Re Telstra Corporation Ltd (No 3)
[2007] ACompT 3
•17 MAY 2007
AUSTRALIAN COMPETITION TRIBUNAL
Telstra Corporation Ltd (No 3) [2007] ACompT 3
TRADE PRACTICES – application pursuant to s 152CE(1) of the Trade Practices Act 1974 (Cth) – application for review of decision of Australian Competition and Consumer Commission to reject access undertakings – unconditioned local loop service (“ULLS”) – whether monthly charge is reasonable – consistency with standard access obligations – averaging of network costs – retail price parity obligation – universal service obligation – universal service fund – whether averaging promotes the long‑term interests of end‑users – whether averaging is in the legitimate business interests of access provider – whether averaging is in the interests of persons who have rights to use the ULLS – direct costs of providing access to the ULLS – reasonableness of network modernisation provisions – efficient costs – reasonableness of PIE II model – international benchmarking – allocation of costs – weighted average cost of capital (“WACC”)
Trade Practices Act 1974 (Cth): ss 152AB, 152AH, 152AL, 152AR, 152AY, 152BS 152BU(2), 152BV(2), 152CE(1), 152CF(3)
Pts IV, V, XIB, XIC
Telecommunications Act 1997 (Cth)
Telecommunications Competition Bill 2002
Legislative Instruments Act 2003 (Cth): ss 55(2), 55(3)
Telecommunications (Consumer Protection and Service Standards) Act 1999 (Cth): ss 11, 11A, 12D, 154, 155, 157
Telecommunications (Universal Service Levy) Act 1997 (Cth)Telstra Corporation Limited [2006] ACompT 4, applied
Application by Optus Mobile Pty Limited & Optus Networks Pty Limited [2006] ACompT 8, applied
Application by Vodafone Network Pty Limited & Vodafone Australia Limited [2007] ACompT 1, applied
Sydney International Airport (2000) ATPR 41‑754, distinguished
Seven Network Limited (No 4) [2004] ACompT 11, distinguished
Re Queensland Co‑Operative Milling Association and Defiance Holdings (1976) 8 ALR 481, applied
Sydney Services Pty Limited [2005] ACompT 7, cited
R v Hunt Ex parte Sean Investments Pty Ltd (1979) 180 CLR 322, appliedFILE No 8 of 2006
RE:FINAL DECISION BY THE AUSTRALIAN COMPETITION AND CONSUMER COMMISSION DATED AUGUST 2006 IN RESPECT OF ORDINARY ACCESS UNDERTAKINGS SUBMITTED BY TELSTRA CORPORATION LIMITED FOR THE UNCONDITIONED LOCAL LOOP SERVICE
TELSTRA CORPORATION LTD (ACN 051 775 556)
Applicant
JUSTICE GOLDBERG (PRESIDENT), MR R DAVEY & PROFESSOR D ROUND
17 MAY 2007
MELBOURNE
IN THE AUSTRALIAN COMPETITION TRIBUNAL
FILE No 8 of 2006
RE:FINAL DECISION BY THE AUSTRALIAN COMPETITION AND CONSUMER COMMISSION DATED AUGUST 2006 IN RESPECT OF ORDINARY ACCESS UNDERTAKINGS SUBMITTED BY TELSTRA CORPORATION LIMITED FOR THE UNCONDITIONED LOCAL LOOP SERVICE
TELSTRA CORPORATION LTD (ACN 051 775 556)
APPLICANT
JUDGE: JUSTICE GOLDBERG (PRESIDENT), MR R DAVEY &
PROFESSOR D ROUND
DATE OF ORDER: 17 MAY 2007 WHERE MADE: MELBOURNE THE TRIBUNAL DECIDES THAT:
1.The decision of the Australian Competition and Consumer Commission on 25 August 2006 rejecting the two ordinary access undertakings given to it by Telstra Corporation Limited on 23 December 2005 is affirmed.
IN THE AUSTRALIAN COMPETITION TRIBUNAL
FILE No 8 of 2006
RE:FINAL DECISION BY THE AUSTRALIAN COMPETITION AND CONSUMER COMMISSION DATED AUGUST 2006 IN RESPECT OF ORDINARY ACCESS UNDERTAKINGS SUBMITTED BY TELSTRA CORPORATION LIMITED FOR THE UNCONDITIONED LOCAL LOOP SERVICE
TELSTRA CORPORATION LTD (ACN 051 775 556)
APPLICANT
JUDGE: JUSTICE GOLDBERG (PRESIDENT), MR R DAVEY &
PROFESSOR D ROUND
DATE: 17 MAY 2007 PLACE: MELBOURNE INDEX
1. INTRODUCTION [1] 2. PARTIES TO THE APPLICATION [5] 3. THE LEGISLATIVE REGIME [6] 4. THE UNCONDITIONED LOCAL LOOP SERVICE [14] 5. THE DECLARATION OF THE ULLS [17] 6. TELSTRA’S UNDERTAKINGS [19] 7. THE COMMISSION’S REASONS FOR REJECTING THE UNDERTAKINGS [32] 8. ISSUES [40] 9. CONSISTENCY WITH THE STANDARD ACCESS OBLIGATIONS [43] 10. AVERAGING OF NETWORK COSTS [53] 10.1 The basis of Telstra’s $30 charge [55] 10.2 Telstra’s Retail Price Parity Obligation (RPPO) [63] 10.3 Telstra’s Universal Service Obligation (USO) and the Universal Service Fund (USF) [69] 10.4 The averaging issue
[75]
11. DOES AVERAGING PROMOTE THE LONG‑TERM INTERESTS OF END‑USERS? [90] 11.1 The meaning of promoting competition in s 152AB(2) of the Act
[92]
11.2 Is averaging likely to promote competition in markets for listed services in urban areas?
[102]
11.2.1 Would increased ULLS charges resulting from averaging lead to less infrastructure‑based competition in urban areas?
[117]
11.2.2 Can Telstra compete in urban areas if it sets cost‑based de‑averaged access charges for the ULLS?
[121]
11.2.3 Conclusion on whether averaging is likely to promote competition in markets for listed services in urban areas
[127]
11.3 Is averaging likely to promote competition in markets for listed services in rural areas?
[133]
11.4 Is averaging likely to achieve the objective of any‑to‑any connectivity in relation to carriage services that involve communication between end‑users?
[147]
11.5 Is averaging likely to achieve the objective of encouraging the economically efficient use of, and the economically efficient investment in, the infrastructure by which listed services are supplied and any other infrastructure by which the services are, or are likely to become, capable of being supplied?
[148]
11.6 Conclusion on whether averaging promotes the long‑term interests of end‑users
[179]
12. THE LEGITIMATE BUSINESS INTERESTS OF TELSTRA AND ITS INVESTMENT IN FACILITIES USED TO SUPPLY THE ULLS
[180]
12.1 Will de‑averaging lead to cream‑skimming?
[193]
12.2 Does the RPPO only constrain the pricing of Telstra’s HomeLine Part and BusinessLine Part line rental services?
[210]
12.3 Would the Universal Service Fund (USF) be expected to compensate Telstra for losses in rural areas?
[225]
12.4 Is Telstra likely to be earning above normal profits from the provision of other services provided over the CAN that might compensate it for losses made when providing line rental services below cost in rural areas?
[240]
12.5 Is bypass of Telstra’s CAN likely if it were to set averaged ULLS charges?
[249]
12.6 Conclusion on the legitimate business interests of Telstra and its investment in facilities used to supply the ULLS
[256]
13. THE INTERESTS OF PERSONS WHO HAVE RIGHTS TO USE THE ULLS
[262]
14. THE DIRECT COSTS OF PROVIDING ACCESS TO THE ULLS [270]
15. THE OPERATIONAL AND TECHNICAL REQUIREMENTS NECESSARY FOR THE SAFE AND RELIABLE OPERATION OF THE CAN AND THE ULLS [275] 16. THE ECONOMICALLY EFFICIENT OPERATION OF THE CAN AND THE ULLS
[279]
17. OVERALL ASSESSMENT OF REASONABLENESS OF AVERAGING [281] 18. NETWORK MODERNISATION PROVISIONS [292] 19. TELSTRA’S NETWORK COSTS –THE PIE II MODEL [329] 20. EXOGENOUS ADJUSTMENTS TO THE COSTS ESTIMATED BY THE PIE II MODEL [367] 21. ALTERNATIVE SUPPORT FOR THE REASONABLENESS OF TELSTRA’S NETWORK COSTS
[374]
22. ULLS SPECIFIC COSTS
[387]
23. THE WEIGHTED AVERAGE COST OF CAPITAL [414] 23.1 The use of two distinct WACCs [415] 23.2 A correction for WACC asymmetry [433] 23.3 Are there asymmetrical effects in estimating the WACC? [440] 23.4 Allowing for the error in estimating the WACC [458] 23.5 Conclusion on a correction for WACC asymmetry [469] 24. CONCLUSION [474] ANNEXURE A GLOSSARY AND ABBREVIATIONS
REASONS FOR DECISION
THE TRIBUNAL: JUSTICE GOLDBERG (PRESIDENT), MR R DAVEY and
PROFESSOR D ROUND
1. INTRODUCTION
Telstra Corporation Limited (“Telstra”) has applied to the Tribunal pursuant to s 152CE(1) of the Trade Practices Act 1974 (Cth) (“the Act”) for a review of a decision of the Australian Competition and Consumer Commission (“the Commission”) to reject two ordinary access undertakings given by Telstra to the Commission under s 152BU(2) of the Act.
The access undertakings set out the price and certain non‑price terms and conditions upon which Telstra undertakes to provide its Unconditioned Local Loop Service. Telstra’s Unconditioned Local Loop Service was described in the undertakings as:
“…a service for the use of a continuous metallic twisted pair between the Network Boundary at the ULL End Customer Premises and a ULL POI associated with the TCAM serving that ULL End Customer.”
