the Proposed Transaction is terminated in 2031, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED].
232 With respect to the TPG undertaking, Optus contends that it would not prevent the weakening of TPG’s competitive position. On exit from the Proposed Transaction, TPG would have access to 60% of its current sites, comprising the 300 sites the subject of the undertaking and the 169 sites provided to Telstra under the Mobile Site Transition Agreement. The 169 sites would be subject to [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], and the 300 sites would not be functional. TPG’s network position on exit from the Proposed Transaction would, therefore, be different – and inferior – to its position on entry.
233 The ACCC acknowledged in its SOFIC that the ACCC’s function in this proceeding is to assist the Tribunal. As Optus has assumed the role of contradictor in the present review, the ACCC confined its contentions to the following topics:
(a) the statutory test and relevant matters of principle;
(b) the issues that the Tribunal will need to consider in undertaking its review;
(c) what is to be authorised by the Tribunal;
(d) the ACCC’s position in respect of the joint undertaking provided by the applicants, dated 20 April 2023; and
(e) clarification of one matter concerning the modelling referred to by the ACCC in its reasons for determination.
234 The Tribunal expresses its appreciation for the assistance provided by the ACCC in this review, and has taken into account the submission advanced by the ACCC. The ACCC’s submissions with respect to the statutory framework governing this review, and particularly the proper application of s 90(7) to the conduct for which authorisation has been sought, have been discussed earlier in these reasons. It is not otherwise necessary to reproduce the ACCC’s submissions, which have been taken into account by the Tribunal.
235 In future applications for review of this kind, the ACCC should not consider itself inhibited from providing assistance to the Tribunal merely because a competitor has intervened as a contradictor. The Tribunal has long recognised the important role played by the ACCC in these reviews, assisting the Tribunal to reach the correct decision in the public interest.122 In a similar review context, the Tribunal observed:
Where there is no natural contradictor on an issue in the application, a decision maker may need to participate in the proceedings in order to assist the Tribunal. In addition, if the statutory regime is consistent with the decision maker taking an active role in the proceedings, it is then appropriate for the decision maker to respond substantively to the review application. The role that the decision maker takes in merits review proceedings is also different to the role taken in actions for judicial review. In the case of merits review, there is a more limited application of the Hardiman principles because the decision maker, as “administrator” of the particular statutory regime, is uniquely placed to assist in that it has an in-depth knowledge of the scheme: Macedon Ranges Shire Council v Romsey Hotel Pty Ltd and Anor (2008) 19 VR 422 at [30]; Bankstown City Radio Co-operative Ltd v Australian Communications and Media Authority [2007] FCA 2053 at [6]; Geographical Indications Committee v O’Connor (2000) 64 ALD 325 at [35].123
236 Given the substantive role assumed by Optus in this proceeding as contradictor, the ACCC was correct to avoid any duplication of the submissions sought to be advanced by Optus. However, it is apparent from the ACCC’s reasons for determination, and the materials before the Tribunal, that a large number of interested persons advanced submissions to the ACCC both for and against the grant of authorisation. In future applications for review of this kind, the Tribunal would be assisted if the ACCC were to present the different views of interested persons in a convenient summary form, to ensure that the Tribunal is fully informed of all perspectives with respect to the relevant conduct and not merely the perspectives of the applicants and a primary contradictor.
E. FACTUAL BACKGROUND TO THE ASSESSMENT
237 At the direction of the Tribunal, the parties filed a document identifying all factual findings made by the ACCC in its determination that are not contested (referred to earlier as the Joint Document of Factual Findings). In the updated version of that document provided to the Tribunal on 11 May 2023, Telstra and TPG included notes or comments with respect to certain of the factual findings. In respect of some findings, Telstra and/or TPG noted that they had not been able to verify the data presented by the ACCC. During the hearing, Telstra and TPG confirmed that notes of that kind were not intended to convey that Telstra and TPG were contesting the data and they also confirmed that it was open to the Tribunal to place reliance on the data notwithstanding the note. In respect of other findings, the notes or comments were in the nature of submissions with respect to the content or relevance of the factual finding. The Tribunal has taken those matters into account.
238 For the purpose of its review, and in accordance with its role in the present application, the Tribunal has examined the findings of fact made by the ACCC, having regard to the evidence on which they were based, the submissions made by the parties about them, and the Joint Document of Factual Findings, applying particular scrutiny in respect of those findings that are contested. Based on that material, the Tribunal makes the following findings with respect to the factual background. Where a finding is not contested by the parties, this has been indicated below.
239 This section sets out relevant background information on the mobile telecommunications industry in Australia, the mobile networks, and their inputs.
240 A mobile network uses spectrum to deliver mobile services such as voice, SMS and mobile data to end-user devices.124 MNOs compete at both the retail and wholesale level, with their retail brands competing for consumer and enterprise customers, and their wholesale arms competing in the provision of wholesale services to MVNOs and other telecommunications providers.125 A mobile network typically has a number of primary components used to deliver these services: the radio access network, transmission networks, the core network, and spectrum.126
Figure 2: Major components of a mobile network127
241 The radio access network consists of base stations (mounted on mobile towers or located at cell sites) that use spectrum to connect end-user devices to the network via radio link. A base station provides mobile coverage to an immediate geographic area called a cell. Importantly, mobile devices will maintain connectivity with the network as the end-user's device moves between cells. This capability is known as inter-cell handover.128
242 In turn, transmission networks connect these base stations to the rest of the network, and the core network. These transmission links can be wireless (microwave, satellite), but are more commonly connected by fibre link. Transmission between base stations and the rest of the network is also known as backhaul. Backhaul is an important component of network quality. A mobile network needs sufficient backhaul capacity to carry traffic across its network.129
243 The core network manages voice, SMS, and data traffic, connects and manages different parts of the network, and handles functions like billing and user management.130 The core network is also where operators’ networks connect to other networks, including the internet. MNOs connect to other operators’ networks at points of interconnection between their respective core networks.131
244 Spectrum is the medium by which signals are carried between the end-user device and the radio access network, and ultimately the MNO’s wider network and beyond. Spectrum is discussed in further detail below.132
245 Telstra is listed on the Australian Securities Exchange. It is Australia’s largest telecommunications company and Australia’s largest MNO by number of subscribers and by the size of its network.133
246 In FY21, Telstra earned total income (excluding financial income) of $23.1 billion, EBITDA of $7.6 billion and NPAT of $1.9 billion.134 Telstra’s 2021 Annual Report stated, in respect of its mobile network:
We continued to offer Australia’s best, largest and most reliable mobile network. We expanded our 5G footprint to cover 75 per cent of Australians where they live and our 4G footprint to 99.5 per cent of all Australians taking it to over 2.5 million square kilometres.
247 In FY22, Telstra earned total income (excluding financial income) of $22.0 billion, EBITDA of $7.3 billion and NPAT of $1.8 billion.135 Telstra’s 2022 Annual Report stated, in respect of its mobile network and the implementation of its T25 strategy:
We will take advantage of the many great strides made in our 5G rollout and boost capacity, speed and population coverage of our mobile network. We will also expand our regional network by an extra 100,000 square kilometres so we can continue to deliver leading mobile coverage and build on our network leadership.136
…
This year Telstra continued to operate Australia’s best, largest and most reliable mobile network. We expanded our 5G footprint so it now covers 80 per cent of Australians where they live and our total network grew to cover 99.5 per cent of all Australians, taking it to over 2.6 million square kilometres. We also are ranked number one for Ookla on Overall Mobile Speeds.137
248 Telstra’s 2022 Annual Report also recorded that, as at 30 June 2022, Telstra’s “communication assets” had a cost of around $62.5 billion, with a written down book value (after accumulated depreciation and impairment) of around $19.7 billion.138 In FY22, Telstra’s total capital expenditure before investments was around $3.1 billion (and excluding intangible assets was around $2.176 billion).139
249 Telstra’s mobile network covers 99.5% of the population140 and 2.6 million km2 of land141 (which is around 33% of Australia’s land mass)142, with Telstra’s 5G network covering 80% of the population.143 Its network includes more than 11,000 mobile base stations nationally144 and approximately 3,700 mobile base stations within the RCZ.145 Telstra’s T25 strategy for its mobile network, as summarised in its 2022 Annual Report, included the following goals:146
Network leadership by FY25:
• ~95% pop. coverage for 5G
• >80% of traffic on 5G
• 3G closed in FY24
Win majority of key surveys for best fixed/mobile network including
• Coverage, and
• Overall customers speeds for mobile FY23-FY25
Double metro cell sites by FY25 to densify the network
Expand regional coverage
• 100,000 km2 new coverage by FY25
250 TPG (which was formerly named Vodafone Hutchison Australia Limited) was listed on the Australian Securities Exchange on 30 June 2020. On 13 July 2020, the newly listed company merged with TPG Corporation Limited (formerly named TPG Telecom) to bring together the two telecommunications companies, TPG and Vodafone. TPG owns and operates nationwide mobile and fixed networks under brands including Vodafone, TPG, iiNet, AAPT, Internode, Lebara and felix. In the year ending 31 December 2021, TPG earned service revenue (which excludes device and installation revenue) of $4.39 billion, EBITDA of $1.73 billion and NPAT of $110 million.147 Its total capital expenditure in that financial year was around $900 million (around $700 million excluding intangible assets).148 As at 31 December 2021, TPG’s “network, equipment and infrastructure” assets had a cost of around $6.4 billion, with a written down book value (after accumulated depreciation) of around $2.7 billion.149
251 TPG is Australia's third-largest MNO by number of subscribers.150 TPG operates more than 5,600 mobile base stations nationally, and currently operates 749 mobile base stations in the RCZ. TPG’s current network covers 96% of the population and is extended under a roaming agreement with Optus.151
252 In TPG’s 2021 Annual Report, the CEO and Managing Director, Mr Berroeta, stated in respect of TPG’s mobile network:
The strength of our existing low-band mobile spectrum holdings ensured TPG Telecom did not have to participate in the Federal Government’s auction of additional 850 MHz and 900 MHz spectrum in December 2021.
Avoided spectrum investment frees up capital to accelerate other investment in our network, specifically the 5G upgrade, as well as the government mandated replacement of Huawei equipment over coming years.
The collective spectrum holdings of the merged Group, combined with the large investments we are making in our 5G network – part of the largest global 5G network under the Vodafone brand – mean we now have our best ever network.
Significantly, we converted more than 1,000 sites to 5G across our radio access network in 2021 and, with the introduction of our standalone 5G network and introduction of 700 MHz spectrum, achieved 5G coverage to 85% of the population in Australia’s top 10 cities.152
253 Optus is a wholly owned subsidiary of Singtel. Optus is Australia’s second-largest MNO. It has the second-largest number of mobile subscribers, and its network covers 98.5% of the population153 with its 5G network covering [REDACTED] of the population (as at August 2022).154 Optus operates around 2,500 mobile base stations in the RCZ.155
254 The CEO of Optus, Ms Bayer Rosmarin, gave evidence about Optus’s investments in its mobile network since 2015, stating:
In 2017, Optus announced that it had committed $1 billion of investment to improve regional mobile services by upgrading 1,800 sites from 3G to 4G and building 500 new 4G sites by the end of June 2018. That followed an ongoing program of investment by Optus which had directed more than $3.6 billion into its mobile networks since 2015. …
Despite this investment, Optus’ market share had not improved, and has remained largely static for the 10 years to December 2020. For the full year ended 31 March 2021, Optus’ EBITDA declined by 25% and EBIT declined by 77%.156
255 On 27 May 2021, Optus issued a media release concerning its financial performance for the financial year ending 31 March 2021 which recorded that operating revenue had declined 7% to $8.32 billion, EBITDA had declined 25% to $1.998 billion and NPAT was a loss of $208 million.157 The media release stated that these declines were “due to the COVID-19 pandemic impacts, lower NBN migration revenues, and market headwinds which led to lower equipment sales and leasing revenues as well as lower fixed broadband margins from higher NBN costs as customer bandwidth consumption continued to rise”.158 The media release also reported that:
Mobile service revenue returned to growth driven by higher postpaid revenue from increased penetration of Optus Choice plans. This resulted in total Optus ARPU postpaid growth of 4% compared to the corresponding half year.159
256 The Group CEO of Singtel, Yuen Kuan Moon, also gave evidence concerning Optus’s financial performance over the last 10 years and its importance within the Singtel group. Mr Moon said that Optus is one of the four group businesses within Singtel, with the other telecommunications businesses located in India, Indonesia, Thailand and the Philippines.160 Mr Moon said that Optus is an important part of the Singtel group, both strategically and in terms of its financial contribution. In FY22, Optus contributed more than 50% of Singtel's total group revenue, as well as EBITDA, and [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. However, Mr Moon said that Optus’s contribution is substantially less when considered in the context of other metrics. Based on underlying profit in FY22, Optus contributed [REDACTED] [REDACTED] [REDACTED] EBIT, and its overall contribution of profit to the Singtel group was [REDACTED] [REDACTED] [REDACTED]. Mr Moon also gave evidence that [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED]) [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED], [REDACTED] [REDACTED] [REDACTED]), [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED]; [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]).161
257 In its reasons for determination, the ACCC refers to only three MNOs: Telstra, Optus and TPG. However, the evidence before the Tribunal indicates that there are other smaller MNOs that have entered or are seeking to enter the market and provide mobile services.
258 An example is the Pivotel group of companies, which made two submissions to the ACCC. Pivotel operates a mobile and satellite telecommunications network as an MNO pursuant to a carrier licence issued by the Australian Communications and Media Authority under the Telecommunications Act 1997 (Cth) (Telecommunications Act). It has points of interconnect in the Australian major capital cities and points of interconnect internationally in Auckland, Los Angeles, and New York. The Commonwealth recently announced that Pivotel had been successful under the Regional Connectivity Program in two of its applications to build new community networks in Victoria and NSW delivering 18 new 4G base stations. These are in addition to other community networks co-funded by State or Federal governments in the Wickepin, Mt Barker and Northern Goldfields regions of Western Australia, as well privately funded 4G networks in the mining and agriculture sectors. Pivotel is also an MVNO reselling the Telstra 4G mobile network services under the Think Mobile brand, an MVNO reselling the TPG mobile network services under the Think Mobile and Reward Mobile brands, and a mobile virtual network enabler (MVNE) enabling other small MVNOs to resell the Telstra and TPG mobile network services.162
259 Another example is Field Solution Group, which is a provider of mobile infrastructure services and is working towards being an MNO by building networks in regional areas that are under-serviced.163
Mobile virtual network operators
260 MVNOs are mobile service providers who acquire wholesale mobile services from the MNOs (Telstra, Optus, and TPG) and provide retail mobile services under their own brands. MVNOs operate under a range of business models. Some MVNOs operate their own marketing and customer support, whereas others resell an entire ‘white-label’ service from an MNO.164
261 The MNOs also operate “sub-brands” that perform a similar role to MVNOs, offering retail services to more value-conscious consumers. These sub-brands also include formerly independent MVNOs that the MNOs have acquired, such as Amaysim which was acquired by Optus in 2021.165
262 In a telecommunications context, spectrum refers to the use of parts of the electromagnetic spectrum for the carriage of signals by radio equipment. Spectrum is the medium by which an end-user’s device is connected to a base station on a mobile network.166 The Tribunal considers it to be uncontroversial that spectrum is a scarce resource (in that the availability of spectrum is limited and subject to regulation of its use, as discussed below) and an essential input into the operation of mobile networks.
263 The legal right to operate radiocommunications devices on specific frequencies and over specific geographic areas is conferred by licence. The ACMA manages Australia’s spectrum and administers its planning and licensing. Spectrum licences confer to the licensee the exclusive right to operate radiocommunications equipment within the listed frequencies and geographic area. Most spectrum licences in Australia are owned by the nationwide MNOs and NBN Co. Spectrum licences are tradeable assets, subject to conditions imposed by the ACMA. Licensees may also authorise other people to operate equipment under their licence. The trade and authorisation of spectrum licences is outlined below.167
264 The purpose of spectrum planning is to enable the use of it by way of managing interference between competing users and uses of the spectrum. Without spectrum planning, interference between devices would render the spectrum less useful for all users.168 As noted above, the Tribunal considers that limiting the use of the spectrum in ways that manages interference necessarily means that planned, useful spectrum is a scarce resource.
Different characteristics and uses of spectrum bands
265 Spectrum is measured in megahertz (MHz) or gigahertz (GHz) and divided into bands based on frequency.169 In these reasons, the Tribunal refers to the 700 MHz, 800 MHz, 2 GHz, and 3.4 GHz bands. These are the names for the bands as they appear on the ACMA’s Register of Radiocommunications Licences. The application for authorisation refers to some of these bands by different names. Other names for these bands include:170
(a) 800 MHz: 850 MHz
(b) 2 GHz: 2100 MHz
(c) 3.4 GHz: 3400 MHz, 3500 MHz, 3600 MHz, 3.5 GHz, 3.6 GHz.171
266 Spectrum bands can be further categorised into “low-band”, “mid-band”, and “high-band”, based on their frequency and typical role within a mobile network. While the boundaries of these categories can vary, typical usage in the Australian context is as follows:
(a) Low-band spectrum is any spectrum below 1 GHz, and in Australia includes 700 MHz, 800 MHz, the 850 MHz Expansion band, and 900 MHz bands. Low-band spectrum has favourable propagation characteristics for wide coverage area networks, travelling comparatively greater distances and providing better service indoors than higher frequency bands.172 Low-band spectrum is well suited to overcoming natural and man-made barriers to signal propagation, such as hills and buildings, as well as foliage and other obstacles.173 Mr Turner explained that, by using lower band spectrum, fewer mobile sites are required to provide coverage over an area relative to the number of sites required when using mid or high-band spectrum.174 Mr Kanagaratnam said that the superior coverage of low-band spectrum makes it more suitable and attractive to MNOs for providing mobile service coverage, particularly over large geographic areas such as regional areas.175
(b) Mid-band spectrum is any spectrum between 1 GHz and 6 GHz, and includes the 1800 MHz, 2 GHz, 2.3 GHz, 2.5 GHz, and 3.4 GHz bands.176 Mid-band spectrum is distinguished by the comparatively greater amount available, allowing for larger bandwidths and more capacity to be provided in the network.177 In comparison to low-band spectrum, mid-band spectrum offers poorer propagation characteristics.178 Mr Lopez of TPG explained that because mid-band spectrum provides significant capacity but transmits information over shorter distances, it is ideal for use in high traffic metropolitan areas; it also means that an MNO will need to build more sites when using this spectrum compared to low-band.179 There was disagreement between Telstra and Optus concerning the suitability of mid-band spectrum in regional areas. Mr Turner of Optus said that mid-band spectrum is an important ingredient when seeking to increase network capacity in regional areas in circumstances where mobile sites are close enough to populated areas for the mid-band spectrum to carry or travel to those areas. He said that Optus’s live network statistics in regional Australia show that mid-band spectrum in the 1,800MHz and 2,100MHz bands can serve customers in a range of up to 6.6 km-14.5 km from network sites, and spectrum in the 2,600MHz band can provide network capacity for customers in a range of 3.5 km-6.6km from relevant sites in regional areas. Approximately [REDACTED] of Optus’s regional mobile customers are within 6.6 km of relevant sites.180 However, Mr Meissner said that Telstra considers that mid-band spectrum is not suitable for use in regional areas because of the shorter distance it can be transmitted. While a large proportion of the population may live within an appropriate distance of a tower, that does not mean that a mobile service can effectively be provided to those customers using mid-band spectrum. That is because customers do not only use their mobile services where they live, they use them where they go. Mr Meissner said that, in his experience, customers in regional areas require continuity of coverage across a large area and not just small islands of coverage near where they live. Only low-band frequencies provide the necessary continuity of coverage and mid-band spectrum is not suitable for this purpose.181 In that regard, the Tribunal notes that the fixed wireless services under the MOCN Service Agreement is proposed to be provided using the combined 3.6 GHz spectrum of Telstra and TPG.182
(c) High-band spectrum is any spectrum above 6 GHz, and includes spectrum at very high frequencies such as the 26 GHz band. High-band spectrum makes very large bandwidths available but offers relatively poor propagation.183 In addition to mobile services, high-band spectrum is also used by the satellite industry to provide satellite services where the end-user device (such as a handset or satellite dish) has line of sight to the satellite overhead, and terrestrial propagation issues are not a concern.184 No high-band spectrum is included in the Proposed Transaction.185
267 Operators may acquire spectrum licences from the ACMA by application or at auction, or in the secondary market from other licensees.186
268 The ACMA may allocate spectrum licences through an auction, tender, or by a predetermined or negotiated price. In practice spectrum licences are generally allocated via auction.187 A spectrum auction is the final step in the ACMA’s process of allocating suitable spectrum to licensees. The Radiocommunications Act governs the ACMA’s process of spectrum band planning and allocation.188 The object of the Radiocommunications Act is “to promote the long term public interest derived from the use of the spectrum”.189
269 The ACMA may allocate nationwide licences or allocate licences for discrete regions of Australia. The 700 MHz band was auctioned as nationwide licences. Other bands such as 800 MHz, 2 GHz, and 3.4 GHz have been auctioned as sub-national licences. The ACMA takes likely demand for the spectrum into account when designing geographic lots.190
270 When allocating spectrum, the ACMA typically divides a band into metropolitan and regional areas, and further into sub-national lots, informed by likely demand for the spectrum. The following figure illustrates the regional lot configuration for the 3.6 GHz auction in 2018, with each colour a different region.191
Regional licence areas in the 3.6 (3.4) GHz band, as auctioned
271 MNO holdings typically vary between these sub-regional lots. For example, the 1800 MHz and 3.4 GHz bands are especially fragmented, with MNOs holding different amounts in different areas. Optus for example holds 65 MHz in the 3.4 GHz band in regional Western Australia (magenta in the above figure), but only 30 MHz in regional New South Wales (red in the above figure).192
272 Under s 60(1) of the Radiocommunications Act, the ACMA is required to determine the procedures to be applied in allocating spectrum licences. Under s 60(5), the procedures may impose limits on the aggregate spectrum to be held by any one person or by a specified person. Under s 60(13A), the ACMA is required to consult with the ACCC about whether the procedures should impose limits.193 The ACCC, in making its assessment of the need for allocation limits, uses the long term interests of end-users (LTIE) test. The ACCC’s assessment of the need for allocation limits is forward-looking, and is provided to the ACMA prior to the ACMA making a determination on the procedures to be applied when allocating spectrum licences.194 However, the ACMA is not required to follow any recommendation made by the ACCC. In contrast, under s 60(10), the responsible Minister is empowered to give directions to the ACMA in relation to the exercise of the power to determine procedures imposing a limit on spectrum allocation. By s 60(12), the ACMA must exercise its powers in a manner that is consistent with any direction given by the Minister.