The Unconditioned Local Loop Service was declared by the Commission under s 152AL(3) of the Act. Two undertakings were given by Telstra to the Commission on 23 December 2005. They contained substantially the same terms and conditions. The first undertaking was in respect of the period from 1 January 2006 to 30 June 2007. The second was in respect of the period from 1 July 2007 to 30 June 2008. Each of the undertakings provided for a monthly charge of $30 per service. The Commission rejected the two undertakings in its Final Decision published on 25 August 2006 on the basis that it was not satisfied that the price and certain non‑price terms and conditions specified in the undertakings were reasonable.
The application for review was filed by Telstra on 14 September 2006. The issues before the Tribunal are whether the price and certain non‑price terms and conditions of the undertakings are reasonable having regard to certain statutory matters to which we shall refer. Annexure A contains a glossary of terms used in these reasons.
2. PARTIES TO THE APPLICATION
The Commission appeared to assist the Tribunal pursuant to s 152CF(3) of the Act, and the following parties were granted leave to intervene in the proceeding:
·Optus Networks Pty Limited (“Optus”);
·AAPT Limited (“AAPT”);
·Agile Pty Limited (“Agile”);
·Chime Communications Pty Limited (“Chime”);
·Macquarie Telecom Pty Limited (“Macquarie”);
·Primus Telecommunications Pty Ltd (“Primus”);
·PowerTel Limited (“PowerTel”).
Agile, Chime, Macquarie, Primus and PowerTel (“the other intervenors”) were jointly represented. All intervenors currently acquire the Unconditioned Local Loop Service from Telstra. AAPT filed a statement of facts, issues and contentions but took no further part in the proceeding.
3. THE LEGISLATIVE REGIME
The telecommunications access regime under Pt XIC of the Act was considered and explained recently by the Tribunal in Telstra Corporation Limited [2006] ACompT 4, in Application by Optus Mobile Pty Limited & Optus Networks Pty Limited [2006] ACompT 8 and in Application by Vodafone Network Pty Limited & Vodafone Australia Limited [2007] ACompT 1. We adopt those considerations and explanations in these reasons and provide the following summary of the provisions of Pt XIC that are relevant to our consideration in this proceeding.
Section 152AR of the Act sets out the standard access obligations (“the SAOs”) with which an access provider that supplies a declared service must comply. The SAOs set out the manner in which an access provider, if so requested by a service provider, must provide an active declared service. These obligations require an access provider, such as Telstra, to:
·supply an active declared service, in this case the Unconditioned Local Loop Service, to the service provider so that the service provider can provide carriage services and/or content services: s 152AR(3)(a);
·take all reasonable steps to ensure that the technical and operational quality of the service supplied to the service provider (which includes ordering and provisioning) is equivalent to that which the access provider provides to itself: s 152AR(3)(b); and
·take all reasonable steps to ensure that the service provider receives, in relation to the service supplied to it, fault detection, handling and rectification of a technical and operational quality and timing that is equivalent to that which the access provider provides to itself: s 152AR(3)(c).
(We describe the second and third of these obligations later in these reasons as “the equivalency SAOs”).
A carrier (or carriage service provider), such as Telstra, who is required to comply with the SAOs must comply with those obligations:
·on such terms and conditions as are agreed between it and an access seeker: s 152AY(2)(a);
·if the carrier (or carriage service provider) and the access seeker cannot reach agreement and the carrier (or carriage service provider) has had an access undertaking accepted by the Commission (or, on review, by the Tribunal), the terms and conditions with which the carrier (or carriage service provider) must comply are, to the extent that they are set out in the undertaking, as set out in the undertaking: s 152AY(2)(b)(i);
·if the carrier (or carriage service provider) and the access seeker cannot reach agreement and the carrier (or carriage service provider) has had an access undertaking accepted by the Commission (or, on review, by the Tribunal) but the undertaking does not specify terms and conditions about a particular matter, the terms and conditions with which the carrier (or carriage service provider) must comply are to be determined by the Commission acting as an arbitrator and its determination on the arbitration must be consistent with the SAOs and any operative undertaking: s 152AY(2)(b)(ii); and
·if there is no undertaking, on such terms as the Commission acting as an arbitrator may determine: s 152AY(2)(b)(iii).
A carrier (or carriage service provider) may submit an ordinary access undertaking to the Commission under which it undertakes to comply with the terms and conditions specified in the undertaking in relation to the applicable SAOs: s 152BS(1). The Commission must accept or reject the undertaking: s 152BU(2), but it must not accept an undertaking unless it is satisfied that:
·the undertaking is consistent with the SAOs applicable to the carrier or provider: s 152BV(2)(b); and
·the terms and conditions specified in the undertaking are reasonable: s 152BV(2)(d).
The sections of the Act critical to the Commission’s decision to accept or reject an undertaking, and any review of the Commission’s decision by the Tribunal, are ss 152AH and 152AB. Section 152AH(1) sets out matters to which regard must be had by the Commission (and by the Tribunal on review) in determining whether particular terms and conditions of an access undertaking are reasonable:
“(a)whether the terms and conditions promote the long‑term interests of end‑users of carriage services or of services supplied by means of carriage services;
(b)the legitimate business interests of the carrier or carriage service provider concerned, and the carrier’s or provider’s investment in facilities used to supply the declared service concerned;
(c)the interests of persons who have rights to use the declared service concerned;
(d)the direct costs of providing access to the declared service concerned;
(e)the operational and technical requirements necessary for the safe and reliable operation of a carriage service, a telecommunications network or a facility;
(f)the economically efficient operation of a carriage service, a telecommunications network or a facility.”
Section 152AH(2) provides that subsection (1) does not, by implication, limit the matters to which regard may be had in determining whether the terms and conditions specified in the undertaking are reasonable.
Section 152AB(2) provides, relevantly, that in determining whether the terms and conditions of an undertaking promote the long‑term interests of end‑users of carriage services or services supplied by means of carriage services (“listed services”), regard must be had by the Commission (and by the Tribunal on review) to the extent to which the terms and conditions are likely to result in the achievement of the following objectives:
“(c)the objective of promoting competition in markets for listed services;
(d)the objective of achieving any‑to‑any connectivity in relation to carriage services that involve communication between end‑users;
(e)the objective of encouraging the economically efficient use of, and the economically efficient investment in:
(i)the infrastructure by which listed services are supplied; and
(ii)any other infrastructure by which listed services are, or are likely to become, capable of being supplied.”
Section 152AB(3) provides that subsection (2) is intended to limit the matters to which regard may be had. Subsequent subsections of s 152AB expand upon the manner in which the Commission (and the Tribunal on review) is to have regard to those objectives.
Where in these reasons we refer to the “reasonableness” of a term or condition, we use the expression as shorthand to describe and explain the task that is committed to us by ss 152AH and 152AB. Section 152AH(1) does not confine the notion of reasonableness articulated in s 152BV(2)(d) because of the provision in s 152AH(2) and it is not exhaustive of the matters to which regard may be had. Nevertheless, when considering whether a term or condition (or matters such as a method or means used to arrive at a particular term or condition) is reasonable, we pay particular regard to the matters specified in s 152AH(1) and the objectives set out in s 152AB. We are not considering the reasonableness of the term or condition (or the method or means used to arrive at the term or condition) in the abstract, unrelated to the matters specified in s 152AH(1) and the objectives set out in s 152AB: Application by Vodafone Network Pty Limited & Vodafone Australia Limited (supra) at par [12].
In determining whether the terms and conditions of an access undertaking are reasonable, there are no absolute answers, nor is there necessarily only one correct approach: Application by Optus Mobile Pty Limited & Optus Networks Pty Limited (supra) at par [15]; Telstra Corporation Limited (supra) at pars [63] and [67].
4. THE UNCONDITIONED LOCAL LOOP SERVICE
The Commission’s Final Decision provides the following succinct description (which we adopt) of the Unconditioned Local Loop Service:
“… a service for access to unconditioned cable, usually a copper wire pair, between an end user and a telephone exchange. The ULLS essentially gives an access seeker the use of the copper pair without any dial tone or carriage service. This allows the access seeker to use its own equipment in an exchange to provide a range of services, including traditional voice services and high speed internet access, to the end‑user.”
The Commission’s Final Decision also describes Telstra as the predominant supplier of the Unconditioned Local Loop Service and describes the declared service as being:
“… used by access seekers to connect their own networks to existing infrastructure and deliver new and innovative high‑speed and data‑based services to end‑users more efficiently. It can also potentially be used to provide voice services more efficiently using voice over IP and DSL technologies. Possible services include high speed Internet access, ‘tele‑working’, distance learning, video‑on‑demand, remote local area network (LAN) access and other multimedia and data applications, as well as traditional local, STD and IDD call services in competition with Telstra.”
The Unconditioned Local Loop Service is an input that access seekers can use to provide retail fixed line services in competition with Telstra in retail markets for telecommunications services.
There are differences between the Commission’s description of the Unconditioned Local Loop Service which it has declared and Telstra’s description of the Unconditioned Local Loop Service which it supplies. We consider these differences later in these reasons. For convenience we generally refer to both descriptions by the acronym “ULLS”.
5. THE DECLARATION OF THE ULLS
Where an eligible service (that is, a listed carriage service within the meaning of the Telecommunications Act 1997 (Cth) or a service that facilitates the supply of such a service) is supplied by a carrier (or carriage service provider), such as Telstra, the Commission may declare the service if it is satisfied that the making of the declaration will promote the long‑term interests of end‑users of carriage services or of services provided by means of carriage services: s 152AL of the Act.
The Commission first declared the ULLS on 11 August 1999 varying that declaration on 18 May 2000. The Commission again declared the service on 28 July 2006 to take effect from the date of its gazettal pursuant to s 152AL(5) and to expire on 31 July 2009. That declaration was published in Gazette No GN. 31 of 9 August 2006. The ULLS was described in the declaration in the following terms:
“The unconditioned local loop service is the use of unconditioned communications wire between the boundary of a telecommunications network at an end‑user’s premises and a point on a telecommunications network that is a potential point of interconnection located at or associated with a customer access module and located on the end‑user side of the customer access module.”
The declaration included the following definitions:
“boundary of a telecommunications network is the point ascertained in accordance with section 22 of the Telecommunications Act 1997;
communications wire is a copper based wire forming part of a public switched telephone network;
customer access module is a device that provides ring tone, ring current and battery feed to customers’ equipment. Examples are Remote Subscriber Stages, Remote Subscriber Units, Integrated Remote Integrated Multiplexers, Non‑integrated Remote Integrated Multiplexers and the customer line module of a Local Access Switch;
…”6. TELSTRA’S UNDERTAKINGS
As noted in par [3] above, Telstra gave to the Commission two undertakings in relation to the ULLS. Telstra said it gave two undertakings:
“… to maximise the prospect of an undertaking being accepted. The Commission may have been or the Tribunal may now for example be satisfied that the terms and conditions of the Undertakings are reasonable in respect of 2006/07 but not in respect of 2007/08. In those circumstances, if one undertaking covered the entire period, it would likely be rejected. However, the existence of two Undertakings may lead to the acceptance [of] one of the Undertakings.”
The second undertaking was in substantially identical terms to the first undertaking save for its commencement and duration. It is to come into operation on 1 July 2007 and to continue, subject to certain provisions not relevant for present purposes, until 30 June 2008.
Each undertaking consisted of Recitals, Operative Provisions and three Attachments (A, B and C). The following paragraphs set out, or describe, the parts of the Recitals, Operative Provisions and Attachments that are relevant to our consideration of their terms and conditions.
Recital B to each undertaking stated that:
“Telstra gives this undertaking pursuant to Division 5 of Part XIC of the Trade Practices Act 1974 (Cth) (“TPA”) in relation to the Unconditioned Local Loop Service (the “Declared Service”) which was declared by the ACCC under section 152AL(3) of the TPA”.
Recital C stated that the terms and conditions specified in the undertaking principally related to matters of pricing and did not cover other matters relating to the supply of the service upon which Telstra and an access seeker must reach agreement prior to the supply of the service.
Clause 2 of the first undertaking set out the period of the undertaking in the following terms:
“2 Commencement and Duration
2.1 This Undertaking comes into operation upon the later of:
(a)acceptance of the Undertaking by the ACCC under Division 5 of Part XIC of the TPA; and
(b) 1 January 2006,
and continues, subject to paragraph 3.2(b), until the earlier to occur of:
(c) 30 June 2007; or
(d)termination or withdrawal in accordance with the TPA.
2.2For the avoidance of doubt, this Undertaking (including, without limitation, any prices in this Undertaking) has no effect in respect of any services that:
(a)have been supplied by Telstra to an Access Seeker prior to the date on which the Undertaking is accepted by the ACCC under Division 5 of Part XIC of the TPA; or
(b)are being supplied by Telstra to an Access Seeker under an existing agreement on the date on which the Undertaking is accepted for as long as that agreement remains on foot.”
Clause 2 of the second undertaking was in identical terms save that the commencement date in cl 2.1(b) was 1 July 2007 and the termination date in cl 2.1(c) was 30 June 2008.
Clause 3 of each undertaking was in the following terms:
“3 Undertaking Terms and Conditions
3.1Telstra undertakes to the ACCC that during the period this Undertaking is in effect pursuant to clause 2.1 it will comply with the terms and conditions specified in Attachments A, B and C of this Undertaking in relation to the standard access obligations applicable to Telstra in respect of the Declared Service.
3.2 For the avoidance of any doubt, this Undertaking:
(a)does not specify all the terms and conditions on which Telstra will comply with the standard access obligations that are applicable to it in respect of the Declared Service referred to in clause 3.1, but only some of them, and therefore does not constitute an offer by Telstra to provide the Service to an Access Seeker; and
(b)does not apply to the Declared Service to the extent that there are no standard access obligations applicable to Telstra in respect of the Declared Service for reasons including, without limitation, the granting of an exemption by the ACCC under section 152AT of the TPA in respect of the Declared Service, the variation or revocation of a declaration by the ACCC under section 152AO of the TPA in respect of the Declared Service or the expiry of a declaration in accordance with the TPA or a determination by the Regulator.”
Attachment A of the first undertaking contained a Service Schedule which provided, inter alia:
“2 Service Description
Telstra Unconditioned Local Loop Service2.1The following service description is provided for Telstra Unconditioned Local Loop Service and applies to the provision of the Telstra Unconditioned Local Loop Service by Telstra to the Access Seeker.
2.2Telstra Unconditioned Local Loop Service is a service for the use of a continuous metallic twisted pair between the Network Boundary at the ULL End Customer Premises and a ULL POI associated with the TCAM serving that ULL End Customer.
2.3Telstra Unconditioned Local Loop Service will support a connection with DC continuity.”
Clause 4 of Attachment A, which dealt with facilities access, noted, in accordance with Recital C, that facilities access was not a matter dealt with by the undertakings and that an access seeker would need to enter into such facilities access arrangements with Telstra as were necessary in order for it to connect its network to the ULLS at the ULLS point of interconnection (“POI”).
Clause 6 related to Network Modernisation and provided:
“6.1 The Access Seeker agrees that:
(a)Telstra has the right to maintain and upgrade its Network:
(b)provision of the Telstra Unconditioned Local Loop Service does not prevent, limit or restrict Telstra from maintaining or upgrading its Network; and
(c)the maintenance and upgrade of Telstra’s Network includes remediation, reconfiguration, enablement, augmentation, maintenance and repair of the Network (including the removal, rearrangement, replacement or decommissioning (for example with fibre optic cable) of the continuous metallic pair used for the supply of the Telstra Unconditioned Local Loop Service to the Access Seeker) (“Network Upgrades”).
6.2 The Access Seeker acknowledges that:
(a)any such Network Upgrade by Telstra may include the installation of TCAMs closer to ULL End Customers than a Telstra exchange building;
(b) any such Network Upgrade may require:
(i)the truncation of Telstra Unconditioned Local Loop Service provided from Telstra exchange buildings;
(ii)the establishment by Access Seekers of a new ULL POI or the relocation by Access Seekers of ULL POIs from one point to another (such as from a Telstra exchange building to the vicinity of relevant TCAMs); and/or
(iii)the alteration of deployment classes of Authorised Equipment used by Access Seekers on Telstra Unconditioned Local Loop Service; and
(c)a Network Upgrade may result in the Telstra Unconditioned Local Loop Service no longer being able to be supplied or the quality of the Telstra Unconditioned Local Loop Service (or any services supplied by Access Seekers to their ULL End Customers using the Telstra Unconditioned Local Loop Service) being adversely affected; and
(d)where a Network Upgrade does not result in the truncation of a Telstra Unconditioned Local Loop Service provided to the Access Seeker, but requires the alteration of the deployment class of Authorised Equipment used on the Telstra Unconditioned Local Loop Service, then subject to the required alteration being made to the deployment class, the Access Seeker may continue to access Telstra Unconditioned Local Loop Service.
6.3Where a Telstra Unconditioned Local Loop Service has been activated or Telstra has received a ULLS Request (as that term is defined in the ULL Ordering and Provisioning Code) and:
(a)the Network Upgrade will require the Access Seeker to take particular action in order to continue to use the Telstra Unconditioned Local Loop Service (such as that described in clause 6.2(b)); or
(b)the Network Upgrade will result in the Telstra Unconditioned Local Loop Service no longer being able to be supplied,
Telstra will provide the Access Seeker with notice of the Network Upgrade. The type of notice given and the length of notice period given in the Network Upgrade notice will vary depending on the type of Network Upgrade, but in the case of Network Upgrades other than Emergency Network Upgrades, Telstra will provide the Access Seeker with not less than 15 weeks’ prior notice of the Network Upgrade.
6.4Where:
(a)a Network Upgrade notice requires the Access Seeker to take particular action (such as that described in clause 6.2(b)) in order to continue to use the Telstra Unconditioned Local Loop Service and the Access Seeker fails to do so within the time specified in the notice; or
(b)the Network Upgrade is one that will result in a Telstra Unconditioned Local Loop Service no longer being able to be supplied,
Telstra has the right to terminate that Telstra Unconditioned Local Loop Service and the Access Seeker must comply with any notice from Telstra for the Access Seeker to submit a Handback (as that term is defined in the ULL Ordering and Provisioning Code).”
Attachment B of the first undertaking contained charges for the provision of the Telstra ULLS and provided, inter alia:
“2 Charges for Telstra Unconditioned Local Loop Service
2.1The charges payable by the Access Seeker to Telstra for Telstra Unconditioned Local Loop comprise:
(a)a once only charge payable at connection (“connection charge”) (although this charge is not a matter dealt with by this Undertaking);
(b) a monthly charge; and
(c)charges for operational aspects of the service such as service qualification inquiries and order withdrawals (although these charges are not a matter dealt with by this Undertaking).
2.2The charges payable by the Access Seeker to Telstra for Telstra Unconditioned Local Loop Service are geographically averaged.
3 Monthly Charge
3.1The monthly charge for each Telstra Unconditioned Local Loop Service is set out in Table x167.1. The monthly charge depends upon whether the Telstra Unconditioned Local Loop Service is connected at an IRIM/RIM/CMUX or an RSS/RSU.”