273 Allocation limits apply only to the auction itself and do not restrict the secondary trading or authorisation of licences subsequent to the auction. Secondary trading or authorisation of spectrum licences is taken to be an acquisition for the purposes of s 50 of the CCA.195
274 The ACMA may issue spectrum licences with a licence term of up to 20 (previously 15) years.196 The ACMA also has the power to renew spectrum licences upon expiry without undertaking an auction or other price-based allocation method. The circumstances under which the ACMA may re-issue a spectrum licence differ depending on when the licence was issued and the terms included in the licence.197
275 The spectrum bands relevant to the Proposed Transaction are all due to expire within the first 10 year term of the agreements. The ACMA has indicated that it will begin consideration of the renewal process for the first of these bands to expire (800 MHz) in 2023.198
276 The ACMA has previously re-issued certain spectrum licences upon expiry, including in the 800 MHz, 1800 MHz, 2 GHz, 2.3 GHz, and 3.4 GHz bands. These spectrum bands are generally issued for a period of 15 years and, to date, they have only expired once since their issue as spectrum licences. Upon expiry, spectrum in these bands was not put back to market but instead re-issued to the incumbent licensees for a further 15 years.199
277 The spectrum licences relevant to the Proposed Transaction and their expiry is set out in the following table:200
Spectrum bands relevant to the Proposed Transaction
Spectrum band | Expiry date | Notes | |
700 MHz | 31 December 2029 | ||
800 MHz | 17 June 2028 | Previously re-issued in 2015 | |
2 GHz | 11 October 2032 | Previously re-issued in 2017 | |
3.4 GHz | 13 December 2030 | See note201 |
278 As mentioned above, spectrum licences are tradeable assets and may be traded and sold to other parties. Licensees are also permitted to trade parts of their licences.202 The subdivision of spectrum licences is subject to rules put in place by the ACMA, but generally allows for the disaggregation (or aggregation) of licences either by frequency or geographic area.203 Licensees may also authorise other parties to operate radiocommunications devices under their licence. This is known as “third-party authorisation”. The Proposed Transaction is a third-party authorisation and involves TPG authorising Telstra to operate equipment on TPG’s licences for the purpose of implementing the network sharing arrangement.204 By operation of s 68A of the Radiocommunications Act, TPG’s authorisation of Telstra to operate radiocommunications devices under TPG’s spectrum licence is taken to be an acquisition by Telstra of an asset of TPG, and conduct engaged in by Telstra, to which s 50 of the Act and related provisions apply.
279 TPG and Telstra have a separate spectrum access agreement in the 3.4 GHz band, with Telstra operating equipment at greater bandwidths than are licensed to them in selected capital cities.205
280 Capacity on a mobile network is a function of site density, spectrum deployments and radio technology. As the MNOs densify their networks, particularly in major metropolitan areas and denser towns, their networks are able to serve more traffic. This enables more customers to use the network, or enables the MNO to provide greater speeds or other capabilities to end-users.206
281 Witnesses on behalf of each of the parties explained the manner in which the capacity of mobile networks is able to be increased, and the commercial advantage of adding spectrum to the network:
(a) Mr Meissner of Telstra stated as follows:207
17 Spectrum is an essential input in the supply of mobile services. Voice and data communications between mobile devices and a RAN cannot occur without the use of appropriate and compatible spectrum. Mobile network operators (MNO) acquire rights to use different frequencies of spectrum in order to provide mobile services to users. Different frequencies have different physical characteristics which can be exploited for different uses. I describe these in paragraph 22 below.
18 Because a MNO’s customers in a geographic area all share use of the same spectrum, there is a direct relationship between the amount of spectrum that a MNO has a licence to use and the per user data speed that its mobile network can support. In effect, if two MNOs have the same number of customers in the same area, in order to deliver the same capacity (i.e. speed) to each customer, a MNO with less spectrum will need to compensate this by investing in more cell sites (i.e. densification) or through better spectral efficiency of its RAN equipment. Spectral efficiency refers to the amount of data that can be transferred over a given spectrum bandwidth with minimal transmission errors. One element of the value of spectrum is therefore that it helps MNOs avoid the need to invest as much in infrastructure in order to achieve the same service performance for customers.
…
24 As described in paragraph 18 above, assuming a fixed volume of spectrum and a fixed number of sites, the data speed that can be supported by a network will decrease when the number of users increase. The number of users connected to a particular site varies, meaning that it is important to consider the amount of spectrum (i.e. MHz) held at each site per customer. This measure provides a more accurate reflection of the network capacity which is available to a customer for downloading or uploading data when compared to measuring a MNOs entire spectrum allocation against its entire customer base.
(b) Mr Katinakis of Telstra explained that the capacity of a mobile network within a given area can be expanded by adding spectrum bands or by adding additional mobile sites, stating:208
As a general principle, as data usage increases without any investment in expanding capacity at a site, customers typically experience increasing degradation of service. As network speeds fall, Telstra will typically therefore make ongoing investments to expand network capacity, such as through sectorisation or the addition of spectrum bands. Sectorisation involves splitting a cell site, so the traffic is split over three cells instead of the original one cell.
(c) Mr Turner of Optus said that the coverage, capacity and capability achievable by a mobile network is dependent on a mix of the number and location of mobile sites (including the types of equipment used on those sites), the technologies deployed and the amount of spectrum, at varying ranges, that an MNO has access to.209
(d) Mr Kanagaratnam of Optus gave similar evidence. He explained that:210
86. The potential capacity and speed of a mobile network may be increased by acquiring more spectrum. All other things being equal, an MNO with higher quantities of spectrum will be able to offer greater capacity and speeds on its mobile network than an MNO with lower quantities of spectrum. The other key way to increase the potential capacity and speeds of a mobile network is by building more sites, which is sometimes referred to as 'densification'.
87. The amount of spectrum, or bandwidth, available to an MNO directly influences the overall amount of traffic that can be carried and speeds that can be achieved on a base station or within a geographical area. This can be the overall throughput demanded by the connected users, in the case of data, or the number of simultaneous connections that can be supported in the case of a voice or video call.
282 Mr Turner of Optus explained the competitive advantages to MNOs from spectrum holdings in the following manner:211
As I describe below, in my experience the true advantage that can be achieved by competing MNOs through spectrum access depends not only on total spectrum holdings, but also the areas in which that spectrum can be used and on access to mobile sites. Overall spectrum holdings are, however, an important measure of an MNO's ability to provide service quality because:
(a) the greater the amount of spectrum held, the greater the capacity and speed achievable by each mobile site in the area where that spectrum is licensed for use; and
(b) spectrum is a finite resource, and so, unlike mobile sites, which subject to capital expenditure constraints can be built or upgraded at any time, there are fixed limits on how much spectrum can be accessed by any one MNO at a particular time.
For these reasons, MNOs view spectrum as a key strategic asset. In my experience, Optus uses overall spectrum holdings as a key indicator of how competitive it can be.
283 The national MNOs are currently operating networks that include 3 generations of mobile technology: 3G, 4G, and 5G. These technology generations are generally defined and standardised at an international level, and require support both within the mobile network and on an end-user’s device.212
284 Each subsequent technology generation has brought increased bandwidth and speeds and improved the capabilities of the network. 5G is the newest technology generation to be deployed, and all 3 MNOs are rolling out their 5G networks currently.213 5G technology makes more efficient use of spectrum, delivers faster speeds and provides better reliability and lower latency as compared to 4G technology. This technological development enables network operators to offer improved services, both fixed and mobile.214
285 Each subsequent technology generation uses spectrum more efficiently, enabling faster speeds or more capacity to be provided using the same parcel of licensed spectrum. A given allotment of low-band spectrum may be used to carry more traffic or cater to more end-user devices simultaneously on a 5G network than on a 4G or 3G network.215
286 MNOs are incentivised to upgrade their networks in order to make use of this more efficient technology and meet evolving consumer needs, but doing so requires large upfront investments. MNOs must also balance repurposing (“re-farming”) their spectrum holdings for newer technology while continuing to operate the older technology simultaneously.216
287 5G technology is designed to be more flexible with the spectrum it is able to use, with more and more equipment expected to become available supporting a wider range of spectrum bands. Currently, the MNOs are making heavy use of the 3.4 GHz band for their 5G networks, with some use of low-band spectrum in the 700 MHz (TPG), 800 MHz (Telstra) and 900 MHz (Optus) bands.217
288 When rolling out new technology generations, MNOs reuse or incrementally add to their existing physical infrastructure. For example, an MNO might add a 5G radio at an existing site where it already operates 4G and 3G equipment.218 The speed at which MNOs are able to do so is determined by spectral and capital availability, but also the availability of suitable sites and sufficient existing backhaul capacity.219
289 In its reasons for determination, the ACCC also produced a number of maps depicting the relative extent of each MNO’s backhaul transmission networks. Backhaul transmission may be a fibre link, but microwave (point to point wireless) links are also extensively used.220 The maps prepared by the ACCC with respect to backhaul transmission are not contested by the parties, although Telstra has noted that they are unable to verify the maps concerning the networks of Optus and TPG.221 The Tribunal does not consider it necessary to reproduce the maps, but records the following facts concerning the backhaul networks of the parties which are not contested.
290 The MNOs operate extensive fibre networks across Australia. For a mobile network, transmission from the mobile site back to the core network is a key component in the performance and cost of the mobile service.222
291 Telstra has an extensive network of regional fibre, built up over time as a legacy of its role as a former statutory monopoly.223 Much of this fibre is subject to access regulation via the Domestic Transmission Capacity Service, and provides capacity for not only Telstra mobile sites, but Optus and TPG sites also.224
292 Optus’s fibre network is [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. However, Optus also operates [REDACTED] [REDACTED]-[REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].225 Optus’s mobile network in regional areas is also [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].226
293 TPG’s fibre network is [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].227 TPG’s mobile sites in regional areas [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED]-[REDACTED]- [REDACTED] [REDACTED].228
294 The commercial and economic benefits of mobile network infrastructure sharing are readily apparent. As Mr Feasey explained in his report:229
11. The sharing or joint exploitation of network assets, as the agreement envisages, is not new to mobile telecommunications markets, either in Australia or elsewhere. In my experience, mobile networks involve fixed and operating costs which represent a significant proportion of the total costs to be borne by the industry and its customers. Sharing these assets allows greater efficiencies or economies of scale to be realised, reducing average costs for those concerned. The benefits of lower costs can be particularly significant in less densely populated areas, where individual operators may otherwise find it uneconomic to provide network coverage, either at all or to the extent that is possible if the network is shared.
12. The sharing or pooling of spectrum can, in my experience, also improve the utilisation of assets which might otherwise remain underexploited. This is particularly so if an operator with significant spectrum holdings lacks network assets in a particular geographic area and so has little or no demand for network capacity in that area, whilst another operator in the same area is capacity constrained. In such a case, pooling or sharing spectrum will allow it to be put to work, allowing the capacity constrained operator to serve demand more quickly and at lower cost. In return, the holder of the spectrum will obtain access to the network (or a better network) in that area, allowing it to begin to offer services and utilize spectrum which would otherwise have been left idle. Mismatches between the demand for additional network capacity and opportunities to meet it by deploying additional spectrum arise because spectrum assignment processes are infrequent and/or because regulators impose constraints on how spectrum is to be assigned or might subsequently be reassigned amongst firms. These can contribute to inefficiencies and mean that demand from customers remains unserved or served inefficiently. Spectrum sharing or pooling arrangements are one way in which some of these inefficiencies can be overcome.
295 The MNOs in Australia have historically shared aspects of their networks. Infrastructure sharing in mobile networks can be broadly classified as either ‘active’ sharing or ‘passive’ sharing.230
296 Passive infrastructure sharing may involve the sharing of non-electronic infrastructure such as cell sites, towers, and buildings, but does not include the sharing of electronic equipment capable of processing or converting telecommunications signals such as radio equipment or spectrum (which is described as active sharing).231 The co-location of mobile sites, a form of passive infrastructure sharing, is facilitated by the Telecommunications Act.232 The MNOs have historically built and managed passive infrastructure as part of their networks. There are also a number of independent third-party infrastructure providers that build and maintain passive infrastructure, supplying the MNOs and other access seekers.233 Telstra, Optus and TPG have all recently divested some of their passive infrastructure into separate entities, in part to fund the rollout of their 5G networks. Telstra, for example, has sold a 49% stake in its tower business (now Amplitel Pty Ltd, Amplitel), while Optus has sold a 70% stake in its tower business (formerly Australia Tower Network Pty Ltd, now named Indara Digital Infrastructure).234
297 One model of active infrastructure sharing is called domestic roaming. Roaming involves a host MNO carrying the traffic of another MNO on its behalf in a specific area. The client MNO is not required to deploy any infrastructure in the relevant area.235 Access to domestic roaming services in Australia is by commercial agreement. The ACCC has conducted public declaration inquiries in 1998, 2005, and 2017, deciding each time to not declare (regulate) a domestic roaming service (under Pt XIC of the CCA).236 In its final report published in October 2017 in its domestic mobile roaming declaration inquiry, the ACCC observed that mobile communications in regional areas was a focus of its inquiry and that the ACCC had examined how the declaration of a roaming service under Pt XIC of the CCA would impact incentives for MNOs to continue to invest in regional areas.237 One of the statutory factors assessed by the ACCC was whether declaration of a mobile roaming service would promote competition in retail mobile services. It should be noted that, at the time of the inquiry, TPG was a new entrant into the mobile services market and supplied services independently of Vodafone Hutchinson Australia. The ACCC concluded that declaration would not promote competition in the retail mobile services market to a significant extent.238 There were two primary reasons for that conclusion. First, the ACCC formed the view that overall geographic coverage is not the primary driver of competition, nor is it essential for MNOs to have equal geographic coverage to compete effectively in the market. Second, and perhaps more significantly, the ACCC formed the view that:239
Removing geographic coverage as a point of differentiation would likely reduce the benefits that an MNO gains from extending its network beyond that of its rivals. We consider that Telstra’s and, to a lesser extent, Optus’ superior network coverage is an outcome of the competitive dynamics in the market and declaration would reduce these dynamics. Networks that differ from each other in terms of coverage, technology and quality provide more choice for consumers and more competitive tension between operators. Declaration would have the effect, therefore, of reducing competition over this dimension of rivalry between MNOs such that it would be likely to distort long-term competitive dynamics and reduce incentives for MNOs to differentiate their services in this way.
298 In its final report, the ACCC also considered whether competition to offer a roaming service is likely to be effective. The ACCC expressed the view that, in geographic areas only served by Telstra, competition may not be effective; in areas served by Telstra and Optus, it is unclear whether the market for roaming services will continue to be competitive (although the ACCC considered that there should continue to be commercial incentives to provide mobile roaming services); and in areas served by three MNOs, the market for wholesale mobile roaming services is likely to be competitive.240
299 MOCN and MORAN arrangements are also models of active infrastructure sharing. Both models involve multiple MNO parties sharing active assets in certain coverage areas of their networks.241 Typical MORAN deployments include the sharing of active base stations, but not spectrum. MOCN deployments typically include the sharing of active base stations, as well as spectrum shared and owned by the operators.242 Mr Lopez of TPG explained that, from a commercial or operational perspective, a MOCN can take a variety of forms, including:
(a) a network as a service (NaaS) whereby one MNO owns and operates all of the RAN equipment, and shares that equipment with another MNO, within a geographical area (with both MNOs retaining their independent core networks); and
(b) a joint venture whereby each MNO owns and operates RAN equipment that is shared by the other MNO within specified geographical areas, and jointly run the shared network.243
300 The applicants refer to the Proposed Transaction as a MOCN, on the basis that the active infrastructure in the RCZ is shared by both parties, and the spectrum utilised is a combination of spectrum owned and operated by Telstra, and spectrum owned by TPG and authorised to Telstra for use in the RCZ and beyond. The ACCC noted in its determination that the characterisation of the Proposed Transaction as a MOCN was contested, largely due to the fact that TPG’s principal contribution to the physical assets of the network is limited to its spectrum. The Tribunal does not consider that the debate over the name to be given to the arrangement furthers the analysis required under s 90(7). The evidence before the Tribunal also included some analysis of comparable arrangements in other countries, particularly Canada. That material did not receive any substantive attention by the parties during the hearing before the Tribunal. The Tribunal notes that there are significant difficulties in making comparisons with commercial arrangements entered into in other countries. To produce meaningful information, the comparison would need to be based on a detailed analysis of the other commercial arrangements and the demographic, economic and commercial circumstances in the other country, including in particular the extent of competition. In the present proceeding, the Tribunal does not consider that the comparative material is of any assistance.
301 A number of the witnesses gave evidence explaining the different competitive effects of the different forms of infrastructure sharing. In the Tribunal’s view, the competitive effects of any form of infrastructure sharing largely depends upon the prevailing market structure. With that caveat, it can be accepted that passive infrastructure sharing has a lower potential for anti-competitive effects and a greater potential for pro-competitive effects. That is because passive infrastructure sharing is a form of asset and capital management and enables the accessing mobile operator to compete freely in the supply of mobile services, having complete control over the price and service quality offered to consumers. At the other end of the spectrum of possible arrangements, roaming has lower potential pro-competitive effects. Mr Chiarelli of TPG explained that, under a roaming arrangement, and similarly when operating as an MVNO, the services that can be offered by the party which is accessing the host network are limited to the services which the host provides to its customers, or a subset of those services, and nothing else.244
302 In between are other forms of active infrastructure sharing such as MOCN or MORAN arrangements. Mr Chiarelli gave evidence that a MOCN arrangement permits each MNO to control its own product and plan development and offer new plans and products to the market in a manner which will better allow for it to compete through product differentiation than is the case with roaming. He explained that product plans (e.g. plans with different data inclusions) and new products (e.g. voice over 5G) are built and controlled in the mobile core network of a party, which, in a MOCN arrangement, continues to be controlled by each party independently.245 Mr Chiarelli also explained that there are significant differences in customer experience between a roaming arrangement and a MOCN. Under a roaming arrangement, customers can experience call failures as the customer moves between networks (from the service providers own network to the accessed network and back again). Such call failures do not occur under a MOCN because the customer’s call remains controlled by the service provider’s core network. That difference also affects the ability of a service provider to diagnose and respond to customer call problems.246
303 Mr Katinakis gave evidence that, in his experience, a MOCN offers a better commercial option for both parties, compared with other infrastructure sharing arrangements such as a MORAN. He said that while a MORAN provides cost savings through shared RAN infrastructure, that shared infrastructure needs to use different spectrum to serve different customer bases. This imposes substantial additional cost and complexity, because the same equipment needs to operate using different spectrum to different customers. A MOCN is more efficient as it allows the MNOs to pool their respective spectrum allocations, which increases the number of usable frequency blocks and improves data speeds, and means that the network equipment uses the same spectrum to service all customers.247
Mobile broadband and fixed wireless
304 Mobile telecommunications networks are increasingly serving multiple purposes. All three national mobile networks in Australia are capable of providing fixed broadband services (known as fixed wireless), comparable to those delivered over the NBN or other fixed broadband networks. 5G networks in particular, with their greater bandwidths and more spectrally efficient technology, are well suited to providing fixed broadband services.248 All three MNOs offer standalone mobile broadband services in addition to the typical mobile broadband service bundled with their retail plans including calls and SMS.249
305 In its Communications Market Report 2020-21 (issued in December 2021), the ACCC stated:250
With the rollout of 5G, network operators have the opportunity to not only offer improved mobile broadband, but also provide fixed wireless alternatives to homes and small businesses, as an alternative to traditional fixed-line broadband.
However, the extent to which 5G mobile and fixed wireless services could become a substitute for fixed-line broadband services in terms of speed, data allowance and price is unclear. 5G services in some areas are becoming increasingly attractive to consumers as an alternative to fixed-line services. Nevertheless, the technology currently has a limited geographic footprint, and will unlikely be able to service many fixed-line broadband end-users for some time.
All 3 of the MNOs now offer some form of 5G home broadband product, which is generally price competitive with comparable NBN plans.
306 The MNOs also compete for customers in the enterprise and government segment. Enterprise and government customers typically have much larger data requirements and require broadband services in multiple locations across Australia.251
Nature and extent of competition between the national MNOs
Relevant markets
307 As stated earlier, the identification of markets is an essential first step in any assessment of competition and the likely competitive effects of commercial agreements or other conduct.
308 There was no real debate between the parties about the markets relevant to the assessment of the competitive effects of the Proposed Conduct and the Proposed Transaction. The principal focus of the evidence and the submissions of the parties was the national market for the retail supply of mobile telecommunications services. Those services include mobile phone services (which typically comprise a bundle of voice, short message services (SMS) and data services) and standalone mobile broadband services. The parties did not contest the ACCC’s finding that the existence of a national market for the retail supply of mobile telecommunications services is supported by the following considerations:252
(a) Mobile services are provided to consumers by their MNOs (and MVNOs) on a national basis, with consumers able to use services anywhere in Australia that their MNO (or MVNO) has network coverage. Mobile services are not advertised as being specifically for customers living in metropolitan or regional areas.
(b) MNOs (and MVNOs) set prices on a nationally consistent basis. MNOs (and MVNOs) offer the same range of retail services in all areas where they have coverage at the same price.
(c) Decisions to build infrastructure in any given geographic area are made having regard to the effect of any associated investment on market shares in all areas of the country.
309 While it is common ground that the market for the retail supply of mobile telecommunications services is national in its geographic extent, the Tribunal agrees with the ACCC that it is important to have regard to geographic variations in the nature and extent of competition within the market.253 This is because the nature and extent of competition in different geographic areas is dependent on the nature and extent of the MNOs access to mobile network infrastructure in the area (whether by ownership or by infrastructure sharing arrangements), and such access varies across Australia and particularly in the RCZ.
310 It is also common ground that MNOs compete in the market for the retail supply of mobile telecommunications services on the basis of price and the quality dimensions of the mobile service. As discussed below, the two most important quality dimensions of mobile services are geographic coverage and service quality (which encompasses the quality of a phone call, the speed of data transmission and overall reliability of the service). Service quality is affected by mobile network capacity in a given location. Other quality dimensions of mobile services include individual services supplied by MNOs and MVNOs such as voice mail, account information and billing services. The following findings made by the ACCC are uncontentious:254
MNOs compete on a range of factors to gain an advantage over their rivals. First, MNOs compete over the coverage, speed, and other quality dimensions of the mobile services they offer to customers. These quality dimensions are directly influenced by the nature and extent of the underlying network infrastructure the MNO owns or has access to. Second, depending on the existing network infrastructure an MNO has at any given point in time, MNOs will compete based on price and inclusions (including data and content) made available in their offerings.
MNOs make ongoing decisions regarding how much to invest in their network coverage and quality. These decisions extend to how much spectrum to acquire (and at what price), how far to extend their geographic network coverage, what generations of mobile network technology to provide over which areas of their network, and the depth/density of their network coverage. Network investment is continuously evolving and ongoing.
As an alternative to building their own infrastructure, MNOs may in some instances enter agreements to access services provided by the network infrastructure of other MNOs or third parties. Such agreements can enable the acquiring MNO to offer higher service coverage and quality than they could using only their own networks.
311 A secondary focus of the evidence and the submissions of the parties was the market for the wholesale supply of mobile telecommunications services. The wholesale supply of mobile telecommunications services encompasses a range of potential service dimensions. The most complete form of wholesale supply are arrangements by which MVNOs are authorised to resell an MNO’s mobile service. Infrastructure sharing arrangements between MNOs are also forms of wholesale supply, although they could possibly be described as infrastructure services. As described earlier in these reasons, such arrangements can comprise domestic roaming (by which a host MNO carries the traffic of another MNO on its behalf) and MOCN and MORAN arrangements. Demand for all such wholesale services is, of course, derivative of the demand for retail mobile telecommunications services. Generally, an MNO has the capability to offer each of the above forms of wholesale service. Many of the characteristics of the retail market for the supply of mobile services are also applicable to the wholesale market in the sense that the nature and extent of competition in the wholesale market in a given area is dependent on the MNO’s access to mobile network infrastructure in the area. However, the wholesale market is also affected by structural considerations, specifically vertical integration. The owners of the mobile networks, the MNOs, compete in the market for the retail supply of mobile telecommunications services. It follows that any decision by an MNO to offer a wholesale service of a particular kind will be influenced by the likely effect of that decision in the retail market. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].255
312 In its reasons for determination, the ACCC also referred to the potential effects of the Proposed Transaction on other markets, particularly:
(a) the input markets being the primary and secondary markets for the acquisition of spectrum and the market for the supply of passive tower infrastructure services; and
(b) the retail supply of fixed wireless services, the supply of enterprise grade mobile services and the supply of NBIoT services.