Table x167.1 was in the following form:
Location of
ULL POIMonthly Charge 1 January 2006 to 30 June 2007 RSS/RSU $30 per month IRIM/RIM/
CMUXNot dealt with by this Undertaking given the
limited demand at this point
The charges for the period of the second undertaking were in identical terms save that the words “given the limited demand at this point” did not appear in Table x167.1 of the second undertaking.
Attachment C to each undertaking was entitled “Standard Access Obligations”. It provided that Telstra would comply with the SAOs in the following terms:
“1.1Insofar as a Service, or part of a Service, is a Declared Service, then Telstra will, as required under Part XIC of the TPA, treat the Access Seeker on a non‑discriminatory basis as required by the standard access obligations in relation to the supply of that Service or that part of the Service, including but not limited to, if requested by the Access Seeker:
(a)taking all reasonable steps to ensure that the technical and operational quality of that Service or part of the Service, is equivalent to that which Telstra provides to itself; and
(b)taking all reasonable steps to ensure that the Access Seeker receives, in relation to that Service or that part of the Service, fault detection, handling and rectification of a technical and operational quality and timing that is equivalent to that which Telstra provides to itself.”
7. THE COMMISSION’S REASONS FOR REJECTING THE UNDERTAKINGS
Having assessed the key components of the undertakings against the matters set out in s 152AH(1), the Commission was not satisfied that the terms and conditions specified in the undertakings were reasonable.
More particularly, it was not satisfied that Telstra’s estimates of its network costs based on its PSTN Ingress and Egress II Model (“the PIE II model”) were reasonable because it had not discharged the onus on it to demonstrate the reasonableness of the model and its underlying assumptions. Thus, the Commission concluded that Telstra had failed to demonstrate that its costs were efficient.
Nor was the Commission satisfied that Telstra’s proposed averaged ULLS charge would be reasonable because the averaged charge would not reflect the costs of the ULLS in different geographic areas. It was of the view that this would adversely affect competition in the markets for basic telecommunications and broadband services and would distort usage and investment decisions, resulting in the inefficient use of, and investment in, telecommunications infrastructure.
Focusing on what it considered to be the high weighted average cost of capital (“WACC”) values used by Telstra to estimate its network costs, the Commission expressed concerns with Telstra’s reliance on Professor Robert Bowman’s methodology to arrive at those values.
In relation to the specific costs that Telstra claimed to incur in the supply of its ULLS, the Commission considered the most significant issue to be the appropriate cost recovery base. It concluded that:
“Even if the ACCC accepted Telstra’s claims in relation to the appropriate cost pool, the recovery of specific costs over all CAN lines or all ADSL [asymmetric digital subscriber line] lines would result in a per service charge substantially below that proposed by Telstra.”
Emphasising that it considered that the undertakings only covered certain forms of the declared services, the Commission formed the view that the undertakings, insofar as they addressed relevant provisions, were consistent with Telstra’s SAOs.
Commenting on the network modernisation provisions in the undertakings (par [28] above), the Commission noted that the issues surrounding network modernisation were inherently complex and considered that such clauses would:
“more usually be determined by bilateral commercial negotiation or by agreed operational procedures through self‑regulatory mechanisms.”
The Commission concluded that the proposed price and non‑price terms and conditions contained in the undertakings:
·were unlikely to promote the long‑term interests of end‑users as they would not promote competition and encourage the economically efficient use of, and investment in, infrastructure;
·resulted in Telstra recovering more than was necessary to promote its legitimate business interests;
·would harm the interests of access seekers and the persons who had rights to use the ULLS;
·exceeded the direct cost of providing access;
·did not have a material effect on the operational and technical requirements necessary for the safe and reliable operation of telecommunications services; and
·were not likely to facilitate the economically efficient operation of the ULLS.
Accordingly, the Commission was not satisfied that the terms and conditions specified in the undertakings were reasonable.
8. ISSUES
The principal issues for our consideration are:
·whether the undertakings are consistent with the SAOs applicable to Telstra: s 152BV(2)(b); and
·whether the terms and conditions specified in the undertakings are reasonable: s 152BV(2)(d).
Deciding whether the terms and conditions specified in the undertakings are reasonable includes a consideration of the following key issues:
·whether Telstra’s proposed monthly $30 per service charge for the ULLS is reasonable; and
·whether the network modernisation provisions are reasonable.
A consideration of the reasonableness of the monthly $30 per service charge raises for consideration the following significant issues:
·whether it is reasonable for Telstra to rely on the PIE II model to estimate the network costs that provide the basis for the $30 charge;
·whether the assumptions made, and the method used, by Telstra in arriving at its WACC are reasonable and, in particular, whether its:
ouse of two WACCs, one to determine the ULLS network costs and another to determine its ULLS specific costs, is reasonable; and
oincrease of a number of input parameters by one standard deviation to take account of claimed asymmetric social consequences of errors in setting the WACC, is reasonable;
·whether the approach taken by Telstra to determine its ULLS specific costs is reasonable and, in particular, whether:
oit is reasonable for Telstra to rely on the Specific Costs Model (“the SC model”) it created for the purposes of developing the undertakings; and
owhat it refers to as its Specific Cost Calculations (“the SC Calculations”) and estimates of its actual costs and its method of allocating those costs are reasonable; and
·whether, as claimed by Telstra, certain touchstones, namely:
ohistoric and current cost estimates derived from Telstra’s record keeping rules accounts (“the RKR accounts”);
oan estimate of its costs derived from a model developed in 1998‑1999 by NERA Economic Consulting (“NERA”) for the Commission and modified by the Commission in 2000 (“the NERA model”); and
ointernational benchmarks,
demonstrate the reasonableness of the $30 charge.
The issues listed above are by no means the only issues addressed in our reasons, nor are they the only issues that were addressed during the course of the review.
9. CONSISTENCY WITH THE STANDARD ACCESS OBLIGATIONS
The SAO imposed on Telstra under s 152AR(3)(a) is to supply “an active declared service” to service providers. The Commission declared the ULLS a “declared service” pursuant to s 152AL. We must not accept Telstra’s undertakings unless we are satisfied that they are consistent with the SAOs that are applicable to Telstra: s 152BV(2)(b). In deciding whether we might be so satisfied, we were faced with the following issues:
(a)whether the undertakings complied with the obligation in s 152AR(3)(a) to supply an active declared service, the ULLS, or whether the undertakings were given in relation to a service, the Telstra ULLS, which was different from the ULLS as declared and so limited in scope as to preclude us from being satisfied, as required by s 152BV(2)(b), that the undertakings were consistent with the SAOs; and
(b)whether either of the equivalency SAOs (s 152AR(3)(b) and s 152AR(3)(c): par [7]) impose an obligation on Telstra to inform a service provider to whom it supplies the ULLS of a network upgrade.
The issue in par [43(a)] arises because the description of the ULLS, which is the subject of the Commission’s declarations is different from, and not co‑extensive with, the description of the ULLS which is the Telstra service, the subject of the undertakings. Specifically:
·Telstra’s ULLS will support a connection with DC continuity but there is no requirement for Telstra’s ULLS to support any other service;
·Telstra’s ULLS involves the use of a continuous metallic twisted pair, whereas the declared ULLS involves the use of an unconditioned copper based wire; and
·the undertakings do not specify prices for the ULLS where the end‑user is connected to IRIM/RIM/CMUX units and therefore do not cover connections to any of these types of network nodes.
Nothing turns on these differences as the ULLS supplied by Telstra falls within the description of the ULLS in the declarations. Telstra’s ULLS answers the description of the ULLS in the declared service in the following respects:
·a continuous metallic twisted pair is an unconditioned communications wire;
·the use of that metallic twisted pair between “the Network Boundary at the ULL End Customer Premises and a ULL POI [point of interconnection] associated with the TCAM serving that ULL End Customer” as described in the undertakings, is a use between “the boundary of a telecommunications network at an end‑user’s premises and a point on a telecommunications network that is a potential POI located at or associated with a customer access module and located on the end‑user’s side of the customer access module” as described in the declarations; and
·although the undertakings do not deal with the charge for a service with ULL POIs which connect with IRIM, RIM or CMUX units, they do deal with services in which ULL POIs connect with RSSs and RSUs so that there is still a use as described in the declarations.
The issue in par [43(b)] arises because of the contention that the Network Modernisation Provisions contained in the undertakings are not consistent with Telstra’s obligation under s 152AR(3)(b) to take all reasonable steps to ensure that the technical and operational quality of the ULLS is equivalent to that which Telstra supplies to itself. Optus argued that in order for Telstra to meet its obligations, Optus would need to be given notice of all the details of network modernisation that would materially affect its operations at the same time as Telstra became aware of such matters, in order to allow it to respond at the same time, and in an equivalent manner, as Telstra. Optus contended that the Network Modernisation Provisions (par [28]) failed to provide such notice and instead provided that access seekers may only be provided with fifteen weeks notice.
We do not consider that the equivalency SAOs (subs 152AR(3)(b) and (c)) impose an obligation on Telstra to inform a service provider of a network upgrade that would impact on the provision of an active declared service. The subject‑matter covered by those subparagraphs relates to matters not directly involved in the fact of notification of a network upgrade. We adopt, for the purpose of the argument, the definition of “Network Upgrades” contained in cl 6.1 of the Service Schedule which is Attachment A to the undertakings (par [28] above). Insofar as that definition is relevant to a notification by Telstra of a Network Upgrade, the notification does not deal with or encompass components of the technical and operational quality (including ordering and provisioning which precede the supply of the service) of the service “supplied” to the service provider but rather relates to a forewarning or notification of an interruption, alteration or change to that service. Further, it does not deal with or encompass components of fault detection, handling or rectification of the service of a technical and operational quality or the timing of such fault detection, handling and rectification of the service. The reference to “timing” in s 152AR(3)(c) is a reference to the timing of the fault detection, handling and rectification of the service; it is not a reference to the timing of the giving of any notice of any matter by an access provider to a service provider.