313 Before the Tribunal, the parties did not advance substantive submissions about those markets. It is not apparent to the Tribunal that the Proposed Transaction will substantially lessen competition in the primary and secondary markets for the acquisition of spectrum and the market for the supply of passive tower infrastructure services. None of the parties advanced substantive submissions that would support a conclusion that the retail supply of fixed wireless services, the supply of enterprise grade mobile services or the supply of NBIoT services warranted separate analysis in the assessment of the likely competitive effects or benefits or detriments of the Proposed Conduct or the Proposed Transaction.
314 For those reasons, the Tribunal’s assessment is principally focussed on the retail and wholesale supply of mobile telecommunications services without differentiating between specific products such as voice calls or data services. The Tribunal is also conscious that fixed wireless services, and particularly 5G services, are a potential substitute for fixed broadband services. It follows that, if and to the extent the Proposed Conduct or the Proposed Transaction increases competition in the supply of fixed wireless services, competition in the wider market for the supply of broadband services would also increase. The Tribunal also notes for completeness that low orbit satellites are currently being used to supply broadband services, and the ACCC found that satellites may, in the future, be able to act as substitutes to terrestrial networks in the provision of NBIoT connectivity.256
315 This identification of markets relevant to the statutory criteria for authorisation is largely consistent with the views of the economists who have provided reports on behalf of the parties.257
Market shares of the MNOs and the rate of churn
316 The MNOs are vertically integrated. They compete at both the retail and wholesale level, with their retail brands competing for consumer and small business customers, and their wholesale arms competing in the provision of wholesale services to MVNOs and other telecommunications providers. The MNOs also compete for enterprise and government customers.258
317 In addition to their “flagship” brands (Telstra, Optus and Vodafone), the MNOs also operate “sub-brands”, which may offer different quality of service at lower prices or with greater inclusions; and which may be targeted at particular customer segments or niches compared to their flagship brands. These sub-brands include Belong (Telstra), Gomo and Amaysim (Optus) and TPG, Lebara and Felix (TPG).259
318 Since the introduction of competition in mobile telecommunications, the relative positions of the three MNOs has been relatively stable. Historically, Telstra has had the greatest share of retail services, followed by Optus and finally TPG (including its precursors).260 In recent years, Telstra’s share of services in operation has been increasing, while Optus’s share has been relatively stable – with the exception of 2021, discussed further below. TPG’s share has steadily fallen since the merger of Vodafone Australia Limited and Hutchison 3G Australia Pty Ltd in 2009.261
319 Estimated retail market shares vary by source. The figure below sets out estimated market shares for retail mobile services from FY03 to FY21, compiled from ACCC Communications Market Reports:262
Estimated market share for retail mobile services between 2002-03 and 2020-21
320 As detailed in the above figure, as at 2021 Telstra continued to have the largest nationwide share of retail services in operation, with approximately 44% of services in operation. Optus had approximately 31% of services nationwide, with TPG approximately 17%.263 The most recent market share data available from ACCC Communications Market Reports dates to the 2020-21 financial year, and may not reflect recent developments such as the Optus data breach in September 2022 which may have caused some customers to churn away from Optus.264 The number of customers acquiring services via MVNOs declined in 2021 from 15% to 9%. In large part, this was due to large MVNOs being acquired by MNOs (for example, Optus’s acquisition of Amaysim).265
321 Estimates of the number of retail services in operation also vary by source. However, the number of services in operation numbers in the tens of millions. Telstra alone reported almost 21 million retail services in operation in 2022. Retail mobile services are acquired by virtually the entire Australian adult population.266
322 In its reasons for determination,267 the ACCC included two figures showing the ACCC’s estimate of each MNO’s metropolitan and regional share of retail mobile services in operation by State and Territory as at June 2022. The data shown in the figures was based on the ACCC’s analysis of data provided by Telstra, TPG and Optus. Each MNO provided its own classifications of locations of services in operation into regional and metropolitan categories. Service location was based on the location of the mobile site used most frequently by the service (Telstra), the typical overnight location of the service (TPG), or the location of the primary cell showing the highest correlation to the service’s movements (Optus). The figures do not include enterprise government services. Significantly, Optus’s services in operation data did not include pre-paid services. Generally, around 36% of services are pre-paid services. It follows that Optus’s market share is likely to be understated in these figures. The two figures prepared by the ACCC are as follows:
MNO shares of retail services in operation in metropolitan areas, June 2022
[REDACTED]
MNO shares of retail services in operation in regional areas, June 2022
[REDACTED]
323 The Joint Document of Factual Findings included the above figures as matters not contested by the parties,268 although the applicants commented that they do not accept that separate markets exist for metropolitan and regional services. No party contended before the Tribunal that there are separate markets for mobile services supplied in metropolitan and regional areas, and all parties accepted that the market for mobile services is national. The Tribunal considers, though, that the nature and extent of competition can vary across a market. As discussed later in these reasons, the Proposed Transaction has implications for customers resident in metropolitan areas who travel to regional areas, and implications for customers resident in regional areas. The effects of the Proposed Transaction are likely to differ across those customer segments. For that reason, it is relevant to consider the different shares of retail mobile services in operation held by the MNOs in different metropolitan and regional areas.
324 Optus’s internal assessments of its market share differ somewhat from the above figures. In an internal document prepared in May 2021 for Optus’s executives, Optus estimated its metropolitan market share as around [REDACTED] and its regional share as around [REDACTED].269 A model prepared by Optus in July 2022 to analyse the effects of the Proposed Transaction, which is discussed further below, similarly estimated Optus’s current metropolitan and regional market shares as around [REDACTED] [REDACTED] [REDACTED].
325 Due to the way retail plans are structured, there is little distinction between market shares for “4G customers” and “5G customers”. Retail mobile plans tend to include a bundle of inclusions, which for the MNOs’ flagship brands includes use of their 5G networks for all current post-paid plans.270 The number of 5G retail customers each MNO has at a given point in time is therefore a function of the relative number of customers with compatible 5G handsets on 5G-enabled plans.271
326 The ACCC also analysed churn data provided by each of TPG, Telstra, and Optus and prepared charts which illustrate the competitive dynamics between the MNOs and how these dynamics differ between regional and non-regional areas. The data only includes churn between the MNOs and not the MVNOs.272
327 The figure below shows total customer churn in and out of TPG in the 2021-22 financial year, overall (on the left) and in the RCZ (on the right):273
TPG mobile customer churn, overall and in regional areas, FY22
[REDACTED]
328 It can be seen from the above figure that:
(a) [REDACTED];
(b) [REDACTED].274
329 The following two figures show customer churn in and out of Telstra and Optus respectively in the 2021-22 financial year, overall (on the left) and in the RCZ (on the right):275
Telstra mobile customer churn, overall and in regional areas, FY22
[REDACTED]
Optus mobile customer churn, overall and in regional areas, FY22
[REDACTED]
330 The above figures show:
(a) [REDACTED];
(b) [REDACTED];
(c) [REDACTED].276
331 In its reasons for determination, the ACCC expressed the view that the above data indicates a higher level of churn and competition between Telstra and Optus than with TPG, particularly in regional areas.277 The Tribunal agrees with that view.
332 The above churn data suggests that the rate of customer churn between the MNOs is [REDACTED], with the rate of churn for Telstra customers being [REDACTED] [REDACTED]. In its FY22 Annual Report, Telstra reported almost 21 million SIOs.278 The above chart in respect of Telstra shows that, in FY22, Telstra gained approximately [REDACTED] customers from, and lost approximately [REDACTED] to, Optus and TPG. The gains and losses were each about [REDACTED] of Telstra’s total SIOs. Mr Ackland’s evidence was that Telstra has [REDACTED] SIOs in the RCZ.279 The above chart in respect of Telstra shows that, in FY22 and in the RCZ, Telstra gained approximately [REDACTED] customers from, and lost approximately [REDACTED] customers to, Optus and TPG. The gains and losses were each about [REDACTED] of Telstra’s total SIOs in the RCZ. While the data excludes churn with MVNOs (and MVNOs may serve more price sensitive customers and thereby generate higher churn rates280), the data suggests an overall [REDACTED] rate of churn.
333 Colloquially, low churn rates within a market is often referred to as customer “stickiness”. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] “[REDACTED] [REDACTED]” ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]) [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].281 Optus’s internal papers refer to the difficulty of “unrusting customers from Telstra”.282
334 The Tribunal considers that the mobile telecommunications market can be described as highly concentrated. In that regard, the Tribunal agrees with the view expressed by CEPA in its expert economic report dated 24 June 2022 prepared for Optus that the Australian mobile telecommunications market can be characterised as an oligopoly with three vertically integrated MNOs – Telstra, Optus and TPG – accounting for 91% of the total market for retail mobile services and a near 100% of wholesale services.283
Barriers to entry and expansion in mobile telecommunications markets
335 In its reasons for determination, the ACCC made findings with respect to the barriers to entry and expansion in the markets for mobile telecommunications services.284 Those findings were not included in the Joint Statement as matters not contested by the parties, and yet none of the parties sought to address those findings in any substantive way. The relevant findings made by the ACCC were that there are high barriers to entry and expansion in the provision of mobile services. As a result of these barriers, the prospect of new entry is relatively low. The barriers to entry and expansion include the following:
(a) Large up-front sunk capital investment required: Mobile networks, and telecommunications networks in general, require large up-front capital investment, which is often sunk.
(b) Economies of scale: The provision of mobile services is characterised by significant economies of scale. Providing mobile services has high fixed costs, and low variable costs. This cost structure results in declining average unit costs as more services are provided. The significant economies of scale can deter new entry, where the entrant may struggle to obtain sufficient scale.
(c) Scarcity of spectrum: A key barrier to entry and expansion is access to spectrum. Spectrum is a scarce, critical input into mobile telecommunications, without which operators cannot offer any service at all. Further, operators require access to a sufficient quantum of spectrum, across a mix of spectrum bands. Spectrum suitable for use in mobile networks has been allocated in Australia over a period of three decades, with licences expiring at various times, and often re-issued to the incumbent users. A potential new entrant would be required to build a spectrum portfolio over time, incurring large costs without any corresponding revenue and at significant risk of not passing a threshold spectrum portfolio on which it may deploy a network.
(d) Brand: Branding and brand perception is also a significant barrier to entry and expansion in mobile telecommunications. Customer perceptions of network quality and reliability are a key driver of consumer decisions regarding mobile services, and these perceptions can be difficult to shift.
(e) Phase in technology cycle: Mobile telecommunications exhibits strong first-mover advantages, with changes in market share due to early advantages in technology lifecycles having the potential to endure through the lifecycle of a technology, such as 4G. The rollout of 5G is now in a critical phase in Australia. When a firm invests in network upgrades ahead of its rivals, it does so with some expectation that it will be able to improve its market share. In turn, this can provide incumbency advantages.
336 The Tribunal considers that the above findings are well-supported by the material that is before the Tribunal in this proceeding. Indeed, the applicants’ commercial rationale for the Proposed Transaction is based on a number of those findings.
Capital intensity and economies of scale
337 It is uncontroversial that mobile networks require large up-front capital investment, which is often sunk. Richard Feasey prepared an economic expert report for Telstra in relation to the Proposed Transaction. In that report, Mr Feasey observed:285
… mobile networks involve fixed and operating costs which represent a significant proportion of the total costs to be borne by the industry and its customers. Sharing these assets allows greater efficiencies or economies of scale to be realised, reducing average costs for those concerned. The benefits of lower costs can be particularly significant in less densely populated areas, where individual operators may otherwise find it uneconomic to provide network coverage, either at all or to the extent that is possible if the network is shared.
338 Australia’s geography is characterised by a very sparse population density on average, paired with a very urbanised population centred in the capital cities and surrounding major regional centres. Australia is one of the most urbanised countries in the world, with the Estimated Resident Population of the 8 capital cities alone totalling more than 17 million people.286 Mr White of Optus gave evidence that the largest 12 cities in Australia (the capital cities of Sydney, Melbourne, Brisbane, Perth, Adelaide and Canberra plus Gold Coast, Sunshine Coast, Newcastle, Central Coast, Wollongong and Geelong) constitute up to 72% population coverage, and the top 19 cities (adding Hobart, Darwin, Townsville, Toowoomba, Cairns, Ballarat and Bendigo) constitute up to 80% population coverage.287 Mobile networks therefore need only cover a very small proportion of the total landmass in order to provide mobile coverage to the homes and workplaces of a majority of the population.288 However, the implication of this degree of urbanisation for mobile coverage is that covering the remainder of the population becomes decreasingly economic as network coverage expands into more remote and less densely populated areas.289
339 In the ACCC’s final report for the 2017 domestic mobile roaming declaration inquiry, the ACCC stated:290
The low population density and large size of regional Australia mean that mobile network operators (MNOs) must incur significant costs to extend population coverage by a small amount in regional areas. This is because they must cover large areas to achieve only a small increase in population coverage. For instance, Telstra covers only 0.8 per cent more of the population than its competitors but its network is over one million square kilometres larger than the next largest network.
…
The ACCC considers that the difference between geographic coverage and population coverage is likely to impact the competitive dynamics between the MNOs and their investment incentives. In areas of high population density, such as metropolitan and large regional centres, there tends to be stronger infrastructure-based competition between the three MNOs. This is because the higher level of population density in these areas means there is likely to be more network usage over which the fixed costs of providing coverage can be recovered. This enables an MNO to reach minimum efficient scale – and therefore be profitable – at a lower share of mobile consumption made in these areas. However, in areas of low population density, particularly regional and remote areas, investment to extend coverage is less likely to be profitable for multiple network operators. There are fewer customers and to some degree, their needs for more contiguous coverage over greater areas are higher. This has meant that there tends to be less infrastructure-based competition in these regions, and consumers can generally only obtain reliable mobile services from either one, or in some areas, two network operators.
340 In its application for authorisation filed with the ACCC, the applicants described the commercial challenge of mobile infrastructure investment in Australia:291
63 Building and maintaining mobile network infrastructure is capital intensive and MNOs face an ongoing infrastructure investment challenge. Mobile networks involve capital investment and fixed operating costs which represent a significant proportion of the total costs to be borne by the industry and its customers.
64 As the ACCC has analysed, the high costs involved in expanding mobile network coverage and quality are correlated to Australia’s highly urbanised population, where revenues from the provision of mobile services to regional and rural customers diminish as population density decreases. These costs are further exacerbated by the need for MNOs to continually deploy new network technologies to market quickly, such as 5G, while earlier network investments become redundant (e.g., 3G). The high cost of expanding mobile infrastructure is a function of:
(a) the level of utilisation of mobile network infrastructure, which impacts the business case for infrastructure investment and the ongoing cost of mobile service provision to recover such investment …
(b) the nature of mobile service provision which requires MNOs to offer services and maintain network infrastructure across a wide coverage area that is inconsistent with customer utilisation of the network. This issue is acute in Australia which is characterised by high urban density and low to very low rural and regional density …
(c) increased consumption of data as newer generations of mobile technology support more data intensive apps and services consume more bandwidth, which means that MNOs face continuing investment demands after the initial deployment of new generations of mobile technology to address these capacity constraints.
65 The provision of mobile services requires ongoing investment, including investments to upgrade from one generation of mobile technology to the next, and within each generation at a localised cell level to continually expand capacity. This requires constant cycles of investment in new mobile infrastructure and technologies in parallel with existing network assets.
341 In his witness statement, Mr Penn, formerly Telstra’s CEO, stated:292
36 … Mobile network deployments, of any kind, involve high fixed costs and these costs are generally even higher in rural and regional areas. The demand for capital investment also increases as demand for capacity increases. 5G technology is designed to handle increased demand for mobile data by providing services using a higher frequency of spectrum. However, because of the propagation characteristics of mid-band and high-band spectrum (i.e. it doesn’t cover the same distance as low-band spectrum), we require more sites to achieve these higher data rates, which further increases the capital cost of our rollout.
37 The vast majority of the Australian population resides in a very small area of land in metropolitan areas. Around 17% of the population resides in a much larger regional and urban fringe areas to which the Proposed Transaction relates (17% Regional Coverage Zone).
38 Customers want mobile coverage as they travel across Australia, including travelling from urban areas to rural and regional areas. As a result, Telstra endeavours to maintain a high-quality mobile network infrastructure across a vast area, including where customer utilisation of the network is much lower in rural and regional areas. As the density of population decreases, the average cost per person of infrastructure used to service the population investment increases. This means that infrastructure investments in regional and rural areas are generally harder to make work economically than investment of an equivalent amount on infrastructure in more highly populated, metropolitan areas.
…
43 As CEO, I ultimately view the commercial importance of spectrum not only in technical or network terms, as much as in terms of what spectrum can deliver for customers and/or for Telstra’s business. In this regard, I see two key commercial benefits from spectrum:
(a) spectrum is critical to improving the improved service outcomes that we can deliver customers, including (in particular) in relation to mobile data speeds and coverage; and
(b) having sufficient spectrum, and using it efficiently, can avoid the need for Telstra to invest as much capital in physical infrastructure - such as mobile sites. This is particularly relevant in relation to regional areas (where the costs of sites can be very high) and 5G, where delivering higher data rates becomes reliant on a denser network than earlier mobile technologies.
342 In his witness statement, Mr Berroeta, CEO and Managing Director of TPG, stated that TPG’s limited regional coverage has historically been, and continues to be, a significant barrier to TPG acquiring and retaining those consumer and enterprise mobile customers who value regional coverage, as well as wholesale mobile customers who wish to offer mobile services as an MVNO to customers seeking coverage in regional areas.293 In his statement, Mr Berroeta provided a history of the commercial arrangements entered into by TPG (and its corporate predecessors) with Telstra and Optus to extend the geographical coverage of its network. It is clear from Mr Berroeta’s statement that he regards the expansion of TPG’s network into regional areas by infrastructure investment as uneconomic, stating:294
The Proposed Transaction aligns with TPG’s ambitions for regional Australia and will enable TPG to deliver regional coverage that rivals its competitors within a timeframe and in a manner not otherwise commercially viable for, or available to, TPG. …
343 In his witness statement, Mr Kanagaratnam explained that:295
A mobile network is capital intensive and has scale economics with ongoing technology upgrades driving ongoing need for capital. Scale economics is driven by two factors. First, if an operator has a larger customer base, they can generate more revenue and invest more capital. Secondly, if an operator has a larger customer base, the costs are lower on a cost per customer basis.
344 The scarcity of spectrum is discussed further below. At this point it is sufficient to refer to the opinion of Ms Ihaia who prepared an economic expert report for Telstra in relation to the Proposed Transaction.296 In that report, Ms Ihaia stated:
46. All three MNOs hold spectrum across a range of low and mid-range bands in regional and rural areas, as described in Table 4 of the Authorisation Application. As explained in paragraph 75 of the Authorisation Application, sub-1 GHz spectrum is particularly relevant to providing services in regional and rural areas because its signal carries further and can penetrate obstacles such as trees. In regional and rural Australia, Telstra holds 2 x 20 MHz in the 700 MHz band, and 2 x 25 MHz in the 850 MHz band. In the December 2021 auction for the 850 MHz and 900 MHz bands, Telstra’s acquisition of 850 MHz took it to the maximum amount of low-band spectrum that it could bid for within the limit that applied to that auction. The entirety of the 900 MHz band available for mobile services at the December 2021 auction was acquired by Optus.
47. Further sub-1 GHz spectrum is unlikely to be available in the near future. As I discuss in more detail in section 5.2, while the 600 MHz (currently used for DTV) could potentially be freed up at some point as a second “digital dividend,” if this does occur, the time taken for past processes indicates that the spectrum is unlikely to become available until at least 2030.
…
56. Telstra reports that network congestion (defined as 4G speeds that are less than [Confidential-Telstra [REDACTED] [REDACTED]] during specific hours within a 4-week period) is a material issue for many sites and customers accessing its network in the 17% Regional Coverage Zone and in the remoter Telstra unique areas beyond. Once speeds drop below these levels, customers’ experience of services such as high-definition video, may start to suffer degradation.
57. Telstra illustrates the extent of the issue in regional areas as compared with metro areas as follows:
a. in major cities around [Confidential-Telstra [REDACTED]] of Telstra’s 4G sites are congested with around [Confidential-Telstra [REDACTED]] of 4G active users currently impacted by congestion.
b. in regional and remote areas around [Confidential-Telstra [REDACTED]] of Telstra’s 4G sites are congested and around [Confidential-Telstra [REDACTED]] active users currently impacted by congestion.
…
63. RAN congestion can be addressed by either acquiring access to further spectrum or densifying sites, as described in paragraph 264 of the Authorisation Application.
64. Where additional spectrum is available, it can be used to increase capacity and address network congestion that is limiting connection speeds. Where additional spectrum is not available, additional sites can be built to reuse the existing spectrum. However, given the sparse population and wide areas in the 17% Regional Coverage Zone, in my view it is likely to be uneconomic and or impractical to invest in additional sites to fully address congestion.
…
80. I understand from the Authorisation Application (paragraph 28) that TPG’s existing regional coverage makes it difficult for TPG to compete for customers in regional areas or to capture significant numbers of metropolitan customers who travel to urban fringe and regional areas. … the low population density in the 17% Regional Coverage Zone means that it would be highly inefficient and costly for TPG to duplicate the networks of Telstra and Optus. [Confidential-TPG TPG estimates that the cost of building out its network to match Telstra’s coverage in regional areas would cost around [REDACTED] [REDACTED] and take at least [REDACTED] [REDACTED] (as described in paragraph 49 of the Authorisation Application).
345 It was common ground that perceptions around mobile network leadership are just as important as actual network quality for how consumers choose a new provider when deciding to change providers.297
346 Evidence relating to the significance of brand perception was given by Mr Ackland of Telstra and Mr Cooney of TPG who each held roles concerned with the sales and marketing to consumer segments within their respective businesses.
347 Mr Cooney of TPG gave evidence concerning TPG’s intended marketing strategy if the Proposed Transaction proceeds. In respect of the initial phase of marketing ([REDACTED] [REDACTED] [REDACTED] [REDACTED]), Mr Cooney explained:298
… I anticipate that TPG’s intention will be primarily to target customers in [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED]) [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ‘[REDACTED]’ [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
348 Mr Cooney stated that most of TPG’s customers who leave TPG choose to move to Telstra or Optus and do so because of [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. Based on an analysis of customer churn data and TPG’s customer exit surveys, Mr Cooney gave evidence that TPG loses customers to both Telstra and Optus due to [REDACTED] [REDACTED] [REDACTED] [REDACTED]. Again, in respect of the intended initial phase of marketing of the Proposed Transaction, Mr Cooney said:299
… TPG’s message to existing customers will be that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ‘[REDACTED]’ [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
349 Mr Cooney also gave evidence concerning TPG’s analysis of consumer preferences and brand perceptions. In his evidence, Mr Cooney identified [REDACTED] distinct consumer groups by reference to demographic factors such as age and location and who have differing preferences with respect to price, coverage, data requirements and customer support. Mr Cooney stated that the identified consumer groups are important to TPG’s marketing and brand strategy.300 Relevantly, Mr Cooney explained that Vodafone is TPG’s key brand which is marketed as having a global network and products to match, and [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. About [REDACTED] of TPG’s mobile customers have services under the Vodafone brand. In the intended initial phase of marketing of the Proposed Transaction, Mr Cooney said that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].301
350 Mr Cooney’s evidence supports the conclusion that customer perceptions of network quality (in particular, coverage) are a key driver of consumer decisions regarding mobile services, and these perceptions can be difficult to shift.