Put shortly, the equivalency SAOs do not give rise to any obligation on the part of Telstra to give any particular type or period of notice of any Network Upgrades that will affect the provision of the ULLS. To that extent, and in that respect, the undertakings are consistent with the SAOs applicable to Telstra. They do not contain any provision relating to the manner or period of notification of any Network Upgrades which impinge upon, or are inconsistent with, those obligations.
It is necessary to distinguish between the nature, content and effect of any Network Upgrade and the period of notification to be given to a service provider of any such proposed Network Upgrade. The former is covered by the equivalency obligations contained in subss 152AR(3)(b) and (c), the latter is not so covered.
It is of course a separate and further enquiry whether the provisions in the undertakings relating to network modernisation – particularly in relation to the period of notification of any proposed Network Upgrades – are reasonable having regard to the matters specified in s 152AH and the objectives in s 152AB.
In any event, we are satisfied that the provisions of the undertakings, including the Network Modernisation Provisions, are consistent with the SAOs that are applicable to Telstra. There is no provision in the undertakings which inhibits or prevents Telstra from:
·ensuring that the technical and operational quality of the ULLS it supplies (including ordering and provisioning) is equivalent to that which Telstra provides to itself; or
·ensuring that service providers receive, in relation to the ULLS it supplies, fault detection, handling and rectification of a technical and operational quality and timing that is equivalent to that which Telstra provides to itself.
For the reasons given above, we do not accept Optus’ submission that in order for the Network Modernisation Provisions in the undertakings to be consistent with the SAOs, Optus would need to be given notice of all of the details of any network modernisation that would materially affect its operations, and allow it to respond at the same time, and in an equivalent manner, as Telstra becomes aware of such matters. We do not accept that such notification is part of the technical and operational quality of the service supplied by Telstra to itself. The notification provision contained in the undertakings (cl 6.2, par [28] above) is not, in its terms, inconsistent with any of the SAOs.
We also do not accept the submission of the other intervenors that s 152BV(2)(b) requires us to reject an undertaking which contains Network Modernisation Provisions unless we are satisfied that Telstra will, in every instance, exercise the Network Modernisation Provisions in a way which is consistent with the obligation under s 152AR(3)(b). As noted earlier, we do not accept that the notification provision in the Network Modernisation Provisions is covered by s 152AR(3)(b) but we accept that the nature of any Network Upgrade may well fall within the terms of ss 152AR(3)(a) and (b). The submissions of the other intervenors misconstrue the terms of s 152BV(2)(b) which requires us to compare the terms of the undertakings with the SAOs and determine whether there are any inconsistencies between them. It is the nature of the obligation undertaken by Telstra which is relevant, not whether it will in fact perform that obligation on every occasion in the future when it is required to do so. If the terms of the undertakings are consistent with the SAOs then our enquiry under s 152BV(2)(b) is at an end. We are not required to go further and determine, that is, predict, whether Telstra will, in fact, in the future carry out and implement the obligations it has undertaken to perform.
10. AVERAGING OF NETWORK COSTS
One of the principal matters for our determination is whether Telstra’s charge of $30 per service per month for access to the ULLS is reasonable. In determining whether that charge is reasonable, we must have regard to the matters set out in s 152AH and the objectives set out in s 152AB. Whether that charge is reasonable requires us to look at the method by which the charge is derived and to determine whether the method is, in all the circumstances, reasonable. We must therefore assess the method adopted to derive the ULLS charge by reference to the same matters set out in ss 152AH and 152AB.
Telstra’s principal contention in relation to the monthly charge for access to its ULLS was that the charge should be averaged across four bands or areas, namely:
·Band 1 – central business districts (“CBDs”);
·Band 2 – metropolitan;
·Band 3 – provincial;
·Band 4 – rural,
to reflect the retail price parity obligation (“RPPO”) (section 10.2 below) faced by it in relation to line rental services at the retail level, so that it can recover the efficient costs of its investment in its Customer Access Network (“CAN”). Although Telstra puts its contention in this form, the contention is more accurately expressed in the following form – the monthly charge for access to its ULLS should be set on the basis that it recovers no more than the efficient costs of the supply of the ULLS in all geographic areas and that this charge is based on the average of the costs across those four areas to reflect the RPPO faced by Telstra in relation to line rental services at the retail level, so that it can recover the efficient costs of its investment in its CAN.
10.1 The basis of Telstra’s $30 charge
We turn to a consideration of averaging having regard to the particular matters set out in s 152AH and the objectives in s 152AB. Telstra’s proposal is to implement a uniform ULLS monthly access charge across the four geographic areas it has identified, based on the average of its estimated efficient costs of supplying the ULLS in these four areas. When we use the expressions “averaging” or “averaged”, we use them as shorthand to refer to Telstra’s geographic averaging of its costs and to the consequent uniform ULLS monthly access charge across the four areas. Similarly, when we use the expressions “de‑averaging” or “de‑averaged”, we are using them as shorthand to refer to the costs of the supply of the ULLS to each of the four areas and to the practice of charging different amounts for access to the ULLS Telstra supplies in each of the four bands or areas.
The issue of averaging arises because the cost of the ULLS which Telstra supplies varies between the four bands or areas. Telstra has estimated its costs of supplying the ULLS in each of the four bands or areas and says that the cost of supplying the ULLS increases successively from one band to the next, with the cost of supplying an ULLS in Band 1 being the lowest and the cost of supplying an ULLS Band 4 being the greatest.
The material before us showed that there are two separate areas in which averaging could most conveniently be considered. The parties considered the impact of averaging in urban areas (in which they included bands 1 and 2) and in rural areas (in which they included bands 3 and 4). We follow the same path, recognising that Telstra’s cost analysis is broken down into the four bands or areas.
In order to set a uniform charge for the ULLS that just recovers the cost of providing the service across all of these bands, a charge must be set that reflects the average cost of providing the service across the whole of the country. In other words, above‑cost charges must be set for the ULLS in some areas in order to generate sufficient revenues to cross‑subsidise below‑cost charges for the ULLS in other areas. Based on Telstra’s estimates of providing the ULLS in each of the four bands, averaged access charges for the ULLS would be above the cost of providing the service in the CBD and metropolitan bands, but below the cost of providing the service in the provincial and rural bands. Telstra argued that averaged access charges must be set for the ULLS in order to generate, and preserve, a cross‑subsidy from CBD/metropolitan bands to provincial/rural bands.
Telstra has derived its monthly charge of $30 per service on the basis that the monthly charge will recover no more than its costs of supplying the ULLS and a reasonable return of, and return on, capital employed in making the service available. Telstra’s cost estimates are based on the following elements:
·ULLS network costs;
·ULLS specific costs; and
·an adjustment for the amount Telstra receives from the Universal Service Fund (“USF”) as a result of its Universal Service Obligation (“USO”).
Based on a number of inputs and assumptions to which we refer later in these reasons, including using a lower and a higher WACC value, Telstra has estimated the averaged costs of supplying its ULLS during the periods covered by the undertakings as set out in the table below:
Lower
value
1/1/06 -30/6/06
Higher
value
1/1/06 -30/6/06
Lower
value
1/7/06 -
30/6/07Higher
value
1/7/06 -
30/6/07Lower
value
1/7/07 -
30/6/08Higher
value
1/7/07 -
30/6/08Network costs $24.95 $34.97 $27.47 $38.37 $28.62 $40.26 plus specific costs $5.65 $5.99 $3.59 $3.90 $2.67 $2.91 less USO adjustment $0.31 $0.31 $0.29 $0.29 $0.26 $0.26 Total $30.29 $40.65 $30.77 $41.98 $31.03 $42.91 The network costs described in this table have been estimated by reference to Telstra’s PIE II model while the specific costs have been estimated by reference to the SC Model and the SC Calculations. We shall return to these cost estimates later in these reasons.
In the case of the lower‑value cost estimates outlined in par [59] above, the averaged costs may be contrasted with Telstra’s estimates of its efficient lower‑value network costs per ULLS per month for each of the four bands based on its PIE II model:
1/7/05-30/6/06 1/7/06-30/6/07 1/7/07-30/6/08 Band 1 $[X] $[X] $[X] Band 2 $[X] $[X] $[X] Band 3 $[X] $[X] $[X] Band 4 $[X] $[X] $[X] Averaged cost $24.95 $27.47 $28.62
Telstra’s estimates of its efficient higher‑value network costs per ULLS per month for each of the four bands were:
1/7/05-30/6/06 1/7/06-30/6/07 1/7/07-30/6/08 Band 1 $[X] $[X] $[X] Band 2 $[X] $[X] $[X] Band 3 $[X] $[X] $[X] Band 4 $[X] $[X] $[X] Averaged cost $34.97 $38.37 $40.26
Telstra averaged its lower‑value and higher‑value network cost estimates using the following weights:
05/06 06/07 07/08 Band 1 [X]% [X]% [X]% Band 2 [X]% [X]% [X]% Band 3 [X]% [X]% [X]% Band 4 [X]% [X]% [X]%
Telstra accepted that if there was no RPPO imposed on it, fully de‑averaged ULLS charges would be a sensible policy from an economic perspective. It contended that the RPPO, in effect, required it to charge the same price for line rental services at the retail level in all areas. (This contention is an over‑simplication of the content of the RPPO and does not take into account the consequence of bundling line rental services with other services. We return to this issue later in these reasons). According to Telstra, a policy of fully de‑averaged ULLS charges would allow both retail and wholesale prices to reflect costs in accordance with standard economic principles. But, it submitted, where an RPPO existed, de‑averaged ULLS charges encouraged “cream‑skimming” by access seekers with the effect that the costs of its CAN would not be recovered and competitive activity in rural areas would be discouraged. By contrast, averaging of ULLS charges assisted in matching the structure of retail and wholesale prices.