351 Mr Ackland of Telstra explained that it is particularly important to his role to understand the factors and drivers that influence a potential customer’s decision to choose mobile services. Market research commissioned by Telstra shows that the reasons that consumers acquire mobile services from Telstra, ranked in order of importance, are [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].302
352 As part of their usual business processes, each of Telstra and TPG conducted financial modelling of the likely effects of the Proposed Transaction on their respective businesses, including the likely changes to market shares as a result of TPG being able to offer consumers better mobile coverage in the RCZ. Mr Ackland was involved in this assessment for Telstra. Mr Ackland said that Telstra assessed that the likely impact of the Proposed Transaction (TPG having access to Telstra’s network in the RCZ) would be that TPG [REDACTED] “[REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]”.303 [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]).304 TPG’s assessment [REDACTED] [REDACTED]. The business case modelling undertaken to assess the Proposed Transaction [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] “[REDACTED] [REDACTED]” [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED] [REDACTED]), [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].305 In that regard, it can be observed that Telstra’s market share in the RCZ greatly exceeds Optus’s market share.
353 Ms Bayer Rosmarin gave evidence concerning the substantial investment made by Optus in its mobile network since 2015. That investment is discussed in more detail below. Ms Bayer Rosmarin said that, despite this investment, Optus’s market share had not improved and had remained largely static for the 10 years to December 2020.306 As noted earlier, an Optus internal paper refers to [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] “the difficulty of unrusting customers from Telstra given historic perceptions about service quality and coverage”.307
354 Overall, the Tribunal considers customer perceptions of network quality and reliability are a key driver of consumer decisions regarding mobile services, and the Tribunal also accepts that these perceptions can be slow to change. That conclusion is supported by the customer churn data set out above, which supports a conclusion of the “stickiness” of mobile service customers.
First-mover advantage in the technology cycle
355 In its reasons for determination, the ACCC expressed the view that leaders in the adoption of transformative new technologies can gain an advantage over competitors, with early adopters able to improve productivity and service delivery and ultimately gain market share as part of the competitive process.308 While the parties did not indicate their agreement with that statement, the Tribunal considers that it is strongly demonstrated in Australia’s mobile telecommunications markets.
356 As discussed further below, the current focus of competition between the MNOs is 5G technology. All three MNOs are in the process of rolling out their 5G networks.309 Telstra has a considerable lead in the deployment of 5G, with its 5G network covering more than 80% of the population.310 Telstra’s public statements (for example, concerning its T25 strategy) and internal papers (including, in particular, board papers) confirm that technology leadership over its competitors, particularly with respect to 5G, is a key element of its competitive strategy with respect to its mobile network.311
357 The terms of the MOCN Service Agreement also demonstrate that Telstra considers that being a first-mover in the technology cycle delivers a competitive advantage. Under that agreement, Telstra will not provide TPG with access to 5G-enabled sites until 6 months after Telstra has activated the sites for 5G. The six-month delay will apply on a site-by-site basis, with the effect that TPG’s access to 5G in the RCZ would be staggered. That limitation will also apply to any future technology generation that is supplied under the MOCN Service Agreement. In the application for authorisation, the applicants acknowledged that the purpose of that limitation “is to provide Telstra with a limited first-mover advantage in 5G and other technologies”.312
358 In its reasons for determination, the ACCC expressed the further view that first-mover advantages in the Australian mobiles sector have created lasting structural impacts. The ACCC considered that additions to market share during the rollout and the early operation of new network technology have tended to be retained by the first-mover throughout the lifecycle of that technology generation.313
359 The Tribunal considers that being a first-mover in the technology cycle delivers a competitive advantage, enabling the first-mover to gain market share from its competitors. It is also an observable fact of the mobile service market that, over a considerable period of time, Telstra has demonstrated an ability to retain its market share leadership while charging premium prices in comparison to its competitors. The Tribunal considers that that ability is most likely attributable to the perception and the reality of Telstra’s network quality in comparison to its competitors, combined with customer “stickiness” as evidenced in [REDACTED] [REDACTED] [REDACTED] [REDACTED].
The geographic coverage of the MNOs’ mobile networks
360 Consumers value mobile coverage in the areas in which they live, work and travel.314 While the degree to which consumers value coverage varies, coverage in regional and remote areas is valued not only by consumers who live and work in those areas, but also by metropolitan consumers. In many cases, consumers place value on remote coverage in areas they may only visit very occasionally or may consider visiting in the future.315 Thus, in the context of the Proposed Transaction, coverage is an important issue not only for those that live and work within the RCZ, but also those that live in metropolitan areas.316
361 While the extent to which the MNOs are willing to invest in coverage will depend upon each operator’s business model, the extent of geographic coverage is a key component in the attractiveness of mobile services. The importance of wide geographic coverage to competition in mobile services can be understood by the high expenditure of MNOs to provide mobile services in regional and rural Australia.317
362 In metropolitan areas, all three MNOs exert competitive pressure on each other and drive investments in infrastructure in these areas, including cell densification, technology upgrades and investments in spectrum and fibre, including backhaul, to serve these denser areas.318 In more remote areas, the MNOs make strategic investments in sites which may not be profitable in isolation.319
363 Operators roll out infrastructure to increasingly sparsely populated areas not only to capture market share in those areas, but to retain existing share and win new share in denser areas where coverage is already available.320
364 The MNOs frequently make coverage claims in the marketing of their mobile services and price their services on a national basis.321 Telstra in particular makes representations regarding its superior geographic coverage and the extent of Telstra’s network features heavily in its marketing material. Optus and TPG (through the Vodafone brand) both offer guarantees regarding their networks’ coverage, allowing customers to exit their contracts within a fixed period if they are not satisfied with the coverage available.322 The MNOs also make representations regarding their level of investment in coverage, particularly coverage in regional areas.323
365 Telstra maintains a significant lead in the number of mobile sites it has deployed nationwide and in regional areas.324 The MNOs’ overall site numbers is set out in the figure below:325
Total number of mobile sites by MNO
366 The above figure was prepared by the ACCC from data published in its Mobile Infrastructure Report 2022 (dated September 2022).326 That report relevantly stated:327
As at 31 January 2022, Telstra had the most mobile sites (11,002) followed by Optus (8,632) and TPG (5,728) …
… in Major Cities, Optus had the most sites (5,294) in these areas, overtaking Telstra (5,257) for the first time since at least 2018. The number of sites each MNO had in Major Cities was relatively comparable in 2022. However, the gap between the MNO with the greatest and fewest number of sites in Major Cities has widened overtime to 963 sites in 2022. …
Outside Major Cities, Telstra had significantly more sites than the other MNOs between 2018 and 2022 (table 4.1). In these areas in 2022, Telstra had 72% more sites than Optus, down from 77% in 2018, and 313% more sites than TPG, up from 286% in 2018.
367 The different number of sites deployed by each of the MNOs in more regional and remote areas is reflected in the different population coverage claimed by each. In remote Australia for example, Telstra has 708 sites in total (Optus has 241 and TPG has 62), and in very remote Australia, Telstra has 898 (Optus has 158 and TPG has 8).328
368 As stated earlier in these reasons, in respect of the RCZ, Telstra has approximately 3,700 mobile sites, Optus has approximately 2,500 mobile sites, and TPG has 749 sites.
369 In its reasons for determination, the ACCC produced a number of maps depicting the geographic coverage of the MNOs’ mobile networks as at January 2022. The ACCC explained that the maps had been prepared using national maps provided by the MNOs under the ACCC’s Audit of Telecommunications Infrastructure Assets – Record Keeping Rules (Infrastructure RKR). The ACCC stated in respect of the coverage maps produced in its reasons that the maps were illustrative only and noted two limitations inherent in the maps:329
(a) the coverage maps are predictive, relying on computational models to predict where coverage will be available based on site locations and propagation characteristics of both the spectrum deployed and the terrain (hills, buildings etc.) in the coverage area; and
(b) each of the MNOs use differing assumptions regarding signal propagation from the mobile site and the end-user device receiving it which means that the maps are not directly comparable to one another on a strict coverage basis.
370 The parties do not contest the information conveyed by the maps, subject to the above limitations. The following maps depict the extent of each of the MNO’s 4G networks, as provided under the Infrastructure RKR.330
371 Telstra’s network covers 99.5% of the population (3G) and 99.4% of the population (4G). It also has the most extensive 5G rollout to date (around 80% of the population). Telstra claims to cover 2.6 million km2 – “1 million square kilometres more than any other mobile network”. The extent of Telstra’s 4G coverage is illustrated in the following figure.331
Telstra’s 4G coverage (blue)
372 Optus coverage footprint is approximately 1.3 million km2, reaching around 98.5% of the Australian population.332 The figure below illustrates the extent of Optus’s 4G coverage. Optus has prepared this map in relation to an “external antenna” standard of coverage, which likely overstates the coverage available with a standard mobile handset.333
Optus’ 4G coverage (“external antenna” coverage standard, green)
373 TPG has the smallest coverage footprint of around 0.6 million square kilometres, reaching around 96% of the Australian population.334 The figure below illustrates the extent of TPG’s 4G network coverage. The vast majority of TPG’s 5,728 sites nationwide are co-located with one or more MNOs.335
TPG’s 4G coverage (red)
374 The figure below overlays the coverage of the three MNOs, and illustrates the extent of infrastructure-based competition as it stands today. The three maps are not directly comparable due to different underlying assumptions, but the comparison illustrates the “good, better, best” competitive dynamic of the three networks.336
Current 4G coverage of the MNOs (without the Proposed Transaction, Telstra in blue, Optus in green, TPG in red)
375 Telstra’s mobile network provides the widest geographic coverage, with 99.4% population coverage on its 4G network.337 Telstra has stated publicly that maintaining network leadership is critical to its growth strategy leading up to FY25 and that maintaining and extending network leadership will underpin its market position and maintain its price premium. Due to uniform national pricing, this price premium covers both customers in regional areas, as well as in metropolitan areas.338
376 In its reasons for determination, the ACCC expressed the view that the extent of Telstra’s mobile network provides an enduring competitive advantage in downstream markets and is a strong contributor to its high market shares, both in metropolitan areas and in regional areas.339 The Tribunal agrees with that view.
377 Customers who live outside the RCZ frequently travel to areas within it. ACCC analysis of Telstra and TPG data (which is not contested by the parties) suggests that, on a weekly basis over the first 5 months of 2022, generally around [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] from outside the RCZ were used inside the RCZ. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].340
378 Mr Cooney of TPG gave evidence that he expected that the increased network coverage that would be provided by the Proposed Transaction would enable TPG to [REDACTED] reduce its churn rate (being the rate of loss by TPG of customers to its competitors). Mr Cooney prepared a spreadsheet that tabulated the reasons given by consumers for leaving TPG and acquiring a service from Telstra or Optus. The spreadsheet was based on TPG’s port out data and customer exit surveys. It contained monthly data from September 2021 to September 2022. The spreadsheet shows that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].341
379 The ACCC conducted analysis of port out survey data provided by TPG which indicates that around [REDACTED] of TPG post-paid respondents cited coverage as the main reason for leaving between January to June 2022, and around [REDACTED] of those respondents moved to Telstra, the MNO with the largest coverage.342
The spectrum capacity held by each MNO
380 MNOs need access to spectrum in order to provide a mobile service. The amount of spectrum and range of spectrum bands held affect the reliability, reach, speed and technologies (such as 5G) of mobile services delivered.343 The characteristics of different spectrum bands is discussed earlier in these reasons.
381 The three national MNOs are the largest commercial users of spectrum licences in Australia. The MNOs, along with NBN Co, hold the vast majority of licensed spectrum across 3 low (sub-1 GHz) bands and 5 mid (1-6 GHz) bands:
(a) low-band: 700 MHz, 800 MHz (including 850 MHz expansion) and 900 MHz;
(b) mid-band: 1800 MHz, 2 GHz, 2.3 GHz, 2.5 GHz and 3.4 GHz.344
382 The MNOs also own high-band (known as “millimetre wave”) spectrum licences in the 26 GHz band. This band has not seen wide deployment. No high-band spectrum is included in the Proposed Transaction.345
383 The table below shows the relevant spectrum holdings of the MNOs in regional Australia. Licence areas vary between bands, but largely reflect the situation in the RCZ:346
Indicative MNO spectrum holdings in regional Australia
384 NBN Co also holds large spectrum licences in the 2.3 GHz and 3.4 GHz bands in regional areas. NBN Co uses this spectrum for delivery of fixed wireless broadband.347
385 The ACMA conducted an auction of spectrum licences in the 850/900 MHz bands in late 2021. The ACCC recommended allocation limits on the auction, such that no person would be permitted to use more than 2 x 40 MHz of sub-1 GHz spectrum (ie low-band spectrum) in regional areas. The Minister made a decision to increase this limit in regional areas to 2 x 45 MHz. Telstra subsequently acquired up to the limit at auction.348 At the auction, Telstra acquired 2 x 10 MHz of spectrum in the 850 MHz band for a price of $615,660,000, and Optus acquired 2 x 25 MHz of spectrum in the 900 MHz band for a price of $1,475,958,000.349
386 As set out earlier in these reasons, TPG’s spectrum that is the subject of the Spectrum Authorisation Agreement is as follows:
(a) in respect of the 700 MHz band, 2 x 10 MHz;
(b) in respect of the 850 MHz band, 2 x 5 MHz;
(c) in respect of the 2100 MHz band, 2 x 5 MHz; and
(d) in respect of the 3600 MHz band, 20-45 MHz.
387 Thus, under the Spectrum Authorisation Agreement, Telstra will have immediate access to 2 x 50 MHz of low-band spectrum (and from 2024, 2 x 60 MHz) which can be used for Telstra’s and TPG’s mobile networks in the RCZ as part of the MOCN, and Telstra will also have exclusive access to that spectrum in areas beyond the RCZ. It can be observed that the low-band spectrum capacity in the RCZ that will become available to Telstra will immediately increase by more than 40%350 (from 2 x 35 MHz to 2 x 50 MHz, and will increase further to 2 x 60 MHz from 2024), but the additional TPG SIOs served by the MOCN represent less than [REDACTED] of Telstra’s SIOs (see the market share data set out earlier in these reasons). Thus, under the Proposed Transaction, the increase in low-band spectrum outstrips the increase in SIOs served by a factor of [REDACTED] [REDACTED] [REDACTED].
388 Currently, the MNOs are operating three generations of mobile technology across eight spectrum bands. The following table sets out the primary uses of these bands by each of the MNOs:351
MNO indicative spectrum use by technology type
389 It needs to be emphasised that the above table is indicative only. As noted earlier in these reasons, while the MNOs are making heavy use of the 3.4 GHz band for their 5G networks, there is also some use of low-band spectrum for 5G networks in the 700 MHz (TPG), 800 MHz (Telstra) and 900 MHz (Optus) bands.
390 Currently, Telstra makes heavy use of its 700 MHz holdings for a broad coverage layer on its 4G network, nationwide. This is expected to continue for the foreseeable future. Telstra’s 800 MHz spectrum is currently deployed in service of its 3G network. However, Telstra is progressively re-farming its 800 MHz spectrum in some areas away from 3G and onto its 5G network. Telstra will close its 3G network in 2024.352
391 Deploying low-band spectrum for 5G will enable the MNOs to maintain their existing coverage footprints, but using a more efficient technology. Low-band spectrum used for 5G is unlikely to enable a step change in user experience, but will deliver the broad coverage required in regional and remote areas.353
392 In mid-band, Telstra uses its 2 GHz holdings as additional capacity for its 4G network. 354
393 Telstra’s 5G network to date has largely been deployed on its 3.4 GHz spectrum. This is broadly in line with international practice, with spectrum in the wider 3.4-4.2 GHz band referred to as the ‘pioneer band’ for 5G. Both Optus and TPG also have deployed the majority of their 5G sites on 3.4 GHz spectrum to date.355
394 The ACMA is currently in the process of allocating spectrum licences in the 3.4 and 3.7 GHz bands. The ACCC has provided advice to the ACMA that this spectrum is a close substitute for existing spectrum licences between 3400 MHz and 3700 MHz (the 3.4 GHz band).356 This process will make a large amount of spectrum available in regional areas which may be used by the MNOs to deploy 5G services, both for mobile and fixed wireless. Following these allocations, all spectrum between 3400 and 3750/3800 MHz in regional areas will be spectrum licensed and allocated.357 Following this auction, the ACMA intends to make up to 200 MHz of spectrum in the 3.8-4.0 GHz band available as area-wide apparatus licences.358
395 After the wider 3.4-4.0 GHz band, the ACMA is unlikely to bring any large amounts of new mobile-optimised spectrum to market. The ACMA’s forward-looking workplan (the “Five-year Spectrum Outlook”) does not include any spectrum earmarked for spectrum licensing and allocation for mobile services in the near future.359
396 The Australian Government released a green paper in 2021 discussing the possibility of a “second digital dividend” via a reallocation of spectrum in the 600 MHz band away from free-to-air television broadcasting and towards wireless broadband. The ACMA has categorised that band as being in the “monitoring” stage of replanning, the first stage of four.360
397 Licences in 7 spectrum bands used by the MNOs expire between 2028 and 2032. This includes all four bands (700 MHz, 800 MHz, 2 GHz, 3.4 GHz) relevant to the Proposed Transaction. This is within the initial ten-year period set out in the agreements. The ACMA intends to begin the process of consulting on these licence bands in the next year.361
398 On 25 July 2022, the ACMA made a written submission to the ACCC in respect of the application for authorisation in response to a request from the ACCC seeking information on specific matters.362 One of those questions was whether TPG has spectrum licences that will expire during the term of the Proposed Transaction and whether the ACMA is likely to renew these licences. In response to that question, the ACMA advised that TPG has a number of extant spectrum licences that will expire in the initial term of the Proposed Transaction, including its spectrum licences in the 700 MHz, 850 MHz, 1800 MHz, 2 GHz, 2.5 GHz, and 3.4 GHz bands. The ACMA also summarised the statutory provisions under the Radiocommunications Act that provide a procedure for the renewal of licences which was introduced by the Radiocommunications Legislation Amendment (Reform and Modernisation) Act 2020 (Cth). The procedure enables licence holders to apply for renewal. The ACMA stated that it has not yet had the occasion to act under this procedure. However, it has published an information paper that sets out its current policy for, among other things, acting under the procedure.363 The document states, amongst other things, that ACMA’s policy is:364
… to engage [spectrum] licensees early and enable full consideration of relevant matters, we expect to undertake work in relation to the renewal process, prior to the renewal application period.
…
In all cases, we would expect that holders of long- and medium-term licences would apply for renewal at the beginning of the renewal application period. This would enable us to decide on renewal well ahead of expiry, providing clarity to licensees over their future arrangements.
Network capacity of each MNO in the RCZ
399 As stated earlier in these reasons, the capacity of a mobile network is a function of mobile site density, spectrum deployments and radio technology. As MNOs densify their networks, their networks are able to serve more traffic. This enables more customers to use the network, or enables the MNO to provide greater speeds or other capabilities to end-users.
400 Mobile service providers compete to offer the fastest data speeds over their networks (typically download speeds). This factor has become increasingly important with more plans having significant data inclusions and is an important factor driving the rollout of the MNOs’ 5G networks.365 Mobile service providers advertise speeds available in regional areas. For example, Telstra advertises “faster speeds in more places”, and often highlights the coverage of its 5G network, which many consumers are likely to associate with faster mobile data speeds.366
401 Before the ACCC and on this review, there was significant disagreement between the parties with respect to the measurement of relative network capacity in the RCZ and the likely effect of the Proposed Conduct and the Proposed Transaction on relative network capacity. The disagreement is at two levels: the first relates to the appropriate methodology to measure relative capacity; and the second relates to the underlying data in respect of each MNO. Evidence as to network capacity was principally given on behalf of Telstra by Aetha and was principally given on behalf of Optus by Mr Turner who is the Director of Spectrum Strategy and Management at Optus. In that respect, the Tribunal notes that Mr Turner has a Bachelor of Engineering in Electronics, Communications and Computer Engineering from the University of Bradford, England, and has worked in the telecommunications industry since 1995.
Methodologies to measure network capacity
402 Mr Kanagaratnam of Optus gave evidence that there are several different metrics to measure network capacity including:
(a) absolute spectrum holdings, being all of the spectrum holdings, channels and bandwidths that an MNO has;
(b) peak data rates, which is a typical measure of maximum network capacity and speed provided by a mobile network; and
(c) spectrum bandwidth per population, which measures how much spectrum is available on a per person basis, with reference to the number of people within a specific population (the relevant formula being the product of available spectrum and the number of mobile sites in an area divided by the total population count in that area).367
403 The Tribunal accepts that each of the above metrics informs an analysis of relative network capacity.
404 In relation to absolute spectrum holdings, Mr Turner of Optus gave evidence that overall spectrum holdings are an important measure of an MNO's ability to provide service quality because:
(a) the greater the amount of spectrum held, the greater the capacity and speed achievable by each mobile site in the area where that spectrum is licensed for use; and
(b) spectrum is a finite resource, and so, unlike mobile sites, which (subject to capital expenditure constraints) can be built or upgraded at any time, there are fixed limits on how much spectrum can be accessed by any one MNO at a particular time.368
405 In relation to peak data rates, Mr Turner explained that the peak sector data rate is calculated by multiplying the spectrum bandwidth available to an MNO (MHz) at a site, by the peak spectral efficiency (bps/Hz) for each spectrum band available, and aggregating that across all the spectrum bands to the specific sector per site level. The result shows the peak data rate of a single sector on the base station.369 A sector is defined by the area covered by a single antenna affixed to a mobile site (typically a mobile site has three antennas oriented in different directions to serve the site).370 The peak data rate for multiple sectors can then be aggregated to show the peak data rate for an entire base station or region on that network.371 The higher the sector peak data rate, the greater the network capacity and therefore the speeds achievable for customers in the area served by that sector.372
406 The third metric measures an MNO’s network capacity as the product of its spectrum bandwidth and its site numbers, and which then takes account of the population that uses or may use that capacity. While that approach to the measurement of network capacity is generally supported by all lay and expert witnesses who gave evidence on this topic,373 there remains disagreement with respect to population count – specifically, whether population count should be based on all potential mobile service users in a given area, or whether population count should be based on the MNO’s customer base (SIOs) in a given area. The Tribunal considers that both metrics are relevant. The former measure, which divides an MNO’s absolute network capacity by all potential mobile service users in given area, is relevant to assessing the relative network capacity of MNOs and their ability to compete for all customers in a given area. Of course, as the population denominator would then be common to each MNO, a measurement of relative network capacity reduces to the product of its spectrum bandwidth and its site numbers. The latter measure, which divides an MNO’s absolute network capacity by its customer numbers in a given area, recognises that an MNO with a large customer base is likely to need greater network capacity to serve those customers.
407 The Tribunal recognises that other technical matters are relevant to the assessment of relative network capacity. In particular, in respect of pooled spectrum, the adjacency of the pooled spectrum affects the potential capacity and speed able to be achieved from the spectrum.374
Measurement of network capacity in the RCZ
408 There was little disagreement between the parties concerning the measurement of absolute low-band spectrum holdings in the RCZ. Data presented earlier in these reasons shows that, in respect of low-band spectrum, Optus holds 2 x 35 MHz (including the recently allocated 900 MHz spectrum which is available post-June 2024), Telstra holds 2 x 45 MHz (including the recently allocated 850 MHz spectrum which is available post-June 2024) and the MOCN would hold 2 x 60 MHz.375
409 The figures presented by each of the parties in respect of mid-band spectrum were not consistent. Mr Turner’s assessment is that, in respect of mid-band spectrum, Optus holds 150 MHz, Telstra holds 280 MHz and the MOCN would hold 350 MHz.376 Aetha’s calculation, based on a site-weighted average of available spectrum, reduced the spectrum available under the MOCN by approximately 20 MHz (and also reduced the available spectrum for a number of other reasons as discussed in its report).377 Mr Turner contested Aetha’s site-weighted average methodology.378
410 The competing contentions of the parties with respect to absolute spectrum holdings in the RCZ is shown in the following table which is an extract of a table produced by Aetha in its supplementary report:379
Source | Operator | Low-band | Mid-band | Total |
Turner | Telstra plus MOCN pooling | 120 | 350 | 470 |
Aetha | MOCN | 120 | 271.3 | 391.3 |
Turner | Optus | 70 | 150 | 220 |
Aetha | Optus | 70 | 134 | 204 |
411 While there are material differences in calculation between the parties in respect of mid-band spectrum, the overall conclusion remains that Telstra has materially greater absolute spectrum holdings in the RCZ in comparison to Optus and the Spectrum Authorisation Agreement will increase its holdings. In low-band spectrum, the increase will be 30 MHz. In mid-band spectrum, the increase will be 30-55 MHz. In respect of the latter, Mr Turner gave evidence that Telstra’s access to additional 3.5 GHz spectrum under the Spectrum Authorisation Agreement provides an advantage in respect of regional 5G rollout because the superior bandwidths will mean that Telstra will enjoy a very significant speed advantage on 5G in this band.380
412 The Aetha report, filed on behalf of Telstra, focussed on relative network capacity measured as the product of the MNO’s available spectrum and the number of mobile sites in an area divided by the MNO’s services in operation (SIO) in the area. Although Aetha’s analysis showed that Telstra has, and the MOCN would have, materially greater spectrum holdings than Optus in each of the sub 1 GHz, 3 GHz and 6 GHz bands (broadly by a factor of 1.5 for Telstra and a factor of 2 for the MOCN),381 and that Telstra has, and the MOCN would have, a greater number of sites in the RCZ than Optus,382 Aetha concluded that Telstra has, and the MOCN would have, lower network capacity per SIO than Optus.383 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED], [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]).384
413 The calculations in Aetha’s report were contested by Mr Turner. Indeed, the parties were in dispute with respect to almost every integer in the calculations. In its supplementary report, Aetha sought to summarise the differing integers used by it and Mr Turner. This is shown in the following table which is an extract of a table produced by Aetha in its supplementary report:385
Operator | Spectrum (MHz) | Sites | SIOs | Spectrum x sites / SIO | |
Optus calculation | MOCN | 470 | 3921 | 4,031,209 | 457 |
Optus | 220 | 2274 | 1,491,018 | 336 | |
Aetha calculation | MOCN | 391.3 | [REDACTED] | [REDACTED] | 293 |
Optus | 204 | 2743 | 1,403,754 | 399 |
414 It can be seen that materially different integers have been used in the calculations. The differences are explained in the Aetha supplementary report and Mr Turner’s statement. It is not possible for the Tribunal, in the context of this review of a merger authorisation, to reach any firm conclusion about the correct or preferable approach. That is a consequence of the procedures that are applicable to the review of a merger authorisation. Whichever integers are used, it is apparent that the absolute capacity of Telstra’s mobile network in the RCZ, measured as the product of spectrum and sites, is considerably larger than that of Optus and will be further enhanced by the Spectrum Authorisation Agreement. However, the Tribunal accepts that an assessment of network capacity must also have regard to the number of customers served by a given network.
415 Mr Turner presented the result of calculations made of the downlink and uplink sector peak data rates in the RCZ able to be achieved by Optus, TPG, Telstra and the MOCN when using low-band spectrum and mid-band spectrum respectively. The results of Mr Turner’s calculations were that Telstra has the potential to increase its:
(a) downlink peak sector data rates on low and mid band spectrum combined in the regional area from 1.8 times that of Optus's currently, to 2.4 times that of Optus; and
(b) uplink sector peak data rates on low and mid band spectrum combined in the regional area from 1.6 times Optus's currently, to 2 times that of Optus.386
416 Mr Turner expressed the following opinions about relative network capacity resulting from his analysis of spectrum:387
74 As a result of such substantial spectrum holdings, especially when those holdings are contiguous and in the absence of further 'native' 5G bands (i.e. 3.5GHz, which was structured for 5G compatibility), no competitor will be able to provide equivalent 5G capacity (due to total available bandwidth) or speed (due to total contiguous bandwidth). This will substantially impact the ability of other MNOs to compete with Telstra because capacity and speed are almost linear functions of the spectrum bandwidth (they improve as the bandwidth increases).
75. Assuming full deployment of spectrum holdings, a competing MNO would, in my view be unable to meaningfully compete with a TPG-Telstra pooled network on capacity and speed due to the sheer imbalance in the holdings. In the absence of sufficient spectrum holdings (which as I have described is a finite resource), the only way to compete would be to build additional infrastructure which is time consuming, costly, and based on my experience and understanding, difficult for Optus to justify in regional areas where Telstra is already at a significant advantage.
417 For a given level of network quality at a given point in time, retail mobile service providers compete on the price and inclusions of their service offerings. This includes among their flagship and sub-brands, as well as MVNOs.388
418 Historically, retail services have included some form of access charge as well as some form of usage charge, but the majority of plans today include unlimited calls and SMS and a fixed data inclusion, for a fixed price. Further, some mobile operators compete on the basis of content inclusions (such as live sport or access to streaming services) associated with their offerings.389
419 Average advertised prices for retail services have risen in recent years, with providers generally choosing to include “more for more” in their retail bundles.390
420 The retail brands of the MNOs, along with MVNOs, also compete on feature inclusions. The cost of calls and mobile data per unit has fallen significantly over the longer term, and the vast majority of plans available on the market today include unlimited national and mobile calls and texts.391
421 Data inclusions also continue to grow strongly. Over the period 2016-17 to 2020-21, feature adjusted prices for mobile phone services declined by over 50%. Providers now advertise primarily on price and data inclusion, eg 40 gigabytes for $40 per month.392 The flagship brands of all the MNOs now also offer no additional charges on excess data usage on higher priced plans, a feature which may also be called “endless” or “infinite” data.393
422 There was a degree of inconsistency in the evidence concerning the relative prices charged by Telstra, Optus and TPG for their retail mobile services. Optus made submissions to the ACCC to the effect that Optus sets its prices at a discount to Telstra, and TPG’s prices are lower than Optus’s prices. That evidence is consistent with data presented by Dr Padilla, which graphed the differences between the post-paid SIM-only prices of Telstra, Optus and TPG’s flagship Vodafone brand, and which showed that the difference in prices increases as the data bundle increased.394 However, Mr Ackland of Telstra gave evidence that, while TPG (and its MVNOs) have less coverage than Optus and Telstra (and their MVNOs), the plans of TPG (and its MVNOs) are offered at similar price points.395 The ACCC found, and it was not contested, that Telstra, Optus and TPG tend to charge differentiated prices; for SIM-only plans, the cheapest plan available from each of the MNOs is $58 per month from Telstra (for 40 gigabytes), $49 per month from Optus (for 30 gigabytes), and $40 per month from Vodafone (for 10 gigabytes, but marketed as including greater data inclusions than stated in its Critical Information Summary).396 The MNOs’ ARPU also reflects the extent of their differentiated pricing. Telstra’s post-paid ARPU for the half-year ended June 2022 was $48.74, Optus’s for the half-year ending March 2022 was $39, and TPG’s for the half-year ending June 2022 was $42.397
423 Whilst in a strict sense ARPU is a measure of revenue and not price, it is a reasonable proxy for relative prices because of the manner in which the MNOs now price their mobile services (generally with unlimited call and text messages and a fixed volume of data inclusions). It is notable that Telstra’s post-paid ARPU for the half-year ended June 2022 was substantially higher than Optus and TPG. The latter were closer in their relative level, although notably TPG was higher than Optus.
The technology cycle – competition to provide 5G services
424 As discussed earlier in these reasons, leaders in the adoption of transformative new technologies like 5G can gain an advantage over competitors, with early adopters able to improve productivity and service delivery and ultimately gain market share.
425 The availability of 5G technology is an increasingly critical focus of competition in the supply of mobile services.398 5G is the newest mobile technology to be deployed, and operators and vendors claim it represents a step-change in the capability of mobile networks. The wide deployment of 5G will enable enhanced mobile broadband services, as well as other capabilities such as reliable low-latency network connections and mass machine communications including NBIoT uses.399 Mr Meissner of Telstra summarised the benefits of 5G technology as follows:
(a) faster speeds: while 5G speeds will vary by area, 5G will be up to 10 times faster than 4G with speeds of 10Gbps or more potentially achievable in future;
(b) lower latency: in the single digit milliseconds compared to several tens of milliseconds for 4G, which will enable 5G to support more advanced cloud and data applications; and
(c) more bandwidth: capacity to support up to 100 times more devices within a cell area than 4G, which will enable 5G to support a high density of NBIoT applications.400
426 5G also enables the deployment of fixed wireless broadband services on a greater scale due to its more efficient use of spectrum. All three MNOs now offer some form of 5G fixed wireless product. 5G fixed wireless has the potential to allow the vertically integrated MNOs to bypass use of the NBN wholesale network in order to serve retail fixed broadband customers.401 5G networks enable the provision of wireless fixed broadband on terms that are competitive with existing fixed line networks and offer an end-user experience that is comparable in terms of speed and data allowances to those fixed line offerings.402
427 All three MNOs are competing in the supply of retail mobile services on the basis of 5G availability, advertising their 5G coverage, faster 5G speeds, or new capabilities enabled by 5G. The provision of 5G is also a basis on which MNOs compete to acquire wholesale customers. The availability of newer product features, such as 5G, to MVNOs is often delayed until after their introduction on the flagship retail brands of the MNOs.403
428 Deploying 5G infrastructure allows MNOs to offer retail and wholesale mobile services that make use of greater capacity and speed, and offer new and differentiated services in the future. Where providers compete on speed, network reliability and the availability of 5G, an advantage in the underlying infrastructure allows an MNO to win market share from its rivals.404
429 Telstra has a considerable lead in the deployment of 5G, with its 5G network covering more than 80% of where the population resides. Telstra has publicly announced that in the 5 years to June 2020 it had invested $7.5 billion in its mobile network nationally, with “much of this on 5G”.405
430 The following figure shows the nationwide number of 5G mobile sites by MNO:406
Total number of 5G sites by MNO, 2020 to 2022
431 The above figure was prepared by the ACCC from data published in its Mobile Infrastructure Report 2022 (dated September 2022).407 That report relevantly stated:408
As at 31 January 2022, Telstra had significantly more 5G sites (4,071) compared to the other MNOs across all areas with more than twice as many sites as Optus (1,932).
432 The Report also records that, as at that date:
(a) in the Major Cities, Telstra had 3,140 sites, Optus had 1,790 sites and TPG had 991 sites;
(b) in Inner Regional Australia, Telstra had 684 sites, Optus had 122 sites and TPF had 38 sites;
(c) in the Outer Regional Australia and remote Australia, Telstra had 245 sites and Optus had 20 sites.409
433 As discussed below, both Optus’s and TPG’s 5G rollouts have been delayed by the government’s imposition of 5G TSSR guidance. The impact of this guidance has been to increase the cost of rolling out 5G for both Optus and TPG, and delay their deployment relative to Telstra, which has been largely unaffected.410
434 Currently, the focal point of competition in the mobiles market is around 5G.411 Consumers are increasingly upgrading their handsets to models that support 5G, with 5G ecosystem support developing around all major device manufacturers.412
435 In the short term, 5G is unlikely to represent a step-change in user experience for end-users. For mobile users, while the average speeds available on Australia’s 5G networks are greater than on 4G, users are likely to be constrained in their usage by data caps, and use-cases that cannot be catered for on existing 4G networks appear to be rare.413 However, perceptions of network quality and capability are crucial in driving consumer decisions, including operator representations about what 5G will enable now and into the future.414 In addition, the MNOs are collectively investing billions of dollars into upgrading their networks to 5G, indicating a widespread belief that 5G is critical to their future ability to compete, or to offer future services that consumers are likely to demand, or both. 5G is also likely to offer significant cost savings over time to the MNOs, when managing increasing traffic demand from users.415
436 Telstra intends to achieve 95% population coverage on its 5G network by 2025 and, as noted earlier, its current strategy is to lead in 5G technology and to maintain network leadership. Prior to the announcement of the Proposed Transaction, Optus’s plan was to achieve “[REDACTED] [REDACTED] [REDACTED] [REDACTED]” [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. This means that while investments in 5G availability and coverage are an important investment for operators over the longer term, the competitive effects of the rollout in terms of customer churn and competitive advantage are happening at the present moment.416 An internal Optus analysis of its 5G upgrade dated 25 May 2022, prepared in response to the announcement of the Proposed Transaction, showed that, under Optus’s current 5G roll out:
(a) [REDACTED];
(b) [REDACTED].417
437 The material before the Tribunal includes a large number of internal papers of each of Telstra, Optus and TPG, including board papers. Those papers demonstrate that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. For example, a Telstra board paper for a meeting on 10-12 February 2020 authored by Mr Katinakis discussed in detail the [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] “[REDACTED] [REDACTED] [REDACTED] [REDACTED]” [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]: “[REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]”.418 The Optus Board paper seeking approval for its proposed 5G Business Case in July 2020 summarised the importance of 5G competition as follows:419
• Our ambition is to become Australia's most loved everyday brand with lasting customer relationships [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED]
• Australian consumers are already adopting 5G technology and all operators are rolling out 5G to take advantage of this switching event
• Telstra are accelerating and promoting their rollout of 5G and are exploiting the natural advantage created by the Huawei 5G ban to re-establish their network leadership position
• [REDACTED]
• [REDACTED]
438 However, an important paper presented to the Optus Advisory Committee on 29 October 2021 commented on the competitive challenges faced by Optus with respect to the 5G roll out. The executive summary to the paper stated as follows (emphasis in original):420
• The Australian market remains highly competitive and fragmented, with 5G a key competitive battleground
• [REDACTED]
• [REDACTED]
• [REDACTED]
• Telstra continue to heavily promote 5G in their above-the-line marketing and are leveraging their claim of "Best 5G Network"
• [REDACTED]
• [REDACTED]:
• [REDACTED]
• [REDACTED]
• [REDACTED]
Commercial and regulatory impediments to expanding network coverage and capacity in the RCZ
Overview
439 Since the inception of mobile technology in Australia, regional and rural investment has been considered by MNOs to be a challenge, and often not commercially viable.421 Due to lower expected returns on network investment in regional and remote areas versus metropolitan areas, the commercial incentives to deploy network infrastructure in these areas are typically lower than in metropolitan areas.422
440 Mr Sweers gave evidence of internal analysis conducted by Telstra to calculate an “average cost per person captured by Telstra sites in metropolitan regional and remote areas”.423 The calculation was based on two formulae. The first calculated the captured population per site by dividing the population in a given area (derived from data published by the Australian Bureau of Statistics) by the number of Telstra mobile sites in the area. The second calculated the cost per person captured by dividing the indicative costs of establishing a Telstra mobile site in metro, regional and remote areas by the captured population per site. The calculations are summarised in the following table which formed Table 1 in the application for authorisation:
Metro areas (81.4% population) | RCZ | Remote (0.7% population) | |
Site count | 6,200 | 3,700 | 900 |
Population | 20,919,800 | 4,471,800 | 179,900 |
Indicative cost per site | [REDACTED] | [REDACTED] | [REDACTED] |
Captured population per site | 3,374 | 1,209 | 200 |
Cost per person captured | [REDACTED] | [REDACTED] | [REDACTED] |
441 The 2021 Regional Telecommunications Review found that there are still connectivity shortfalls in regional, rural and remote Australia, and that while mobile coverage continues to improve, expanding reliable coverage to “priority areas” such as major transport corridors, disaster-prone communities, tourist areas, and public facilities is becoming more difficult. Similarly, Infrastructure Australia has identified 24 of Australia's 48 regions as having an “Infrastructure Gap” regarding broadband and mobile connectivity.424
442 The history of regional mobile network investment suggests that the two primary ways in which MNOs have sought to make regional investment commercially viable is by obtaining government assistance and by entering into agreements with their competitors to form joint ventures or share network infrastructure.425 In many areas, it is unlikely that operators would roll out coverage without government co-contributions, such as from the Mobile Black Spot Program or other State or Territory programs.426
443 The evidence before the Tribunal indicates that Telstra faces lower impediments to expanding regional coverage in comparison to Optus and TPG. The evidence demonstrates that Telstra has the broadest 4G and 5G network coverage measured by population and landmass. There is no suggestion that Telstra considers that the Proposed Transaction is necessary to make its ongoing 5G rollout commercially viable. The Tribunal considers that Telstra’s rollout is supported by its considerable scale advantages.
444 In contrast, the evidence indicates that each of Optus and TPG face significant impediments to expanding regional coverage. Those impediments arise from their relative scale (in comparison to Telstra) and government regulatory requirements arising from the TSSR guidance. Historically, Optus and TPG have sought to overcome those impediments through network sharing arrangements. This section describes those matters.
445 The announcement of the TSSR guidance was made by the Commonwealth Government in August 2018,427 and the changes came into effect in September 2018.428
446 Both Optus and TPG are affected by the TSSR guidance requiring them not to use radio access network equipment provided by “high-risk” vendors in their 5G rollouts.429 For both Optus and TPG, this TSSR guidance prevents the use of Huawei equipment in their 5G networks, which they had previously used in their 4G networks and had planned to continue using alongside Huawei 5G equipment.430
447 Mr White of Optus gave evidence that a consequence of the TSSR guidance is that Optus is not only prevented from using Huawei equipment for 5G, but is also required to swap out and replace its 4G Huawei equipment if and when it deploys 5G at a site with 4G Huawei equipment, and that this requires the removal and write-off of high quality well-performing 4G equipment that is still well within its useful life.431 Mr Kanagaratnam further explained that the Huawei 5G ban meant that MNOs would not be allowed to operate a 5G mobile network in which transmissions using 5G mobile technology passed through Huawei RAN equipment at any stage (ie including any Huawei 4G RAN equipment to which the network was connected when providing 5G mobile services). Optus’s initial 5G mobile network rollout is based on a "non-standalone" network, which means that elements of Optus’s existing 4G network (including the 4G RAN equipment) are required to be used in providing 5G mobile services on the network. This meant that Optus has to remove the Huawei 4G RAN equipment from its network and replace it with equivalent equipment provided by a non-Huawei vendor (Ericsson and Nokia).432
448 The position of TPG is similar. The application for authorisation, verified by Mr Berroeta,433 stated that:434
87 TPG has made limited investments in the 17% Regional Coverage Zone in recent years, focusing more on the 5G roll out in the metropolitan areas. Prior to its merger with VHA, TPG had ceased the roll-out of its mobile network in Australia in 2019 because, following the Australian Government’s Security Guidance, TPG considered that it did not make sense to invest in a network that could not be upgraded to 5G. While TPG’s merger with VHA means it now has its own infrastructure in the 17% Regional Coverage Zone for 4G services up to 96%, VHA’s infrastructure also largely utilised Huawei equipment and was similarly impacted by the Security Guidance, which led to a delay in its 5G network rollout. Following the Security Guidance, VHA needed to find a new equipment vendor and swap out its existing Huawei 3G/4G equipment from sites to be upgraded to 5G. That network, now owned by the merged TPG/VHA, supplies its 3G/4G/5G network services in metropolitan areas using Nokia equipment.
88 The Security Guidance continues to impact TPG’s network. Wherever TPG seeks to offer 5G services, it is required to swap out its existing Huawei 3G/4G RAN equipment and replace it with non-Huawei 3G/4G/5G equipment. In 2020, TPG commenced network upgrades replacing its metropolitan 3G and 4G Huawei network with Nokia 3G and 4G substitute equipment on sites where 5G was being deployed. The additional costs of the full replacement of 3G/4G Huawei equipment has meant less capital has been available to expand or upgrade its regional network. [[REDACTED] [REDACTED] [REDACTED]] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] - [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
89 [REDACTED]
449 Telstra is not similarly affected by the TSSR guidance. Telstra uses very limited high-risk vendor equipment in its network.435
450 The TSSR guidance has imposed additional costs on both Optus and TPG in rolling out their 5G networks. Optus estimates that the TSSR guidance [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] “[REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]” [REDACTED] “[REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]”).436 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].437
451 Internal Telstra documents demonstrate that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. Telstra documents estimate that “[REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]” and that “[REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]”.438
Optus’s history of network investment
452 Optus operates a substantial mobile network in Australia which has been established with large capital investments. The evidence before the Tribunal looks more closely at the capital investments decisions made by Optus in the last 5 or 6 years, and specifically its 5G Business Case and more recent investment decisions relating to the 5G rollout.
453 Ms Bayer Rosmarin gave evidence that, in 2017, Optus announced that it had committed $1 billion of investment to improve regional mobile services by upgrading 1,800 sites from 3G to 4G and building 500 new 4G sites by the end of June 2018. That followed an ongoing program of investment by Optus which had directed more than $3.6 billion into its mobile networks since 2015. Ms Bayer Rosmarin said that, despite this investment, Optus’s market share had not improved and had remained largely static for the 10 years to December 2020. For the full year ended 31 March 2021, Optus’s EBITDA declined by 25% and EBIT declined by 77%.439
454 Ms Bayer Rosmarin gave detailed evidence about the commercial considerations that were relevant to the development and ultimate approval of Optus’s 5G Business Case in July 2020.440 She explained that the 5G Business Case was always challenging because of competing considerations. On the one hand, investing in 5G made commercial sense because of competitive pressures from Telstra and TPG. Ms Bayer Rosmarin explained that:441
(a) in 2020, both Telstra and TPG were focused on their 5G network rollout;
(b) coverage is one of the most important factors for winning and retaining customers, and investment in 5G is important to attract and retain customers in both regional and metropolitan areas as 5G is important to increasing network capacity and therefore the quality of the network that can be provided to customers;
(c) in a three player market, maintaining scale is important to offsetting high capital costs of mobile network infrastructure (which is compounded by frequent technological changes); and
(d) Optus’s modelling showed that Optus would lose value if it did not invest in 5G technology, given its main competitor, Telstra, was rolling out 5G.
455 On the other hand, the 5G investment was challenging because:442
(a) [REDACTED];
(b) the advantages that Telstra has over Optus have been further exacerbated by the Huawei 5G ban which means that upgrading a site to 5G is more expensive and time consuming for Optus than Telstra because existing 4G Huawei equipment in Optus sites needs to be replaced. The Huawei 5G ban gave Telstra a cost advantage; and
(c) [REDACTED].