Cream‑skimming, as explained by Telstra, occurs when access seekers take advantage of the arbitrage opportunity that exists if cost‑based charges are set for the ULLS (a key input into the provision of a retail line rental service) in urban areas in circumstances where Telstra needs to set above‑cost prices for retail line rental services in these areas in order to meet its RPPO. If access seekers chose only to operate in urban areas, they would be able to set prices for the retail line rental services they offer that reflect the cost of providing the service in these areas if they were able to acquire the ULLS for de‑averaged charges. However, given such retail prices would be cost‑based, they would necessarily be below the above‑cost prices for the retail line rental services set by Telstra in order to comply with its RPPO. If Telstra chose to respond by lowering its own retail prices towards its cost of production, it would therefore reduce the above‑cost revenue it would receive from providing retail line rental services in urban areas. In turn, this would undermine its ability to fund the cross‑subsidy arrangement it believes is necessary to meet the RPPO in rural areas. Alternatively, Telstra could choose not to respond to competition and retain above‑cost prices for retail line rental services in urban areas. If it did so, it would be expected to lose most (if not all) of its retail line rental customers in urban areas. Telstra argued that under either strategy it would no longer be able to earn sufficient revenues in total to recover the efficient costs of its CAN.
10.2 Telstra’s Retail Price Parity Obligation (RPPO)
The RPPO arises in the following circumstances. Under cl 19A of the Telstra Carrier Charges – Price Control Arrangements, Notification and Disallowance Determination No. 1 of 2005 (as amended by the Telstra Carrier Charges – Price Control Arrangements, Notification and Disallowance Determination No. 1 of 2005 (Amendment No. 1 of 2006)) (“the Price Control Determination”), Telstra must offer “basic line rental services” to residential, charity and business end‑users, in non‑metropolitan areas, at the same or a lower price and on the same price‑related terms it offers to such end‑users in metropolitan areas. Further, Telstra must offer “basic line rental services” within a bundle in all non‑metropolitan areas at the same or lower price and on the same price‑related terms as it offers “basic line rental services” in a comparable bundle in metropolitan areas.
For the purposes of cl 19A:
·“metropolitan area” means the inter‑carrier charge area for Sydney, Melbourne, Brisbane, Perth or Adelaide;
·“non‑metropolitan area” means any area of Australia other than a metropolitan area;
·“price‑related terms” means terms relating to price or a method of ascertaining price;
·“basic line rental service” means a line rental supplied in conjunction with a service supplied to a customer in order to comply with the requirement to provide pre‑selection under a determination made by the Australian Communications and Media Authority under Part 17 of the Telecommunications Act 1997, other than a line rental in respect of which the customer contractually agrees not to exercise the right to pre‑select in favour of a carriage service provider other than Telstra.
(The Price Control Determination is made pursuant to ss 154(1), 155(1) and 157(1) of the Telecommunications (Consumer Protection and Service Standards) Act 1999 (Cth) (“the TCPSS Act”)).
It was accepted that at the present time, and on the date of the Price Control Determination, Telstra’s basic line rental services were those supplied under Telstra’s HomeLine Part and BusinessLine Part services.
The consequence is that Telstra is required to provide its HomeLine Part and BusinessLine Part line rental services in non‑metropolitan areas at the same or a lower price, and on the same price‑related terms, as it offers these services to customers in metropolitan areas. These two services are unbundled services. HomeLine Part service is defined by Telstra in the terms and conditions on which it offers its Basic Telephone Service. The HomeLine Part service is available where a customer does not preselect Telstra for long‑distance calls, international calls and calls to mobile numbers. A subscriber to the HomeLine Part service obtains:
·connection to Telstra’s Public Switched Telephone Network (“PSTN”);
·the ability to make and receive certain types of calls, basically local calls;
·a telephone number;
·a free listing of the telephone number in a telephone directory.
A subscriber to the HomeLine Part service who wishes to make long‑distance calls, international calls or calls to mobile numbers must either preselect a carrier other than Telstra for these calls or use an override code to select Telstra, in which case further charges will be incurred.
The BusinessLine Part service is in similar terms but applies to business end‑users rather than residential.
Telstra also supplies line rental services as a part of bundles that include various calling services (such as fixed‑to‑mobile calls, long distance calls and international calls). However, the RPPO requirement in the Price Control Determination does not apply to these line rental services when they are supplied as part of a bundle of services so long as the line rental service supplied in these bundles is not a “basic line rental service” as described in par [64] above.
10.3 Telstra’s Universal Service Obligation (USO) and the Universal Service Fund (USF)
Underpinning Telstra’s argument for the averaging of charges was its proposition that it needed to subsidise the provision of telecommunications services in rural areas in Australia because of its RPPO. It was accepted that the cost of providing these services in provincial and (especially) rural areas was significantly higher than in urban areas, and that Telstra was required to provide line rental services in provincial and rural areas, and to end‑users, that were unprofitable.
The need to subsidise the provision of telecommunications services in rural areas raises for consideration Telstra’s USO which in turn gives rise to a consideration of what Telstra receives from the USF.
The USO is described in s 9(1) of the TCPSS Act in the following terms:
“For the purposes of this Act, the universal service obligation is the obligation:
(a)to ensure that standard telephone services are reasonably accessible to all people in Australia on an equitable basis, wherever they reside or carry on business; and
(b)to ensure that payphones are reasonably accessible to all people in Australia on an equitable basis, wherever they reside or carry on business; and
(c)to ensure that prescribed carriage services are reasonably accessible to all people in Australia on an equitable basis, wherever they reside or carry on business.”
Telstra is subject to the provisions of the USO by virtue of the provisions of ss 11, 11A and 12D of the TCPSS Act.
The USF arises through the imposition of a levy upon industry participants (telecommunications carriers), including Telstra, under the Telecommunications (Universal Service Levy) Act 1997 (Cth).At the present time, Telstra is the only universal service provider. The levy is assessed, collected, recovered and distributed under Div 13 of the TCPSS Act. The amount of the subsidy to be paid to a universal service provider in relation to its USO is determined by the Minister for Communications, Information Technology and the Arts under Div 9 of the TCPSS Act: s 16. The TCPSS Act does not, however, set out any mechanism or formula by reference to which the subsidy to Telstra is to be determined, nor is the subsidy linked to Telstra’s costs in fulfilling its USO. The Minister may take advice from the Australian Communications and Media Authority. On 21 June 2005, the Minister set the subsidies for the financial years 2005/2006, 2006/2007 and 2007/2008.
Telstra takes into account the amount of the subsidy it receives when calculating its ULLS charge but it contends that the subsidy does not provide for full recovery of the deficit incurred in serving end‑users to whom it provides a service at a loss.
The Commission submitted that Telstra had not provided any evidence to suggest that the USF would be incapable of meeting any cost shortfall that would exist in rural areas for an efficient forward‑looking network. The Commission also submitted that in order to determine whether the ULLS charge proposed by Telstra was reflective of its efficient costs, the USF subsidy should be calculated by reference to the subsidy that Telstra would receive if a forward‑looking network were built. The Commission argued that Telstra’s actual subsidy was irrelevant. We do not accept this line of reasoning. The fact is that the Minister has already set the USF subsidies for the periods covered by the undertakings. The proposition that we have to consider, whether Telstra’s costs underpinning its proposed monthly charges are efficient costs, does not mean that we have to undertake a hypothetical exercise to determine the appropriate amount of the USF subsidy during the periods of the undertakings. The existence of issues in relation to Telstra’s costs, whether they are efficient costs and whether the PIE II model accurately estimates efficient ULLS network costs, does not mean that we should disregard the USF subsidies actually fixed by the Minister in relation to the periods covered by the undertakings.
10.4 The averaging issue
Telstra supported its case for averaging by relying on the report of Professor David Sappington, an eminent expert in the field of economics and regulatory policy in the telecommunications industry.
Professor Sappington’s conclusions, in summary, were that uniform ULLS charges:
·were consistent with, and helped to mitigate the deleterious effects of, the RPPO;
·would limit cream‑skimming in urban areas that undermined the RPPO;
·could enhance competitive activity in rural Australia and help to limit the operation of inefficient suppliers in urban areas;
·could help to support a low uniform charge for an unbundled basic access service throughout Australia; and
·constituted a reasonable policy in Australia as long as Telstra continued to face the RPPO.
Professor Sappington’s analysis was based upon three explicit assumptions:
·Telstra was required to charge the same price for unbundled basic access services to residential (and business) end‑users throughout Australia;
·the price that Telstra charged for unbundled basic access services to residential end‑users could exceed Telstra’s unit production cost in urban areas but often was below Telstra’s unit production cost in rural areas; and
·Telstra’s efficient unit cost of producing the ULLS was generally higher in rural areas of Australia than in urban areas.
It appears that there was a fourth assumption underpinning some of Professor Sappington’s analysis, namely, that uniform ULLS access charges, or averaging, is only effective to mitigate the effects of the RPPO and avoid the consequences of cream‑skimming when new entrants in urban areas, or retail competitors of Telstra in urban areas, are precluded from investing in alternative urban networks, or if it is not technically or commercially feasible for competitors to roll out alternative urban networks.
There was no material placed before us to the effect that bypass, whether efficient or inefficient, by access seekers would not be possible. Indeed there was some evidence to the contrary, namely, that bypass of Telstra’s network was possible. In its response in July 2006 to the Commission’s Position Paper, A Strategic Review of the Regulation of Fixed Network Services, Telstra stated that:
“In these CBDs [the central business districts of Sydney, Melbourne, Brisbane, Adelaide and Perth], there are multiple competing fixed and wireless access networks that provide effective competitive infrastructure to Telstra’s network.”