456 The modelling that underpinned the 5G Business Case assessed [REDACTED] [REDACTED] [REDACTED]: [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] “[REDACTED] [REDACTED] [REDACTED]”; [REDACTED] [REDACTED] [REDACTED] [REDACTED] “[REDACTED]” [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]; [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] “[REDACTED] [REDACTED]” [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]).443 Optus’s modelling estimated that [REDACTED] “[REDACTED]” [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED]) [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] “[REDACTED] [REDACTED]” [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
457 The 5G Business Case that was approved by the Singtel Board in July 2020 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].444 The minutes of the Board meeting noted that “[REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]”.445 In that regard, the financial metrics presented in the 5G Business Case ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]) [REDACTED]:
(a) [REDACTED];
(b) [REDACTED];
(c) [REDACTED].446
458 The 5G Business Case recognised and evaluated [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].447
459 Mr Kanagaratnam also gave detailed evidence about [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].448
460 Ms Bayer Rosmarin gave evidence that, since the 5G Business Case was approved in 2020, Optus management has [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:449
(a) [REDACTED].450 [REDACTED].
(b) [REDACTED].451 [REDACTED].
461 Ms Bayer Rosmarin said that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].452 In that regard, Ms Bayer Rosmarin gave evidence that [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED]) [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. 453 [REDACTED] [REDACTED], Telstra’s ROIC for FY22 was 7.1%.454
462 It is also significant to note that Optus has a history of reducing the cost and risk profile of capital investment in its mobile network through infrastructure sharing. Each of Mr White and Mr Kanagaratnam gave evidence that Optus and TPG have been parties to a joint venture arrangement (referred to as the eJV) since 2004. The eJV provides for [REDACTED] [REDACTED] [REDACTED] [REDACTED] in metropolitan areas ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]) and 3G roaming in regional and urban fringe areas. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].455
463 [REDACTED].456
464 [REDACTED].457 [REDACTED].458 [REDACTED].
465 Ms Bayer Rosmarin explained Optus’s commercial rationale for [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:459
[REDACTED]:
(a) [REDACTED].
(b) [REDACTED].
(c) [REDACTED].
(d) [REDACTED].
466 Mr White gave evidence that he considered there were clear incentives for both Optus and TPG to conclude a network sharing arrangement in that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].460 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:461
30. [REDACTED]
31. [REDACTED]
32. [REDACTED]
33. [REDACTED]
34. [REDACTED]
467 Mr Kanagaratnam gave similar evidence, although with greater emphasis to the benefits of Optus being able to pool spectrum with TPG. Mr Kanagaratnam said there are various reasons why an agreement with TPG would make sense for Optus:462
151. First, there would be a reduced cost to Optus due to the revenue generated from TPG's payments to access Optus' network. This would allow Optus to generate a higher return on its investment and to build and deploy more greenfield sites in areas where sharing was to take place.
152. Second, by pooling Optus and TPG spectrum, Optus's customers (as well as TPG's customers) would benefit from higher speeds and better network performance. I describe this benefit in further detail in section 5 below.
153. Third, with access to TPG's low band spectrum, Optus would be able to provide wider 5G coverage per site and therefore rapidly expand its 5G coverage footprint. Optus would also be able access additional sites from TPG that Optus does not have network equipment on at present resulting in increased coverage for Optus – this is similar to what Telstra and TPG are proposing to do in the Proposed Transaction. As discussed at paragraph 161 below, over time Optus would extend its network coverage by utilising [REDACTED] [REDACTED] [REDACTED] of TPG's sites.
154. Fourth, Optus would be able to deploy and sell fixed wireless access services due to the additional mid band 5G spectrum that Optus would be able to access through potential spectrum pooling between Optus and TPG. I describe this in further detail in section 7 below.
155. Overall, in my view an arrangement between TPG and Optus would make both Optus and TPG significantly stronger competitors to Telstra in regional areas. This will apply across multiple important network attributes; coverage, speed, and overall network quality and performance.
TPG’s history of network investment
468 TPG has the smallest mobile network of the three MNOs. The network was largely established by Vodafone which merged with the former TPG Telecom Limited in 2020 to become the new TPG.
469 As is apparent from the discussion concerning Optus, TPG also has a history of reducing the cost and risk profile of capital investment in its mobile network through infrastructure sharing. Mr Berroeta also gave evidence about the eJV between Optus and Vodafone, noting that:463
(a) the original agreement reached in November 2004 was to form a joint venture to share 3G network sites in the [REDACTED] [REDACTED] [REDACTED] [REDACTED];
(b) in May 2012, the eJV was expanded to share access rights on 4G network sites [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED];
(c) also in May 2012, Optus and TPG entered into a 2G and 3G domestic roaming agreement in the 80% to 96% population coverage area.
470 Mr Berroeta said that, at the time the roaming agreement was entered into, Vodafone’s total population coverage on its own network was around 94% and the roaming agreement extended coverage to [REDACTED]. Today, TPG’s mobile network reaches 96% of the population without Optus’s 3G roaming and [REDACTED] with Optus’s 3G roaming.464
471 Mr Berroeta said that, from TPG's perspective, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].465
472 The evidence before the Tribunal indicates that, in the last 5 to 10 years, TPG has pursued the expansion of its mobile network in the >80% population coverage area through a combination of targeted infrastructure investment and infrastructure sharing arrangements.
473 [REDACTED].466 [REDACTED].467
474 [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. The ACCC conducted a declaration enquiry but in October 2017 the ACCC issued its final decision not to declare a regional domestic mobile roaming service. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].468
475 [REDACTED].469
476 Mr Berroeta said that, [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] the Commonwealth Government issued the TSSR guidance, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]).470
477 Mr Berroeta also said that, [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].471
478 [REDACTED].472 [REDACTED].473 The Project Hannibal negotiations resulted in the Proposed Transaction.
479 [REDACTED].474
F. COMPETITION IN THE FUTURE WITHOUT THE PROPOSED CONDUCT/PROPOSED TRANSACTION
Overview
480 In order to assess the effects of the Proposed Conduct and the Proposed Transaction on the nature and extent of competition in the retail and wholesale mobile services markets and any other relevant market, it is necessary first to consider the nature and extent of competition in the future without the Proposed Conduct or the Proposed Transaction occurring. The Tribunal considers that reliable predictions about the future can be made by analysing the manner in which the principal market participants, Telstra, Optus and TPG, have competed in those markets in the recent past and up to the present day. As was observed in the joint judgment of Middleton and O’Bryan JJ in Pacific National (in the context of a proceeding alleging a contravention of s 50 of the CCA):475
In the usual case, predictions about the nature and extent of competition in the future with and without the acquisition will be rooted firmly in past and present market conditions, which are susceptible of proof in the ordinary way. Most markets have a history from which an assessment of substitution possibilities, concentration, barriers to entry and other commercial behaviours and conditions can be undertaken and reliable predictions about the future can be made. Further, some future facts are more certain than others. For example, commercial firms and governments make plans about investment or entry into markets, which are observable facts able to be proved in the ordinary way. Such facts provide a platform on which a court is able to undertake the assessment required by s 50.
481 In this section of the reasons, the Tribunal draws on the factual background contained in the previous section, and other material that is before the Tribunal, to assess the likely competitive behaviours of each of Telstra, Optus and TPG in the wholesale and retail mobile markets, and thereby the nature and extent of competition, in the future without the Proposed Conduct or the Proposed Transaction occurring.
482 One important matter should be mentioned at the outset. Part of the evidence before the Tribunal comprises witness statements from executives of Telstra, TPG and Optus which contain predictions about the future behaviour of their respective companies if the Proposed Transaction does not proceed. While the Tribunal has had regard to, and places some weight on, those statements, the Tribunal is also conscious that the opinions expressed by the executives is likely to be coloured, whether consciously or unconsciously, by the interest that their respective company has in the application for authorisation. That factor has particular significance with respect to the opinions expressed by the executives about the prospects that, in a future where the Proposed Transaction does not proceed, TPG and Telstra may enter into another form of network sharing arrangement with each other and TPG and Optus may enter into some form of network sharing arrangement with each other. As noted earlier in these reasons, the applicants contend that, if the Proposed Conduct or the Proposed Transaction does not proceed, there is no prospect that Telstra and TPG would enter into any other form of network sharing arrangement with each other. The applicants submit that Telstra would have no ability or incentive to enter another arrangement both because of the finely balanced economics of the current MOCN arrangement, and for fear of being again opposed by the ACCC. TPG also contends that, if the Proposed Transaction does not proceed, there is no commercially realistic prospect of a comparable active network sharing arrangement between TPG and Optus. In assessing those prospects, the Tribunal gives more weight to the internal business documents of each company than opinions expressed in statements filed in these proceedings. The Tribunal considers that the former are a more reliable guide to the relevant commercial and economic considerations that will influence commercial decision-making in the future.
483 As these reasons will explore, the assessment of whether the Proposed Transaction would be likely to have the effect of substantially lessening competition is finely balanced. The Tribunal considers that there are a number of ways in which the Proposed Transaction might be modified by Telstra and TPG in order to avoid any material anti-competitive effects. Whether or not Telstra and TPG would ultimately find an altered arrangement commercially attractive is a matter that cannot be assessed by the Tribunal. However, the Tribunal does not consider it to be self-evident that an alternative arrangement could not be mutually beneficial to the parties.
Telstra as an independent competitor
484 Telstra is, by a considerable margin, the strongest competitor in the retail mobile services market. It is the largest MNO by both network size, spectrum holdings and customer numbers and its 5G deployment is the most extensive in Australia. Telstra has the most extensive fibre backhaul network. Telstra has the greatest national market share of retail services by a significant margin (44% as at 2021, compared to 31% for Optus and 17% for TPG) and also has the highest ARPU by a significant margin ($48.74 per month for post-paid customers, followed by TPG at $42 and Optus at $39). Telstra holds very large customer shares in regional areas, ranging from [REDACTED] in Queensland to almost [REDACTED] in Western Australia. It currently operates around 3,700 mobile sites in the RCZ, compared to Optus’s 2,500 and TPG’s 749 sites in the RCZ. In his s 155 examination, Mr Penn acknowledged that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].476
485 Telstra’s T25 strategy for its mobile network, as summarised in its 2022 Annual Report, included the following goals (to be achieved by FY25):
(a) extending 5G coverage to approximately 95% of the population and with more than 80% of mobile traffic using 5G;
(b) to win the majority of key surveys for best mobile network including with respect to coverage and speed;
(c) expand regional coverage by 100,000 km2.
486 While Telstra will gain commercial advantages from the Proposed Conduct and the Proposed Transaction (as discussed later in these reasons), there is no suggestion in the evidence that Telstra’s T25 strategy is dependent upon the Proposed Transaction or any aspect of it.
487 The additional spectrum that Telstra will gain under the Spectrum Authorisation Agreement will alleviate congestion in certain parts of Telstra’s network in the RCZ and enable Telstra to improve the quality of service (particularly speed) to customers experiencing congestion. As discussed in the following section of these reasons, the evidence shows that this will be a significant competitive benefit to Telstra. However, the evidence also shows that Telstra has the resources to overcome congestion issues in its network without the Spectrum Authorisation Agreement, albeit at a higher capital cost, by investing in additional mobile sites. In his statement, Mr Penn said:477
66 Absent the Proposed Transaction, Telstra will therefore continue to make investment decisions in accordance with its current T25 Strategy to seek to address congestion where it occurs. Especially in regional areas, we will remain constrained by our limited low-band spectrum holdings. This means that we would need to continue to deal with anticipated congestion concerns as they arise, on a case-by-case basis.
67 However, without the spectrum pooling provided by the MOCN, Telstra will need to invest materially more in network infrastructure in the RAN in order to compensate for our lower spectrum holdings. This will require both additional time and cost and we may find that it is not commercially feasible to deliver the same level of service to the same areas as we would otherwise have been able to achieve through pooling of spectrum with TPG.
488 In his s 155 examination, Mr Penn further explained:478
[REDACTED]
Optus as an independent competitor
489 Optus is behind Telstra in most of the key competitive indicators in the mobile services market: network size, customer numbers and 5G deployment. It is also behind TPG on some indicators; for example, while the data suggests that Optus has a larger market share than TPG on a national basis, TPG has a larger number of SIOs in metropolitan areas. Optus has the lowest ARPU of the three MNOs.
490 Nevertheless, Optus is a significant competitor in the retail mobile market. Its market share of 31% is substantial. The evidence indicates that, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:479
[REDACTED]
491 Optus is presently rolling out 5G technology across its mobile network and its current strategy is to [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. This is [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. As at 31 January 2022, in the major cities, Telstra had 3,140 5G sites and Optus had 1,790 5G sites. In relation to the RCZ, the vast majority of Optus’s rollout of 5G sites will occur [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. The TSSR guidance has imposed additional costs on Optus in rolling out its 5G networks. Huawei equipment is required to be removed from [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] and Optus estimates that the TSSR guidance will [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. Optus also has a significant scale disadvantage in comparison to Telstra in rolling out its 5G network in the RCZ. It has far fewer customers who reside within the RCZ, and also has a smaller customer base in metropolitan areas (some of whom would travel to areas within the RCZ and wish to obtain mobile coverage). The evidence before the Tribunal, summarised earlier in these reasons, shows that Singtel [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
492 Overall, the Tribunal’s view is that Optus suffers from significant competitive disadvantages in comparison to Telstra in seeking to upgrade its mobile network to 5G which will have medium term implications for competition in the supply of retail mobile services.
TPG as an independent competitor
493 TPG is also behind Telstra in most of the key competitive indicators in the mobile services market: network coverage, customer numbers and 5G deployment. It is also behind Optus on most indicators. The most recent ARPU data shows that TPG is substantially lower than Telstra but marginally higher than Optus.
494 TPG has an extensive mobile network in metropolitan areas, with 4,337 mobile sites in major cities as at 2022 (compared with 5,294 for Optus and 5,257 for Telstra).480 Its mobile network in the [REDACTED] [REDACTED] [REDACTED] [REDACTED] is also supported by the eJV with Optus under which the parties share their [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. In relation to its 5G rollout, as at 31 January 2022 in the major cities, TPG had 991 5G enabled sites compared with 3,140 for Telstra and 1,790 for Optus.481 Mr Chiarelli gave evidence that, in the 12 months to 30 September 2022, TPG removed 4G Huawei equipment, and replaced it with 5G equipment, on 1,093 metropolitan sites.482 [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].483
495 TPG’s network is substantially smaller than Telstra and Optus in the RCZ. As stated earlier in these reasons, in respect of the RCZ, TPG has 749 sites compared with 3,700 for Telstra and 2,500 for Optus. TPG’s 4G coverage extends only to 96% of the population and it relies on a 3G roaming agreement with Optus to provide coverage up to [REDACTED] of the population. TPG faces the same competitive disadvantages in the RCZ as Optus: additional costs are imposed on it by reason of the TSSR guidance and its faces very significant scale disadvantage in comparison to Telstra.
496 The parties to this review do not contest the ACCC’s finding that, in the future without the Proposed Transaction (and without an alternative Optus deal), TPG would undertake a targeted build of mobile sites in the RCZ and that a targeted build would involve approximately the following:
(a) TPG gradually building a limited number of sites;
(b) TPG not decommissioning mobile sites in the 80%+ population coverage area;
(c) TPG’s geographic coverage remaining less than Telstra’s and Optus’s coverage; and
(d) TPG retaining ownership of all of its active infrastructure (both the radio access network and core network).484
497 [REDACTED]:
(a) [REDACTED];
(b) [REDACTED]:
(i) [REDACTED]
(ii) [REDACTED]
(c) [REDACTED];
(d) [REDACTED].485
498 It is also supported by Mr Berroeta who said that, if the Proposed Transaction does not proceed, the alternative available to TPG (aside from a deal with Optus) is a targeted build of potentially around [REDACTED] new sites in the 80%-96% population coverage area over the next 10 years (targeted at more highly populated towns and holiday hotspots).486
499 Further, if the Proposed Transaction does not proceed, it would be necessary for TPG to find a means to commercialise its excess spectrum holdings in the RCZ. In that regard, Mr Berroeta stated:487
[REDACTED]
500 One of the questions the ACCC asked the ACMA was whether there was demand for TPG’s existing spectrum holdings. The ACMA responded as follows:488
As a general proposition, the ACMA is seeing increasing demand for spectrum to deliver increased capacity and data rates for a variety of use cases including wireless broadband, localised area wireless broadband and fixed wireless broadband, with interest in particular spectrum bands expressed to the ACMA through various consultations and feedback on the draft FYSO [Five Year Spectrum Outlook 2022-2027]. This increasing demand for low, mid and high band spectrum is relevant to spectrum currently held by TPG, as well as spectrum held by a range of other licensees. Beyond that, we cannot comment on any specific level of demand for TPG spectrum.
501 Submissions before the ACCC also provide evidence of the demand for spectrum in the RCZ from a range of infrastructure and wholesale service providers (including neutral hosting) and smaller MNOs including BAI Communications, Pivotel, and Field Solutions Group.489
502 Overall, the Tribunal’s view is that TPG suffers from significant competitive disadvantages in comparison to Telstra in seeking to upgrade its mobile network to 5G across its network, and disadvantages in comparison to Optus in seeking to upgrade its mobile network to 5G in the RCZ, which will have medium term implications for competition in the supply of retail mobile services.
Wholesale mobile services including infrastructure services
503 The above discussion of the competitive position of the three MNOs in the future without the Proposed Transaction is largely focussed on the retail mobile market. However, each of the MNOs have the ability to use their network assets to supply wholesale mobile services and infrastructure services more generally. The relative competitive capacity of each of the MNOs to supply wholesale or infrastructure services is dependent upon the relative extent of their network assets, which has been discussed above.
504 Historically, the MNOs have competed to supply wholesale mobile services to MVNOs. In the future without the Proposed Transaction, that competition would be expected to continue in largely the same way. Although the Tribunal was directed to little evidence about such competition, it would be expected that competition for the supply of wholesale mobile services to MVNOs would reflect the relative competitive positions of the mobile networks of the MNOs.
505 Supplying MVNOs is likely to be a means by which MNOs can effect a degree of price discrimination, whereby lower prices are charged to more marginal customers without the MNO lowering prices to customers who are attracted to the MNO’s principal brand. However, the Tribunal would expect that MVNOs would impose only a relatively weak competitive constraint on MNOs. This is for two reasons. First, with respect to price, it would be expected that an MNO would seek to recover its full wholesale costs of supply from the MVNO and accordingly price competition from the MVNO would be limited to the available resale margin. Second, with respect to service quality, the value added by an MVNO would be limited to the reselling component of the service, such as customer service, although MVNOs may be able to compete with respect to product bundles or plan inclusions. As noted earlier, the availability of newer product features, such as 5G, to MVNOs is often delayed until after their introduction on the flagship retail brands of the MNOs.
506 Different considerations apply with respect to infrastructure sharing, which involves the supply of infrastructure services and can also be regarded as a form of wholesale supply. That is because the incentives of the MNOs to engage in infrastructure sharing can differ. As Mr Feasey observed:490
13 Whether an individual operator will have an incentive to share its network assets with rivals will, in my experience, depend on its individual circumstances. This explains why we see a wide range of different sharing arrangements and why their character can change over time. Catalysts for operators to reconsider their network strategies include the prospect of additional costs which arise when the industry moves from one generation of mobile technology to another, when significant new spectrum is assigned or when existing sharing arrangements are due to expire or existing equipment must be decommissioned. Sharing requires that all participants be better off as a result of the arrangement, but the benefits that each obtains need not be of the same kind or of the same magnitude. Operators may contribute different assets to the sharing arrangement and this will be reflected in the commercial terms that are reached between them.
507 Optus and TPG have a history of infrastructure sharing in order to reduce the cost and risk profile of capital investment in their mobile network. The evidence suggests that, until the Proposed Transaction, Telstra had little involvement in infrastructure sharing and it can be inferred that it had little incentive to share its superior network with its rivals.
508 In a briefing note prepared by Optus for its CEO [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED]) [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].491 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:492
[REDACTED]
[REDACTED]
[REDACTED]
509 It is relevant to the assessment of the competitive effects of the Proposed Conduct and the Proposed Transaction to consider the likelihood that, in the future without the Proposed Transaction, any other form of network sharing would be agreed between TPG and Telstra or between TPG and Optus.
Network sharing between Telstra and TPG in another form
510 Telstra and TPG contend that, if authorisation is refused, there is no prospect that Telstra would agree another form of network sharing with TPG. They say that Telstra would have no incentive to do so both because of the finely balanced economics of the current MOCN arrangement, and for fear of being again opposed by the ACCC. On that basis, they contend that, in the future without the Proposed Transaction, Optus would be the only MNO with which TPG could agree a potential network sharing deal and this would increase Optus’s bargaining power and decrease TPG’s bargaining power, thereby lessening competition in the market for mobile network infrastructure services.
511 [REDACTED]:493
63 [REDACTED]
512 Other executives of Telstra and TPG gave evidence that they considered that a MOCN arrangement, which facilitates spectrum sharing, is commercially more advantageous to them than, say, a MORAN arrangement or a roaming arrangement.494 As noted earlier in these reasons, typical MORAN deployments include the sharing of active base stations, but not spectrum.
513 The Tribunal accepts that the assessment of the likelihood that, in the future without the Proposed Transaction, Telstra would compete with Optus to offer another form of network sharing arrangement to TPG must have regard to Telstra’s incentives to so compete. In order to enter into a network sharing arrangement with TPG, Telstra must conclude that the commercial benefits of the arrangement outweigh the commercial detriments. In practical terms, the question is whether, if the Proposed Transaction did not proceed because it is not authorised, Telstra would be willing to agree to a varied arrangement with TPG which is less damaging to competition than the Proposed Transaction.
514 It is apparent that the Proposed Transaction entered into between Telstra and TPG has mutually beneficial elements, although the value of each of the benefits differs between the parties. From the MOCN service, TPG obtains the ability to provide a 4G and 5G service to its customers in the RCZ at a scale and in a timeframe that it could not achieve on a standalone basis, while Telstra obtains the financial benefit from the wholesale revenue earned. From the spectrum component, Telstra obtains network capacity benefits while TPG benefits financially from the spectrum payments. From the site component, TPG avoids costs with respect to mobile site closures, while Telstra again obtains network capacity benefits from additional mobile sites.
515 Despite the fact that the Proposed Transaction has mutually beneficial elements to the applicants, the Tribunal does not accept that no other mutually beneficial arrangement could be negotiated between the applicants. In particular, the Tribunal does not accept that an arrangement could not be negotiated which excluded the spectrum or site components, or included those components to a lesser extent, or reduced the geographic area covered by the MOCN service, and thereby diminished the potential anti-competitive effects of the transaction. The Tribunal has reached that view on the basis of the available evidence. The following matters are significant in that regard.
516 First, Telstra’s expert, Mr Feasey, expressed the opinion that Telstra would have an incentive to negotiate alternative arrangements, such as a MORAN, with TPG. Mr Feasey summarised the available options and the parties’ differing incentives as follows:495
39. If TPG and Telstra were not to implement the agreement TPG could pursue one of a number of alternative network sharing arrangements, and it could do so with either Telstra or Optus. This could include a Multi-operator RAN or MORAN arrangement with Telstra, a MOCN or MORAN arrangement with Optus or a domestic roaming arrangement with Optus. Any of these counterfactuals would enable TPG to compete more effectively with Telstra and Optus in the provision of 4G and 5G services in the relevant area than it does today and so would represent more competitive counterfactuals than the conditions of competition which prevail today.
40. I consider that Telstra would have an incentive to enter into a network sharing arrangement with TPG, including a MORAN arrangement, in relation to the provision of 4G and 5G services in the relevant area if it thought TPG to otherwise be able to obtain a network sharing arrangement of some kind with Optus. The fact that Telstra decided to enter into the MOCN agreement with TPG is evidence as to Telstra’s (and TPG’s) commercial incentives.
41. On the same reasoning, I consider that Optus would have an incentive to enter into an arrangement with TPG, but whether it would offer a MOCN, a MORAN or a domestic roaming arrangement would depend, amongst other things, on the extent to which Optus would expect TPG to favour the Telstra network over the network which Optus was able to offer, the nature of any commercial offer from Telstra, and the extent to which Optus required access to additional spectrum in order to meet both its own capacity requirements and the additional demand of TPG’s customers.
42. Since the limited reach of its own network in the relevant area means that TPG would have no alternative use for the spectrum it holds in those areas it does not cover, I would expect TPG to prefer a MOCN arrangement in which the value of the spectrum it could contribute to the arrangement could be reflected in the commercial terms of the transaction. It would prefer this to alternative arrangements, such as a MORAN or domestic roaming, in which TPG would retain exclusive use of its spectrum in the relevant area.
43. Although any of these alternative network sharing arrangements would enable TPG to be a more effective competitor to Telstra and Optus than pre-transaction, I consider that a domestic roaming arrangement with Optus would position TPG as a less effective competitor than it would be with either a MOCN or MORAN arrangement with Optus or a MORAN arrangement with Telstra. A MORAN arrangement with Telstra would offer TPG greater coverage and earlier provision of 5G services in the relevant area compared to a MOCN or MORAN arrangement with Optus. A MOCN arrangement with Optus would enable Optus to benefit from access to the TPG spectrum and TPG to obtain additional revenues in return from Optus in a way that a MORAN arrangement with either party would not. Overall, the competitive conditions associated with a counterfactual represented by a MOCN or MORAN agreement with Optus will in my view be sufficiently similar to those associated with a MORAN arrangement with Telstra (with any differences depending more on the precise ways in which each is implemented) to mean that it is not necessary to distinguish between them when undertaking the competitive assessment in this case.
517 Second, Mr Katinakis gave evidence that, when he joined Telstra in October 2018, the Australian telecommunications industry was entering a particularly challenging period with changing market dynamics, increasing data usage and shifts in customer expectation. Telstra had also publicly committed to a significant opex cost reduction program across the organisation, targeting a further $1 billion annual reduction in underlying core fixed costs by FY22 (in addition to the previous target of $1.5 billion).496 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. Mr Katinakis raised the prospect of a network sharing deal, and ideally MOCN, with TPG within the leadership team in or around November 2020. In or around December 2020, the leadership team agreed that Mr Penn would call Mr Berroeta to test TPG’s appetite for some kind of MOCN or similar deal as a way of starting their 5G deployment.497 In the joint announcement that was made to the Australian Securities Exchange on 21 February 2022, Mr Penn is quoted as saying that the transaction “allows Telstra to have an innovative way of monetising some of our active mobile infrastructure, in areas where the population coverage is much smaller and more challenging in terms of returns and further investment and where there are already a number of competitors”.
518 Third, the evidence of Mr Katinakis, and the documentary evidence before the Tribunal, shows that the form of the proposed network sharing arrangement between Telstra and TPG was not cemented for a considerable period of time. As Mr Katinakis explained, Telstra’s incentive to engage in negotiations with TPG was to a significant extent the product of a determination to monetise its 5G network, thereby reducing unit costs.
519 [REDACTED].498 [REDACTED].499 [REDACTED].500
520 A paper prepared by Mr Katinakis for a Telstra Board meeting on 12-14 October 2021 stated that, for the proposed commercial arrangements between Telstra and TPG, Telstra would be contributing the MOCN spectrum and was proposing to provide TPG with all the available 4G bands in the proposed MOCN coverage area in addition to 5G low and mid-bands, as they get built, in that area.501
521 The Telstra Board gave its final approval to the Proposed Transaction over a series of meetings in February 2022. [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. The paper prepared for the Telstra Board meeting on 14-16 February 2022 by Mr Sweers under the supervision of Mr Katinakis contained [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].502 The minutes of the Telstra Board meeting held on 14-16 February 2022 record that the Board approved entry into the MOCN Service Agreement and authorised the CEO to execute all ancillary documents including the Spectrum Authorisation Agreement and the Mobile Site Transition Agreement.503
522 It is apparent that a further Telstra Board meeting was held on Sunday, 20 February 2022. Mr Sweers prepared a further paper for the Board [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:504
• [REDACTED]
• [REDACTED]
• [REDACTED]
• [REDACTED]
• [REDACTED]
• [REDACTED]
• [REDACTED]
• [REDACTED]
523 [REDACTED]:505
[REDACTED]:
- [REDACTED]
- [REDACTED]
- [REDACTED]
524 On 20 February 2022, the Telstra Board confirmed its approval for the Proposed Transaction.506 The transaction was announced the following day.
525 Subsequently, and for the purposes of the application for authorisation, Telstra undertook further analysis of the potential financial and commercial benefits arising from having access to additional spectrum under the Spectrum Authorisation Agreement, including particularly the benefits of reducing congestion. In his s 155 examination, Mr Sweers confirmed that while the above description of the service quality benefits of additional spectrum (included in his paper for the Board meeting on 20 February 2022) was done at that time, the further analysis contained in the authorisation application was undertaken in preparing the application and not for the Board’s consideration.507
526 Having regard to the evidence before it, the Tribunal considers that, in the future without the Proposed Transaction, Telstra and TPG would have considerable incentive to attempt to negotiate an alternative arrangement which excluded the spectrum or site components, or included those components to a lesser extent, or reduced the geographic area covered by the MOCN service, and thereby diminished the potential anti-competitive effects of the transaction.
Network sharing between Optus and TPG
527 TPG contends that, in the future without the Proposed Transaction, there is no realistic prospect that TPG would agree an alternative network sharing arrangement with Optus, save for a 4G roaming agreement. Optus contends that an alternative network sharing arrangement is a realistic prospect.
528 In support of their respective contentions, TPG and Optus filed voluminous evidence [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. The evidence comprised statements by the executives [REDACTED] [REDACTED] [REDACTED] [REDACTED] and documents [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
529 In his statement, Mr Berroeta expressed the view that, [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].508 The Tribunal places little weight on statements of that kind in assessing the likely future behaviour of Optus and TPG. While that may be the present view of Mr Berroeta, the Tribunal considers that the view involves hindsight and is likely to be coloured by the fact that the statement has been prepared to support the application for authorisation. The Tribunal considers that the evidence filed by Optus shows that, [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED], [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. Similarly, TPG submitted to the Tribunal that the commercial relationship between TPG and Optus has deteriorated since the Proposed Transaction was announced, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] “[REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED],509 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]) and does not carry the weight that TPG seeks to place on it.
530 The Tribunal has considered the detailed evidence filed by the parties on this topic. The documentary record [REDACTED] [REDACTED], and the statements that explain the documentary record, are informative of the technical capability and the incentives of the companies to reach an agreement. Taking the evidence as a whole, the Tribunal draws the following conclusions.
531 First, both parties have a commercial incentive to share network assets in the 80%+ population coverage area to support the financial viability of their mobile businesses. As set out earlier in these reasons, Ms Bayer Rosmarin explained [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]: [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]; [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]; [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].510 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED],511 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
532 Second, [REDACTED]:512
[REDACTED]
533 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].513 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. As described earlier in these reasons, Telstra’s network in the 80%+ population coverage area was superior to Optus’s network in coverage, capacity and technology (in that Telstra’s 5G rollout was well advanced in comparison to Optus and 95% of the population would be reached by FY25). [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED]:
[REDACTED].514
534 [REDACTED].515
535 The delay in Optus rolling out its 5G network in the 80%+ population zone also has implications for spectrum sharing in the short term. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] ([REDACTED] [REDACTED] [REDACTED] [REDACTED]) [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].516
536 Third, [REDACTED].517 [REDACTED].518 [REDACTED].519 [REDACTED].520 [REDACTED]:
(a) [REDACTED].
(b) [REDACTED].
(c) [REDACTED].
(d) [REDACTED].
(e) [REDACTED].
537 [REDACTED].521
538 Fourth, it is common ground that, in the 80%+ population area, Optus and TPG do not share the same degree of contiguous low-band spectrum as is enjoyed by Telstra and TPG (although Optus and TPG have contiguous spectrum in the 1,800MHz and 3.5GHz bands).522 As discussed earlier in these reasons, spectrum assets which are contiguous are more valuable as they bring efficiencies in network deployment and enable MNOs to achieve greater speeds and capacity. Nevertheless, Mr Turner’s evidence is that, by pooling Optus’s and TPG’s spectrum, the combined Optus/TPG spectrum holding would be close to Telstra’s current spectrum holdings and would enable Optus/TPG to achieve a similar peak data rates to Telstra.523 Mr Kanagaratnam’s evidence was to the same effect.524 Mr Kanagaratnam also gave evidence that in the 80%+ population area, [REDACTED] [REDACTED] of Optus data traffic is carried on the low-band spectrum and that the addition of TPG's low-band spectrum could improve the experience and reduce the congestion of both Optus and TPG customers on the low-band coverage layer.525
539 There is some disagreement in the evidence as to whether, [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED],526 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].527 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].528 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].529 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].530
540 In the Tribunal’s view, in the medium term without the Proposed Transaction, Telstra will remain the strongest competitor in the retail mobile services market. Optus and TPG will remain vigorous competitors, but their respective market positions will be weaker than that of Telstra by reason of a number of factors including:
(a) less network coverage overall;
(b) less 5G network coverage; and
(c) delay and increased costs of 5G rollout due to the effects of the TSSR guidance.
541 There is a realistic potential for TPG to improve its network coverage through a network sharing arrangement with Optus or an arrangement with Telstra that addresses the anti-competitive features of the Proposed Transaction as discussed below.
G. THE EFFECTS, BENEFITS AND DETRIMENTS OF THE PROPOSED CONDUCT/PROPOSED TRANSACTION
Overview
542 Assessing the effects, benefits and detriments of the Proposed Conduct and the Proposed Transaction is also a forward-looking exercise. It involves a comparison of the future without that conduct, as discussed in the preceding section, to the future with the conduct. The evidence before the Tribunal primarily addressed the likely impacts of the Proposed Conduct and the Proposed Transaction on each of the MNOs. The Tribunal considers that a convenient framework in which to consider the likely effects, benefits and detriments of the Proposed Conduct and the Proposed Transaction is to consider the likely direct effects on each of the MNOs in turn, and then to consider the likely effects on relevant markets more generally (such as price and market share effects) to the extent reliable predictions can be made. For the reasons explained earlier, it is also necessary to assess the likely impact of the Proposed Conduct and the Proposed Transaction separately.
Telstra
543 Mr Penn explained that he engaged in negotiations with TPG with respect to a network sharing arrangement because he believed that such an arrangement could potentially offer a number of commercial benefits for Telstra, including:
(a) enabling Telstra to use its network investment more efficiently, including through monetising some of this through an additional wholesale revenue stream;
(b) giving Telstra access to additional spectrum to help alleviate the congestion challenge it faces more quickly and cost-effectively than rolling out additional sites; and
(c) [REDACTED].531
544 In respect of the Proposed Transaction that was ultimately agreed, Mr Penn said that he believed that the transaction would enable Telstra to achieve better utilisation and long term capital efficiency of its existing RAN infrastructure.532 He explained that the capital benefits of the deal have two aspects:
(a) first, the revenue from fees paid by TPG would contribute to meeting Telstra’s infrastructure costs in regional areas; and
(b) second, the pooled spectrum would provide Telstra with the ability to deliver increased capacity for regional customers at lower cost, by reducing the need to invest in densifying Telstra’s physical infrastructure through more sites and/or radios.533
545 Mr Penn said that access to pooled spectrum would help Telstra to grow its business in urban fringe and regional areas by better meeting the challenge of network congestion and the consequent impact on service quality in those areas.534
546 Mr Penn said that the Proposed Transaction also creates commercial risks for Telstra. The immediate network benefit it provides to TPG means that the Proposed Transaction will almost certainly result in Telstra losing some retail market share to TPG (and MVNOs that use the TPG network), especially in regional areas.535
547 Each of those benefits and risks is considered in turn. At the outset, the Tribunal notes that those benefits and risks were each referred to in the internal papers provided to Telstra’s Board for its meetings on 14-16 February 2022536 and on 20 February 2022537 at which the Proposed Transaction was approved, and were also discussed in the joint ASX announcement made by Telstra and TPG on 21 February 2022538 and an investor briefing given on 22 February 2022 by Mr Penn, Telstra’s CFO Vicki Brady, and Mr Katinakis in relation to the Proposed Transaction.539 References are made to those documents, and other evidence, in the following discussion.
548 Following that discussion, the Tribunal considers the potential effect of the Spectrum Authorisation Agreement and the Proposed Transaction on Telstra’s incentives to compete, including, in particular, its pricing incentives.
549 In relation to the MOCN Service Agreement, the paper prepared for Telstra Board meeting on 20 February 2022 stated that the financial modelling that had been undertaken by Telstra [REDACTED] [REDACTED]:
(a) [REDACTED];
(b) [REDACTED].
550 The paper also [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:
(a) [REDACTED];
(b) [REDACTED];
(c) [REDACTED];
(d) [REDACTED];
(e) [REDACTED];
(f) [REDACTED];
(g) [REDACTED].
551 The joint ASX announcement made on 21 February 2021 was reproduced earlier in these reasons. The benefits of the MOCN Service Agreement and the Mobile Site Transition Agreement to Telstra were described as follows:
(a) the MOCN Service Agreement would “provide significant value to Telstra’s wholesale mobile revenues”, “allows Telstra to have an innovative way of monetising some of our active mobile infrastructure, in areas where the population coverage is much smaller and more challenging in terms of returns and further investment and where there are already a number of competitors”, and “additional scale from this agreement therefore supports return on invested capital in these areas and makes ongoing investment in the network and innovation more sustainable”; and
(b) the Mobile Site Transition Agreement would allow Telstra to “deploy infrastructure on up to 169 TPG Telecom existing mobile sites, improving coverage for TPG and Telstra customers in the zone”.
Benefits of additional spectrum
552 The papers presented to Telstra’s Board for the meetings on 14-16 February 2022 and 20 February 2022 referred to the benefits of Telstra gaining additional spectrum under the Spectrum Authorisation Agreement.
553 As noted earlier in these reasons, the paper prepared for the meeting on 14-16 February 2022 contained [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED]:
[REDACTED]
554 The paper prepared for the meeting on 20 February 2022 contained [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED]:
[REDACTED]:
• [REDACTED].
• [REDACTED].
• [REDACTED].
555 Those benefits were verified in the evidence given by Mr Sweers.540
556 In a later section of the paper, the paper recorded [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED]:
(a) [REDACTED];
(b) [REDACTED].
557 Similarly, the joint ASX announcement made on 21 February 2021 stated that the Spectrum Authorisation Agreement would allow Telstra to “grow its network, increase capacity and continue to provide the country’s largest and fastest network”, it would “ensure that regional and rural customers will now experience faster speeds in more locations on their mobiles” and it would “also ensure that Telstra customers will experience significantly reduced congestion at busy times”.
558 The benefits to Telstra of the additional spectrum obtained under the Spectrum Authorisation Agreement was the subject of further comment from Telstra’s executives at Telstra’s Investor Briefing on 22 February 2022. Mr Penn said that the additional spectrum:
(a) exceeds any incremental capacity that Telstra would need to build to provide MOCN services to TPG’s customers; and
(b) provides capex efficiencies for Telstra in “hundreds of millions”.
559 Mr Penn’s observation reflects the quantum of additional spectrum that will be gained under the Spectrum Authorisation Agreement. As noted earlier in these reasons, and acknowledged by Mr Sweers and Mr Meissner, the low-band spectrum capacity in the RCZ that will become available to Telstra will immediately increase by more than 40%541 (from 2 x 35 MHz to 2 x 50 MHz), and will increase further to 2 x 60 MHz from 2024, but the additional TPG SIOs served by the MOCN represent less than [REDACTED] of Telstra’s SIOs. Thus, the increase in low-band spectrum outstrips the increase in SIOs served by a factor of [REDACTED] [REDACTED].
560 At Telstra’s Investor Briefing on 22 February 2022, Mr Katinakis described Telstra’s future spectrum use in the RCZ as follows:
(a) Telstra is currently using its 850 MHz spectrum for the 3G network which will be shut down by 2024;
(b) Telstra’s 700 MHz spectrum is fully utilised for its 4G network;
(c) as Telstra rolls out its 5G network, it will squeeze some capacity into its existing 850 MHz spectrum and it had recently acquired additional 850 MHz spectrum that will become available in 2024 onwards; and
(d) the additional spectrum acquired from TPG is available immediately and can be used to relieve congestion on the 4G network without incremental capital.
561 Mr Katinakis summarised the benefits of the additional spectrum as follows:
The other spectrum we can use for 5G, again immediately. As we deploy our own network it becomes a lot better, so not only makes us a lot more competitive, it makes us also a lot more able to serve the customers with a lot more capacity. So from a capex savings then, we are maintaining our capex guidance as it is; so we’re not going to increase our capital for any of this, we are going to absorb the TPG capacity in that deployment. And we are going to be left with a tonne of extra capacity that we are going to use for our own subscribers. And delay and avoid any deployment of new gear until later in the deal when it’s required. So that’s where the savings are coming from.
562 The evidence filed by Telstra placed significant emphasis on the congestion currently being experienced in Telstra’s network in the RCZ and the benefit that would be gained from using TPG’s spectrum to reduce that congestion.542 Some of that evidence has been referred to earlier in these reasons. Mr Meissner gave evidence that, as at March 2022, around [REDACTED] of Telstra’s 4G sites in regional and remote areas were congested, impacting approximately [REDACTED] of active 4G users. However, and as referred to earlier in these reasons, Telstra does not suggest that congestion cannot be addressed by means other than acquiring additional spectrum, such as site densification. The real benefit from access to additional spectrum is that congestion can be reduced at lower cost compared with site densification.
563 There was disagreement about the quantum of the capex savings that Telstra would gain from the additional spectrum. Telstra placed reliance on the calculation stated in Ms Ihaia’s report of [REDACTED] which was based on the capex forecasts she was given on instructions.543 Those capex forecasts were titled “Without the Proposed Transaction” and “With the Proposed Transaction” and relate to the financial years FY23 to FY31. Mr Sweers gave evidence that the forecasts were drawn from a model maintained by Telstra of its mobile network and used in its ordinary course of business.544 As Optus submitted, the capex savings calculated by Ms Ihaia (on the instructions given to her) are not a calculation of the capex savings that result from the additional spectrum able to be used by Telstra under the Spectrum Authorisation Agreement. The calculation made by Ms Ihaia is the difference between Telstra’s forecast capex without the Proposed Transaction and with the Proposed Transaction. The forecast with the Proposed Transaction includes the additional capex costs associated with carrying TPG’s customer traffic. To calculate the capex savings that would result from the additional spectrum, it would be necessary to compare Telstra’s forecast capex with the MOCN Service Agreement in place, but with and without the Spectrum Authorisation Agreement. The Tribunal was not taken to any calculation of that kind in the evidence. Optus observed that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. The Tribunal does not accept the validity of that calculation [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. However, the Tribunal does accept that the capex savings that would be generated from the additional spectrum are likely to greatly exceed the figure of [REDACTED] [REDACTED] [REDACTED] [REDACTED] referred to in Ms Ihaia’s report. That conclusion is consistent with Mr Penn’s statements at the Investor Briefing on 22 February 2022.
Avoiding regulation under Pt XIC of the CCA
564 The paper presented to Telstra Board meeting on 20 February 2022 listed [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. The Tribunal understands that this is a reference to [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. As noted earlier in these reasons, access to roaming services in Australia is by commercial agreement. The ACCC has conducted public declaration inquiries in 1998, 2005, and 2017, deciding each time to not declare (regulate) a domestic roaming service under Pt XIC of the CCA. It is apparent from the evidence [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
565 In the Tribunal’s view, it is neither possible, nor useful, to predict how the Proposed Transaction might affect any potential future declaration of a telecommunications service under Pt XIC of the Act. Any decision by the ACCC in the future would have regard to the market circumstances existing at the time.
566 Telstra recognised that the Proposed Transaction would improve the mobile network services that TPG is able to provide to retail and wholesale customers in the RCZ, including a 5G service, and thereby improve TPG’s competitive position in the retail and wholesale mobile service markets.545 That improvement could lead to TPG gaining customers at the expense of Telstra. As Mr Sweers explained, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].546
567 [REDACTED].
568 In its reasons for determination, the ACCC observed that the wholesale payments from TPG to Telstra:547
… could, all else staying the same, lessen the incentive for Telstra to compete with TPG and lessen the competitive constraint on Telstra’s pricing decisions. This is because the wholesale payments from TPG could be expected to reduce the cost (in terms of lost margin) that Telstra would otherwise face if it raised its prices and some customers switched from it to TPG under the Proposed Transaction.
569 Telstra’s submissions contrast the relatively large absolute size of the margin earned on their retail customers (estimated by Dr Padilla to be $27.48) compared to the [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] (estimated by Dr Padilla to be around [REDACTED]).548 However, this is not a relevant comparison for assessing the effect of the wholesale payments on the incentive to raise prices. The Tribunal agrees with the ACCC that, all else being equal, the effect of the variable elements of the wholesale payments is to increase the incentive for Telstra to raise prices. This is because the potential gain from charging higher prices to infra marginal (retained) customers is now offset by a smaller net loss from marginal (lost) customers if Telstra were to raise prices. It is not the absolute size of retail margins that is relevant to this assessment, but the changed trade-off between the increase in margins charged to retained customers compared to the lower loss of margin on lost customers.
570 Consistent with this, Dr Padilla recognises in his second report that the variable wholesale payments made by TPG to Telstra would offset to some extent the downward pressure on prices he predicts as a result of closer substitution between TPG and Telstra.549 Dr Padilla derives an estimate of Telstra’s current own price elasticity of demand from information on margins and estimates (on the basis of survey data regarding consumer preferences) a significant increase in that estimated elasticity of demand, from -1.54 to [REDACTED]. The net result is that the effect of the wholesale payments is swamped by the effect of the expected change to Telstra’s own price elasticity of demand and Dr Padilla predicts an overall negative effect on Telstra’s prices.550
571 In its reasons for determination, the ACCC concluded that the wholesale payments from TPG would not significantly affect Telstra’s pricing in the short term.551 The Tribunal agrees with this assessment. While the wholesale payments will increase the incentive for Telstra to raise prices, this will be offset to some extent by TPG’s service offer becoming a closer substitute for Telstra’s offer. The extent of the latter effect on Telstra’s demand elasticity and market share is unknown and subject to wildly different estimates by the parties themselves. Nevertheless, there will be some effect, at least in the short term. In the longer term, other factors will also come into play, such as the potential impact of [REDACTED] [REDACTED] [REDACTED] [REDACTED] on substitution and Telstra’s own price elasticity of demand and how wholesale charges may change relative to retail margins.