Further, Optus pointed to the fact that bypass of Telstra’s existing network had already occurred to some extent. In the Commission’s report, Telecommunications Infrastructure in Australia 2004 (June 2005), it was noted that the traditional copper network accounted (in 2004) for 87.2% of subscriber connections and that other technologies (such as Hybrid Fibre‑Coaxial Cable (“HFC”) – 11.86% and optical fibre – 0.6%) accounted for the remainder.
The Commission also noted in its Final Determination on its Declaration Inquiry for the ULLS, PSTN OTA and CLLS (July 2006) that a number of local access networks based on microwave, fixed wireless, optical fibre and satellite technologies had emerged in recent years. It identified the alternative network infrastructure in place in capital cities as of June 2004 and said they were mostly in discrete geographic areas. It observed:
“While these networks may have the technical capability to deliver services that are, to a degree, substitutable for those offered via Telstra’s copper customer access network (CAN), most of these networks are located in discrete geographic areas. For instance, optical fibre networks are mostly located in central business district (CBD) areas and are targeted toward corporate customers. Some fibre‑based and/or HFC networks are also deployed in certain regional areas such as the ACT and some regional cities in Victoria, NSW and other areas. Many of the wireless networks that have been developed recently are targeted at end‑users in regional and remote areas.”
We also note, and take into account, the following observation of the Commission:
“In addition, the mere existence of alternative networks does not necessarily indicate that there is effective competition in particular areas. Effective competition will depend on factors including, but not limited to, the height of barriers to entry, competitors’ wholesale and retail market concentration levels, and the prices and costs of services provided. Hence, it is when conditions for competitive new entry exist and there is evidence of effective competition in an appropriately defined market (or the prospect of this in a clearly defined time‑frame) that removal of regulation should be considered.
Therefore, in addition to the evidence that shows alternatives are fragmented, the Commission has not received any evidence that these alternative networks actually constrain Telstra’s prices and behaviour. Instead, it appears that Telstra generally has a large degree of market power in originating and terminating voice calls, as well as the provision of fixed services more generally.”
The significance of the existence, or possibility, of bypass, whether it be efficient or inefficient, is that Telstra’s rationale for averaged ULLS access charges is undermined if new entrants or retail competitors in urban areas can bypass the Telstra network and avoid having to pay the averaged ULLS charge. To the extent that bypass of the CAN is significant in urban areas, there will be less users of the CAN in urban areas from which Telstra can extract above‑cost revenues through its pricing of retail line rental services. In this situation, Telstra’s ability to subsidise the losses incurred in rural areas with the excess of revenue over cost obtained in urban areas will be lost (or significantly diminished) and the consequence will be similar to that which Telstra contends will occur as a result of cream‑skimming.
Telstra submitted that no compelling case had been made that inefficient bypass would be widespread under averaging or that the risk of such bypass told against averaging. It relied upon the views expressed by CoRE Research that entry into the market, even if subsidised, would not occur unless it was efficient and that entrants would make efficient build versus buy decisions. There is some merit in this submission as there is in the other submissions advanced by Telstra as to why the risk of bypass did not totally undermine its arguments in support of averaging. Nevertheless, as noted earlier, there was some evidence before us that bypass had occurred.
However, the reasons for that bypass, specifically whether they were cost‑driven or demand‑driven, were never articulated. We cannot be confident what the drivers of bypass might be, nor the quantum of bypass, should demand‑side signals change, as would be the case with an averaged access charge which would be higher than hitherto had existed in urban areas.
One of Professor Sappington’s basic propositions was that de‑averaged ULLS charges, coupled with the RPPO, would enable inefficient competitors to thrive in urban areas while precluding the operation of efficient competitors in rural areas.
Professor Sappington noted that, under de‑averaging, the price structure imposed by the RPPO encouraged competitors to service end‑users in urban areas of Australia and discouraged competitive activity in rural areas. He observed that even a competitor with a cost structure similar to Telstra’s could service end‑users profitably in urban areas by charging a lower price than Telstra charged, as Telstra’s price is set at an above‑cost level in urban areas in order to help finance the below‑cost pricing that is implemented in rural areas to comply with the RPPO.
Professor Sappington considered that, in theory, Telstra could lower the price it charged for unbundled basic access services in urban areas to match the price charged by competitors, but under the RPPO, this price reduction would require an identical price reduction in rural areas, which together would reduce Telstra’s earnings unduly. He considered that Telstra might conceivably reduce the prices it charged for bundled telecommunications services in urban areas only, leaving the price of unbundled basic access unchanged throughout Australia. Telstra might then consider raising the prices of bundled services in rural areas to offset the revenue reduction resulting from the lower prices of bundled services in urban areas. However, Professor Sappington noted that rural end‑users could avoid the high prices for Telstra’s bundled services in rural areas by purchasing basic access from Telstra (at the RPPO price) and securing any additional desired services from competitors.
It was critical to Professor Sappington’s analysis that the typical discipline of competitive markets, whereby prices were driven to the level of costs so that prices for retail line rental services would be set at levels that reflected the de‑averaged costs of providing these services, could not influence Telstra’s pricing because of the RPPO. He contended that once Telstra’s prices for basic access services, were precluded from reflecting the de‑averaged costs of producing those services, it was best to set wholesale charges for the ULLS that similarly departed from the de‑averaged costs of producing the wholesale services.
Professor Sappington noted that in the report Averaging vs. De‑averaging, provided to the other intervenors by Marsden Jacob Associates, it was suggested that uniform charges would enable Telstra to achieve a competitive advantage that would allow it to undercut its urban competitors and capture greater market share. This suggestion is not warranted if you accept Professor Sappington’s thesis that averaging of charges is necessary to enable Telstra to subsidise its rural services, having regard to the RPPO. Professor Sappington noted that the RPPO prevented Telstra from reducing the price it charged for unbundled basic line access in urban areas of Australia without implementing the same price reduction in rural areas. Therefore, Professor Sappington argued, Telstra was not at liberty to engage in the selective price cutting in urban areas that Marsden Jacob Associates appeared to envision. This is not necessarily correct. Telstra might not be at liberty to engage in selective price cutting of the retail line rental service which is subject to the RPPO, but it is at liberty to engage in selective price cutting in urban areas in relation to bundled products which are not subject to the RPPO. This proposition also applies to any other line rental service which is not subject to the RPPO.
However, Marsden Jacob Associates might be correct in their suggestion if the position is that Telstra is able, by reference to other services provided over its CAN, to accumulate profits from the provision of those services which enable it to subsidise rural services thereby releasing its urban charges from the need to subsidise rural services.
11. DOES AVERAGING PROMOTE THE LONG‑TERM INTERESTS OF END‑USERS?
The answer to this question requires us (s 152AB(2)) to have regard to the extent to which averaging is likely to result in the achievement of the objective of:
·promoting competition in markets for listed services;
·achieving any‑to‑any connectivity in relation to carriage services that involve communication between end‑users; and
·encouraging the economically efficient use of, and the economically efficient investment in:
othe infrastructure by which listed services are supplied; and
oany other infrastructure by which the listed services are, or are likely to become, capable of being supplied.
We turn first to the objective of promoting competition in markets for listed services.
11.1 The meaning of promoting competition in s 152AB(2) of the Act
The ULLS can be used to provide a range of telecommunications services, including fixed-line voice services (such as the provision of line rental, local call and long distance call services) and broadband internet services. We consider the markets for these services to be the relevant markets within which to determine whether averaging is likely to promote competition.
Telstra argued that averaging was likely to promote competition in urban areas as de‑averaged charges for the ULLS, in the face of the RPPO, would prevent Telstra from competing with access seekers in urban areas, even if it was equally as efficient as, or more efficient than, access seekers, because of the existence of cream‑skimming.
Telstra also argued that averaging would be likely to promote competition in rural areas because access seekers could obtain access at charges less than cost and so compete in an area otherwise not open to them. If charges were de‑averaged, competition would be unlikely to occur in rural areas. Even an efficient access seeker would be unable to compete as retail prices would be held below cost by Telstra’s obligation to meet its RPPO. Put shortly, Telstra contended that de‑averaged charges encouraged inefficient competition in low cost urban areas and inhibited or prevented efficient competition in high cost rural areas. Averaged charges avoided these consequences because, as long as Telstra must meet its RPPO, they were the only means for encouraging efficient infrastructure‑based competition in rural areas.
It is important to remember that in this context we are considering the likelihood of the promotion of “competition”, not the promotion of competitors. We are concerned with the process of competition and whether effective competition is likely to be promoted. We are not concerned to determine whether particular market outcomes will be achieved by averaging.
In addition, no attempt was made by the parties to inform us as to whether the use of one WACC or two made any significant quantitative difference to the overall return that Telstra could earn. Given the relative imbalance in the sizes of the two groups of assets, the net effect on Telstra’s earnings of choosing one approach over the other may in effect have been of negligible practical proportions, depending on the actual asset and WACC values that are brought into the estimation.
Generally, we found the level of assertion by both the Commission and Telstra on this issue unhelpful. We are unable to be satisfied, having regard to the matters set out in s 152AH and the objectives in s 152AB, that the claim by Telstra for two WACCs, one for the assets in the CAN and one for the ULLS‑specific assets, is reasonable.
23.2 A correction for WACC asymmetry
Telstra submitted that the social consequences of getting the estimate of the WACC wrong were not symmetrical. An overestimate of the true WACC, it said, produced far less serious long run social damage than did an underestimate of that value.
Telstra argued that should a regulator set a WACC that was too high relative to its true value (we note that such a value can never be ascertained), end‑users might suffer in the short run, but the excessive return would attract new entrants and this competition would restore prices to an acceptable level. That is, a WACC set too high would be self‑correcting over time.
In contrast, a WACC that was set at commercially unrealistically low levels would lead to firms delaying new investments or possibly exiting the market. Under this scenario, the quality of service would deteriorate, and ultimately the service might not be offered at all. Telstra did not see any likely self‑correcting mechanism being possible under this scenario.
In putting its case that an allowance needed to be made for the asymmetric consequences of error in estimating the WACC, Telstra again relied heavily on the reports of Professor Bowman.
The Commission, Optus and the other intervenors provided little by way of expert analysis and reports to counter the propositions put forward by Professor Bowman. Rather, they attacked the premises that formed the foundation of his estimates, arguing that the process adopted by him in deriving his recommended WACC values could not produce reasonable parameter estimates.
The only direct counter to Professor Bowman put into evidence was the critique by Mr Jason Ockerby, the head of regulatory economics at Optus. He made several criticisms of Professor Bowman’s reports, which Telstra submitted had been addressed by Professor Bowman. However, Mr Ockerby’s critique with respect to the asymmetrical effects of underestimating and overestimating the WACC were not dealt with in Professor Bowman’s responses to Mr Ockerby.
This issue comprised two separate enquiries: are there asymmetrical effects caused by estimating the WACC either too high or too low and, if so, should an allowance be made for these errors, and to what numerical extent?
23.3 Are there asymmetrical effects in estimating the WACC?
Optus submitted that in order for an uplift to the WACC to take account of asymmetrical effects to be justified, it had to be shown that the capital supplied in response to a change in the WACC must be asymmetric around the optimal level of the WACC (less being supplied when the WACC was set too low and more when it was too high), and also that the economic loss associated with the underinvested capital must be shown to exceed the economic loss associated with an equivalent overinvestment of capital.
Telstra accepted that there would be deadweight losses both in cases where the WACC is set too high and where it is set too low, but argued that this acceptance did not imply that the impacts associated with overestimation or underestimation of the WACC were symmetric, and that the impacts depended on the shape of the demand and cost functions. It contended that the deadweight loss associated with setting the WACC slightly too high was likely to be small compared to the deadweight loss from the contraction of output caused by the refusal to invest that might be prompted by setting a WACC that was below the actual cost of capital.
Telstra did not provide, other than by way of assertion, any support for the proposition that a WACC that was set below the true rate would lead to a decline in investment in the relevant assets. Whether such a result would occur would depend on the size of the error, as well as on the distribution of investors’ perceptions of the true WACC value. It provided no evidence in respect of these matters and therefore we are not able to be satisfied of the extent of the consequences that it asserted would follow in the case of such estimating errors.
The Commission and Optus challenged Telstra’s claim that there was an asymmetry in the social consequences in mis‑estimating the WACC for the purposes of calculating the ULLS network costs and specific costs on the basis that the perceived return required by investors was uncertain and was distributed over a range, and that any reduction in output would be limited because investors would have different perceptions of the WACC. Telstra responded that the fact that investors had a distribution of views as to the correct WACC did not invalidate the conclusion that there would be less investment if the WACC were set too low. If prices were set below long‑run efficient costs, investment by the access provider would fall below the efficient level because in a system based on voluntary supply, firms only supply when the expected price is at least equal to the expected cost.
Telstra provided no evidence or analysis of the economic losses associated with a too‑high WACC or a too‑low WACC to satisfy us that an upwardly‑biased WACC would be reasonable. It also provided no evidence to support the likelihood of the outcome said to follow from a too‑low WACC value, nor of the time period over which such a socially damaging outcome might occur.
With regard to circumstances where a WACC was set too high, it is not immediately obvious to us that new entry and competition would be expected to occur so as to restore prices to acceptable levels. While competitive markets might be expected to ensure above‑cost prices were not sustainable as a result of the prospect of new entry, regulated services are not provided in competitive markets. A key reason for regulating the provision of particular services is that they are provided in markets that tend to be in the nature of a natural monopoly. Prices need to be regulated in these markets because natural monopoly providers of services are less likely to be constrained by the threat of competitive entry if they seek to raise their prices above those levels necessary to recover their costs. Accordingly, given the cost structure likely to exist in relation to the provision of the ULLS (high fixed costs, low variable costs), it is not likely that if the regulator were to set too high a WACC that other firms would be able easily to enter such a market and restore prices to acceptable levels given the existence of an incumbent supplier of network services in this market. It is also important to realise that even if a market might remedy the situation caused by a too‑high WACC, a regulator may also be capable of recognising in its next determination that it had erred if it set a WACC too low, and accordingly could allow a higher WACC value for the next regulatory period, or even allow for some recovery of past unrecognised or under‑recovered costs.
Optus contended that as the undertakings were for a short term of three years, it was unreasonable to accept Professor Bowman’s references to the long‑term consequences of any underestimation or overestimation of the WACC. In a similar vein, the Commission contended that as the undertakings were short‑term instruments, the WACC could be reset thereafter.
Telstra rejected these submissions, arguing that although the undertakings were for a relatively short term, Telstra and access seekers made investment decisions that had long‑term implications and consequences and that to ignore the long‑term nature of the ULLS network business would be a mistake and inconsistent with the statutory criteria.
The Commission submitted that there was no clear empirical evidence of the asymmetries contemplated by Professor Bowman as a matter of fact, such that one or the other was relatively more likely to promote the long‑term interests of end‑users.
We accept that it is possible that there may be asymmetric consequences associated with setting a WACC too high or too low. However, it is not clear to us that the asymmetry would always imply that overestimation of the WACC led to a lesser social cost than underestimation of the WACC. The nature of the asymmetric consequences of incorrectly setting a WACC is likely to depend on the circumstances of a given matter that may be before the Tribunal. Telstra and Professor Bowman submitted that the long‑term social costs of underestimating the WACC would be greater than the long-term social costs of overestimating it in this particular instance, largely because in circumstances where the WACC was set too low, there was a risk that this would lead to the cessation of services, or a failure to develop services at a socially desirable rate. In order to convince us of this submission, however, it was incumbent upon Telstra to provide evidence that these circumstances actually existed or would exist in relation to the ULLS. Professor Bowman assumed that they did, but he did not provide any evidence or support for the proposition that this was, or would be, the case.
Telstra assumed that setting a WACC that was too low would deter investors. However, different investors will inevitably have different attitudes to risk. Setting the WACC below the true value may deter some investors and therefore result in less investment taking place in the short run, but it will not be likely to cause all investors to cease providing funds. Of course, the service provider might be forced to cut back on maintenance or service quality if it perceived the return on these investments to be too low, but no evidence was advanced by Telstra that consumers’ valuations of different levels of quality was asymmetric. It is possible, at least in theory, that consumers might value lower quality, or less innovation, that might follow from less than efficient levels of investment no differently than they value the surplus lost from greater‑than‑efficient quality, or wasteful innovation, that could arise from too much investment.
The Commission argued that a firm faced with a WACC that had been set too low might choose to improve its efficiency rather than allow its service quality to downgrade. This would effectively be an incentive‑based response of the sort claimed by Telstra to come into force if the WACC value were to be set too high.
Telstra rejected this proposition, arguing that it was misconceived and lacked any academic or empirical support. It also argued that setting the WACC slightly too low could not be self‑correcting as doing so deterred investment in competing facilities by both the access provider and potentially efficient access seekers.
Much of the capital invested in the ULLS network is sunk. Accordingly, Mr Ockerby argued that the effect of an additional return on capital in the regulated WACC would not affect the decision whether to invest, because this decision is irreversible, and that the only relevant risk to investment would be that which related to maintaining and incrementally expanding the existing network. From this he concluded that Telstra would not fail to maintain its CAN asset base because of a small error in the WACC, and that to do otherwise would put at risk billions of dollars of revenue.
In other words, if Telstra were to fail to reinvest in the CAN, it would suffer a large foregone revenue opportunity, a high opportunity cost of its failure to act. For Professor Bowman’s analysis to be correct, it required the assumption that Telstra’s costs of not investing due to a WACC that was set too low were zero. In Mr Ockerby’s words, “in reality, the costs of not investing almost certainly exceed the costs of investing”.
Once these costs of not investing are recognised, the asymmetric costs of regulatory error could well be reversed – setting the WACC too high would likely lead to greater social costs through higher prices for consumers, lost consumer surplus, and overinvestment, while the social costs of a too‑low WACC would not be of major proportions.
It was also pointed out by Mr Ockerby that the WACC allowed for in the PIE II model was based on a hypothetical network investment that was continually re‑estimated by such a forward‑looking TSLRIC‑based model. Thus, the hypothetical asset base to which the WACC was applied was independent of what Telstra actually invests. No evidence was offered by Telstra that the WACC value it sought would in fact directly lead to actual dollar‑for‑dollar new investment by it in the network used to provide its ULLS. Further, this actual expenditure by Telstra would represent only a very small proportion of that which was modelled by a forward‑looking model like PIE II, as a large amount of Telstra’s ULLS network had already been sunk.
We accept that there exists as a matter of theory the potential for asymmetrical consequences should the WACC be set too low or too high. Which of these consequences will carry with it the greatest social damage is not a matter solely for theory, however, but for robust empirical examination, well‑guided by theory, of the actual facts of any particular case. In the present matter, based on the material before us, we do not consider that it is possible to form a concluded view as to where, on balance, the greatest long run social damage would lie, when considering the impact of a WACC set too high compared with the impact of one set too low.
23.4 Allowing for the error in estimating the WACC
Even if we were minded to accept Telstra’s arguments that the social costs of underestimating the WACC were so high relative to those of overestimating the WACC that an upward lift in the estimated WACC should be made to compensate for this error, the question would remain as to the method by which this compensatory adjustment should be estimated, and the way in which its quantum should be determined.
As a matter of principle, we do not necessarily disagree with the arguments put forward by Professor Bowman. As he acknowledged, “the ‘true’ WACC is not known; it can