TPG
572 Mr Berroeta gave evidence that there are a number of features of the Proposed Transaction that are particularly important for TPG, and which will improve its ability to compete to attract and retain retail and wholesale mobile customers. Those features are:
(a) an increase to its network coverage;
(b) access to a 5G network;
(c) maintaining control of its core network;
(d) receiving services on a non-discriminatory basis; and
(e) seamless handover of mobile services between networks. 552
573 Mr Berroeta also said that the Proposed Transaction provides other commercial benefits to TPG, particularly:
(a) allowing TPG to monetise unused and underutilised spectrum;
(b) the Proposed Transaction is non-exclusive and enables TPG to concentrate its greenfield investments in metropolitan and peri-urban regions of Australia; and
(c) TPG is able to exit the Proposed Transaction after 10 years and provides a transition-out period of up to 36 months.553
574 Each of those benefits is considered in turn.
575 Mr Berroeta’s evidence is that the Proposed Transaction will provide TPG with coverage for 98.8% of the Australian population. This is an increase from TPG's existing coverage provided by its own network, which covers about 96% of the population. Further, TPG’s depth of coverage in the area covering 81.4% to 96% of the population will substantially increase from present, with the increased site density provided by the Proposed Transaction. The number of sites to which TPG has access in regional Australia will increase five-fold under the Proposed Transaction from around 725 sites to around 3,700 sites in the RCZ. In terms of geographic coverage, TPG’s existing 4G network covers around 650,000 km2 of Australia. With the MOCN, TPG’s 4G network geographic coverage will more than double to around 1,500,000 km2.554
576 In a paper dated 16 February 2022 prepared by Mr Czinner and presented to the TPG Board for its approval of the Proposed Transaction, Mr Czinner said:555
[REDACTED]
577 The Tribunal has previously noted Mr Berroeta’s evidence that TPG’s limited regional coverage has historically been, and continues to be, a significant barrier to TPG acquiring and retaining those consumer and enterprise mobile customers who value regional coverage, as well as wholesale mobile customers who wish to offer mobile services as an MVNO to customers seeking coverage in regional areas.556 The Tribunal has also previously noted Mr Cooney’s evidence that he expected that the increased network coverage that would be provided by the Proposed Transaction would enable TPG to [REDACTED] reduce its churn rate (loss of customers to its competitors). Internal TPG analysis shows that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
578 Although the Proposed Transaction will provide TPG with coverage for 98.8% of the Australian population, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. Given the competitive significance of customer perceptions of network coverage and quality, this limitation will reduce to some extent the competitive benefits obtained by TPG from the Proposed Transaction.
579 The Tribunal has earlier noted that perceptions of network quality and capability are crucial in driving consumer decisions. The MNOs are collectively investing billions of dollars into upgrading their networks to 5G, indicating a widespread belief that 5G is critical to their future ability to compete, or to offer future services that consumers are likely to demand, or both.
580 Under the Proposed Transaction, Telstra is required to make 5G available to TPG at a particular site in the RCZ 6 months after the site is activated for 5G for Telstra customers. This will also apply to any other future technology generation agreed to be added to the MOCN services. Some 5G sites will be immediately available to TPG on implementation of the Proposed Transaction because 5G will have been deployed at that site for at least 6 months by that time. Mr Berroeta said that, in his experience, a 6-month difference in the parties’ 5G network in discrete geographic areas will not have a material impact on TPG's competitiveness vis-a-vis Telstra. TPG has typically lagged behind Telstra by more than 6 months in terms of technology and coverage in the RCZ. The Proposed Transaction provides TPG with access to 5G in the RCZ a number of years earlier than it would otherwise be able to deploy such technology itself or access it through an agreement with another MNO (ie Optus), and in some areas TPG would not have deployed 5G at all. It will also provide customers seeking 5G coverage in the RCZ with a third alternative where currently only Telstra and Optus offer those services.557
581 The Tribunal accepts Mr Berroeta’s evidence that the Proposed Transaction would enable TPG’s network to provide a 5G service at a much earlier time than any alternative option available to TPG (continuing as an independent competitor, or acquiring a service from Optus in circumstances where Optus’s 5G rollout is a number of years behind Telstra’s rollout). However, the effect of the 6-month time lag in making 5G available to TPG is to enable Telstra to maintain a first-mover advantage in respect of 5G availability. As discussed earlier in these reasons, the first-mover advantage gives rise to a barrier to entry by reason of the “stickiness” of customers. It can be expected that the 6-month lag will give Telstra a significant competitive advantage to TPG in circumstances where TPG is otherwise sharing the Telstra network.
582 Under the Proposed Transaction, TPG will maintain control over its mobile core network and is solely responsible for operating its core. Mr Berroeta said that this means that TPG retains control to make enhancements, upgrades, interconnection arrangements and to acquire any goods or services from third parties for the purpose of developing the core network. Mr Berroeta explained that product plans and new products are built and controlled in the mobile core network.558 Mr Chiarelli gave examples of the types of products and services that MNOs can provide to customers under a MOCN, which could not be provided, or as easily, under a roaming or wholesale arrangement: 559
(a) 5G standalone services which will enable VoNR (i.e. voice over 5G), mMTC (massive Machine Type Communication) and uRLL (ultra Reliable Low Latency) services;
(b) caps on data volume and variable throttling of data, where customers pay for a plan under which their download and upload speeds are reduced once they have used a certain level of data in a defined period (TPG currently has 87 of these plans, servicing more than 2.8 million customers);
(c) real time alerts to customers based on the customer’s behaviour (of which TPG currently issues more than 6 million per month);
(d) commercial offers of additional services, or different plans, directed to customers based on their activity in the roaming areas (for instance, an offer of a plan by reference to the fact that a customer has just called a particular destination overseas).
583 Mr Berroeta said that the Proposed Transaction contains important non-discrimination principles, such that Telstra is not permitted to discriminate between the treatment of, and quality of service provided to, TPG's end users and Telstra's retail customers. The non-discrimination obligations mean that TPG will automatically derive any benefits Telstra makes to its RAN in the RCZ.560
584 The Tribunal accepts that non-discrimination principles are an important element of a network sharing arrangement that better facilitates competition between the service provider and the service acquirer. However, as noted earlier in these reasons (and acknowledged by Mr Berroeta in his statement), there are significant exceptions to the non-discrimination provisions, including:
(a) TPG will not have access to 5G-enabled sites until 6 months after Telstra has activated the sites for 5G;
(b) the non-discrimination provisions do not apply in respect of Telstra customers who have an enterprise grade product or a “special services”;
(c) fixed wireless services will only be supplied to TPG over 3.6 GHz spectrum on a 5G standalone basis, while the fixed wireless service supplied to Telstra comparison customers may use 3.6GHZ spectrum on a 5G non-standalone basis and may use other spectrum bands; and
(d) the provisions do not apply to NBIoT capability.
585 Mr Berroeta explained that seamless handover refers to the ability to maintain connectivity when switching from one network to another. It is possible under MOCN architecture because the core network does not differentiate between the radio access sites of the different operators. In roaming arrangements, the absence of seamless handover results in a very poor customer experience.561
586 The Tribunal has earlier noted that there are significant differences in customer experience between a roaming arrangement and a MOCN. Under a roaming arrangement, customers can experience call failures as the customer moves between networks whereas such call failures do not occur under a MOCN because the customer’s call remains controlled by the service provider’s core network.
Monetise TPG’s spectrum holdings
587 Mr Berroeta said that the Proposed Transaction has commercial benefit to TPG because it enables TPG to monetise unused and underutilised spectrum.562 In regional Australia, TPG has a relatively small number of sites (around 750), has not rolled out 5G and has a small share of supply (around [REDACTED] in the RCZ), all of which means that TPG’s low-band spectrum is presently under-utilised. Mr Berroeta said that the pooling of the parties' spectrum under the Proposed Transaction is an efficient method of utilising and monetising TPG's spectrum holdings in the RCZ, resulting in increased capital efficiency and reducing the overall operating expenditure of TPG's mobile network in that area.
588 The Tribunal accepts that the Proposed Transaction provides that commercial benefit to TPG. As noted in the preceding section, pooling spectrum also provides benefits to Telstra. The Tribunal is not persuaded, though, that without the Proposed Transaction TPG would be unable to commercialise its spectrum holdings in the RCZ though a transaction with another party, particularly Optus. The Tribunal considers it highly unlikely that there would not be an alternative use for the spectrum, and rejects the submissions (faintly advanced by the applicants) that without the Proposed Transaction the spectrum would lie “fallow”.
589 Mr Berroeta said that a key component of the Proposed Transaction is that it is not exclusive. The non-exclusive nature of the Proposed Transaction provides TPG with flexibility and an opportunity to pursue complementary deals, such as an expansion of the Optus eJV [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] and/or greenfield developments.563
590 The Tribunal accepts that a non-exclusive arrangement reduces the potential anti-competitive effects of a network sharing arrangement. However, the Tribunal places little weight on the prospect that TPG will, in the future with the Proposed Transaction, be able to expand the Optus eJV [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] or undertake any significant greenfield development of TPG’s network in the RCZ. As to the former, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. As to the latter, in circumstances where TPG entered into the Proposed Transaction to avoid incurring the costs of expending its own network in the RCZ, and where TPG is planning to decommission its mobile sites in the RCZ, the Tribunal considers that the prospect of TPG undertaking any material network investment in the RCZ during the term of the Proposed Transaction to be remote.
591 Mr Berroeta said that the Proposed Transaction preserves TPG's ability to exit the arrangements after 10 years (or after any optional extensions exercised by TPG) and provides a transition-out period of up to 36 months at TPG’s election.564 Mr Berroeta also made reference to the application for authorisation which claimed that, upon exit, a number of options will be open to TPG including:
(a) rebuilding its network in regional areas using any existing infrastructure that could host RAN equipment including existing towers and rooftops (not limited to those which it currently uses) with some greenfield development;
(b) a targeted site build in in the RCZ supplemented by other services such as low earth orbit satellite (LEOSat) services or neutral host provider services; and
(c) entering into one or more network sharing arrangements with other MNOs.
592 Mr Chiarelli gave more detailed evidence on those options.565
593 As stated earlier in these reasons, the Tribunal considers that any assessment of TPG’s commercial and competitive options upon exit is a wholly speculative exercise because the earliest time at which exit will occur is 10 years hence. For that reason, the Tribunal has placed no material weight on the contractual provisions governing exit nor the claimed options of TPG upon exit. The Tribunal’s focus is on the effects, benefits and detriments of the Proposed Conduct and the Proposed Transaction over the medium term, which is the period from the day and continuing during the initial 10 year term of the arrangements.
Optus
594 Evidence concerning the likely effects of the Proposed Transaction on Optus was given by a number of witnesses, including particularly:
(a) the CEO of Optus, Ms Bayer Rosmarin;
(b) the Managing Director of Wholesale and Strategy at Optus, Mr White;
(c) the Chairman of Optus, Mr O’Sullivan; and
(d) the Group CEO of Singtel, Mr Moon.
595 The evidence concerned Optus’s current market position, the challenges it currently faces with respect to its 5G rollout in regional areas, the expected effects of the Proposed Transaction on its current business model (and 5G rollout) and the likely changes to its business model as a consequence. Part of the evidence concerned Optus’s modelling of the financial impact of the Proposed Transaction on its current 5G Business Case. It is convenient to commence with the evidence concerning the financial modelling, before considering the evidence relating to the broader competitive pressures that will be faced by Optus if the Proposed Transaction proceeds in its current form.
Optus’s modelling of the financial impact of the Proposed Transaction
596 Ms Bayer Rosmarin gave evidence that, following the announcement of the Proposed Transaction, she considered it necessary to carry out an assessment of its effect on Optus’s mobile business, in particular the implications for Optus’s 5G rollout. Ms Bayer Rosmarin said that the 5G Business Case approved by the Singtel Board in July 2020 – and updated and implemented during 2020 and 2021 – relied on critical assumptions to support its viability. It was clear that the Proposed Transaction cut across a number of critical assumptions and could threaten the commercial viability and feasibility of aspects of the 5G investment. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].566
597 Ms Bayer Rosmarin said that she asked Mr White to lead the preparation of an initial assessment of the likely impacts for Optus of the Proposed Transaction. She explained that a key part of the work was leading a review of the impact of the Proposed Transaction on Optus’s 5G Business Case.567 That work culminated in the preparation of a paper for Optus’s Executive Committee dated July 2022 and titled “Regional 5G Plan in response to Telstra and TPG Network Sharing” (the July 2022 ExCo Paper).568 The ExCo Paper was co-authored by Mr White and Mr Kanagaratnam, together with Andrew Sheridan and Matthew Williams.
598 The central component of the July 2022 ExCo Paper was modelling undertaken by Optus of the likely impact of the Proposed Transaction on Optus’s 5G Business Case (the 2022 Optus Model). As set out earlier in these reasons, the 5G Business Case that was approved by the Singtel Board in July 2020 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. The July 2022 ExCo Paper recorded that market developments since 2020 had impacted the value of the approved 5G Business Case. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. For the purposes of the July 2022 ExCo Paper, Optus modelled the impact of the Proposed Transaction on Optus’s updated 5G Business Case [REDACTED] [REDACTED] [REDACTED]. The July 2022 ExCo Paper recorded that, [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:569
[REDACTED]
[REDACTED]
[REDACTED]
599 The July 2022 ExCo Paper recorded the following [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:
[REDACTED]
600 As can be seen from the above table, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. Mr White explained that this figure [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].570 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
601 The July 2022 ExCo Paper concluded that:571
[REDACTED]
[REDACTED]
[REDACTED]
602 The July 2022 ExCo Paper contained the following recommendations:572
[REDACTED]
• [REDACTED]
• [REDACTED]
• [REDACTED]
• [REDACTED]
• [REDACTED]
• [REDACTED]
603 Mr White explained the process undertaken by Optus in preparing the modelling that underpinned the July 2022 ExCo Paper. The underlying assumption is that the Proposed Transaction would [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].573 Mr White acknowledged that modelling of this kind is notoriously difficult and is based on the application of commercial judgment to a range of data points and factors.574 Nevertheless, Mr White explained that he expected the Proposed Transaction to [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:
(a) [REDACTED];
(b) [REDACTED];
(c) [REDACTED];
(d) [REDACTED].575
604 Mr White further explained that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:
(a) [REDACTED];
(b) [REDACTED];
(c) [REDACTED].576
605 Mr White said that the modelling showed that any future in which the Proposed Transaction is authorised would be seriously detrimental to Optus. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].577
606 Mr White also explained that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:
(a) [REDACTED];
(b) [REDACTED];
(c) [REDACTED];
(d) [REDACTED].578
607 Mr White also gave evidence that, since the July 2022 ExCo Paper, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].579 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].580 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].581 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].582 [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:583
[REDACTED]
Reliability of Optus’s modelling
608 The applicants submitted that the Tribunal should not place any weight on Optus’s July 2022 ExCo Paper or the modelling underpinning it. They argued that it was prepared by Optus, at least in part, to persuade the ACCC to refuse authorisation for the Proposed Transaction, and that the paper was one element in Optus’s “regulatory advocacy” against the Proposed Transaction. In support of those arguments, the applicants contrasted the analysis in the July 2022 ExCo Paper with earlier assessments of the Proposed Transaction made by Optus’s executives, and indeed earlier drafts of the July 2022 ExCo Paper, which did not contain the same pessimistic assessment as ultimately appeared in the final July 2022 ExCo Paper.
609 There was a substantial body of evidence before the Tribunal concerning Optus’s initial assessments of the Proposed Transaction and the preparation of the July 2022 ExCo Paper. That evidence included internal Optus communications following the announcement of the Proposed Transaction, drafts of the July 2022 ExCo Paper and internal responses to the drafts, the witness statements of Mr White and Mr Kanagaratnam who were two of the four authors of the July 2022 ExCo Paper, as well as transcripts of the s 155 examinations of Ms Bayer Rosmarin, Mr White and Mr Kanagaratnam in which they were asked questions about the July 2022 ExCo Paper. In forming a view about the weight to be given to Optus’s July 2022 ExCo Paper, the Tribunal has considered that evidence.
610 The Tribunal does not accept the applicants’ submission that no weight should be placed on Optus’s July 2022 ExCo Paper. To the contrary, the Tribunal considers that the paper represents a serious analysis of the possible effects on Optus from the Proposed Transaction. While the evidence shows that Optus’s assessment of the possible effects of the Proposed Transaction developed over the weeks and months following its announcement, that is hardly surprising and does not undermine the credibility of the analysis contained in the July 2022 ExCo Paper. The Tribunal considers that the explanations of the July 2022 ExCo Paper, and Optus’s overall analysis, provided by Ms Bayer Rosmarin, Mr White and Mr Kanagaratnam are credible.
611 The Tribunal accepts, though, that all financial models are dependent upon their assumptions. That was expressly acknowledged by Mr White in his evidence with respect to the Optus financial model. As was further acknowledged by Mr White, a key assumption underpinning the Optus financial model is [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. The important question is the reasonableness of those assumptions.
612 In his second report filed on behalf of Telstra, Dr Padilla undertook an assessment of the likely costs to Optus of continuing to invest in regional areas versus the cost to Optus of not investing and losing customers who value good coverage and quality.584 Dr Padilla assumed that the capex cost of upgrading 2,500 sites to 5G would be [REDACTED] [REDACTED]. With respect to the cost of not investing, Dr Padilla assumed (on instructions) that:
(a) if Optus invests to maintain reasonably competitive regional coverage, Optus’s market share would decline from [REDACTED] in Region 2b and [REDACTED] in Region 3 to [REDACTED] and [REDACTED] respectively in 2031;
(b) if Optus does not invest sufficiently to maintain reasonably competitive regional coverage, it would gradually lose its customers in the RCZ over the period to 2031 – which would risk around [REDACTED] [REDACTED] of profit (in present value terms with an assumed real cost of capital of 2.5%) over a ten year period – and would lose [REDACTED] of customers in Region 1 and 2a – which would place a further [REDACTED] [REDACTED] of Optus’s gross profit (in present value terms) at risk over a ten year period.
613 On the basis of the above figures, Dr Padilla expressed the following opinion:585
Given one-off costs of upgrading part or all of its network in the Regional Coverage Zone to 5G (of [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]) compared with a potential cost of not investing of close to [REDACTED] [REDACTED], I consider that Optus will continue to upgrade its network under the Proposed Transaction.
614 The differences between the Optus model and Dr Padilla’s model were considered in the ACCC’s reasons for determination. The ACCC prepared its own model to test the sensitivities of the assumptions used in each model (which has been provided to the Tribunal on this review).586 The ACCC observed that the models put forward are very dependent on the different assumptions regarding market share loss, margins likely to be earned by Optus on mobile subscribers, and the discount rate used to estimate the net present value of future investments. In that regard, the ACCC noted the following matters:587
(a) Dr Padilla’s model assumes that Optus will lose [REDACTED] subscribers living in the RCZ by 2031 if it makes no further 5G network investment in this area. The ACCC considered that, if Optus does not invest further but responds to the Proposed Transaction by reducing prices for its services, it may retain [REDACTED] [REDACTED] [REDACTED] price-sensitive customers in the RCZ (noting that TPG still has a market share of [REDACTED] of those subscribers living in regional Australia despite its present network inferiority in these areas).
(b) In contrast, Optus’s model assumes it will [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. The ACCC considered that this figure may be optimistically high, especially in light of estimates of TPG’s current market share in regional Australia. The ACCC observed that, using Optus’s modelling, if Optus’s market share in the RCZ fell to [REDACTED] by 2031, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
(c) The results in Dr Padilla’s model are highly sensitive to assumptions made about the margin Optus earns on additional customers it serves on its network. Dr Padilla assumes Optus would earn a margin of 65% on each customer on its network. This contrasts with Optus’s estimate of the margin on its subscribers in its modelling of [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. By adjusting the margin in Dr Padilla’s model to [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], the results of Dr Padilla’s model would be reversed such that the NPV of his estimate of Optus investing to extend its 5G network in a future with the Proposed Transaction becomes [REDACTED] [REDACTED].
(d) Dr Padilla’s modelling also uses a real discount rate of 2.5% to estimate the NPV of future incremental revenues and costs of different investment decisions for Optus. This number is [REDACTED] [REDACTED] than that used by Optus itself in its internal business case modelling. Adopting a nominal discount rate of [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] and adjusting this for an expected rate of inflation of 2.5% in line with the Reserve Bank of Australia’s long term target inflation level, this would imply a real discount rate of approximately [REDACTED]. If Dr Padilla’s model is adjusted for this discount rate, and all else being equal, the NPV of Optus’s return on investing in a regional 5G network would fall significantly, by [REDACTED] [REDACTED] to [REDACTED] [REDACTED].588
615 The ACCC sought to test the outcome of Dr Padilla’s model if it changed the following three assumptions used in that model (and held all other assumptions constant):
(a) Optus’s market share in regional Australia falling gradually to [REDACTED] (rather than to [REDACTED]) by 2031;
(b) the margin earned on additional customers being [REDACTED] (rather than 65%); and
(c) applying a real discount rate of [REDACTED] (rather than 2.5%).
616 The ACCC’s estimate is that those change in assumptions would cause [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].589
617 The Tribunal considers that the adjustments made by the ACCC with respect to margin and discount rate are reasonable adjustments, reflecting Optus’s internal financial metrics for decision-making. The Tribunal considers that the ACCC’s adjustment with respect to regional market share is unduly pessimistic. Based on regional market share figures produced earlier in these reasons, the adjustment requires an assumption that Optus’s regional market share would decline from around [REDACTED] to [REDACTED] notwithstanding that Optus has an extensive 4G network in regional areas. The Tribunal considers that Optus’s own estimate, that its market share would decline to [REDACTED] [REDACTED], is more reasonable having regard to the current market shares held by each of the MNOs, the competitive position of their respective mobile networks, and the evidence given by Mr White with respect to Optus’s estimate.
618 Overall, the Tribunal does not consider that the assumptions applied in Dr Padilla’s model are reasonable. Varying those assumptions to accord with Optus’s internal financial metrics and predicted market share changes reverses the results of Dr Padilla’s model.
619 TPG prepared two “aide memoire” documents, which are more accurately described as submissions, with respect to the 2022 Optus Model that underpinned the July 2022 ExCo Paper. Those documents advance the following criticisms of the 2022 Optus Model.
620 First, TPG observes that, in respect of each scenario modelled, the NPV figures are calculated as the NPV of free cash flow over the 15 year period of the model, discounted at the weighted average cost of capital (WACC). The free cash flow is calculated as the EBITDA of the mobile business (including 5G), less total capex (including network investment, mobile allocated capex, spectrum and other costs) – in other words, the NPV reflects free cash flow after capex costs have been subtracted from earnings. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].
621 While the Tribunal accepts that the Optus model shows that each of the modelled scenarios [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. More significantly, TPG’s observations do not undermine the conclusions expressed in the July 2022 ExCo Paper that [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. The further evidence of Ms Bayer Rosmarin, Mr O’Sullivan and Mr Moon in that regard is discussed below.
622 Second, TPG suggests that the discount rate used to calculate NPV in the model, being a WACC of [REDACTED], is higher that Optus’s “standard WACC” of [REDACTED]. The source relied on by TPG for Optus’s “standard WACC” is a footnote within the 5G Business Case.590 It is not at all clear what that footnote is referring to. However, it is clear that, as part of its ordinary business planning, Optus used a WACC of [REDACTED] for its assessment of the 5G Business Case and in making its investment decision. Another Optus document in evidence before the Tribunal shows that, in February 2021, Optus used the same WACC [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED].591 Optus has done likewise in re-assessing its future 5G investment decisions in light of the changed circumstances brought about by the Proposed Transaction in the 2022 Optus Model. TPG has not demonstrated that the WACC is unreasonable.
623 Third, TPG submitted that an assumption made by Optus with respect to a [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]:
(a) [REDACTED].
(b) [REDACTED].
(c) [REDACTED].
624 [REDACTED].
625 In response, Optus advanced two submissions. First, it observed that the 2022 modelling was based on different circumstances to the 2020 modelling. In the 2022 modelling, [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED], [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED] [REDACTED]. [REDACTED] [REDACTED] [REDACTED], [REDACTED] [REDACTED]: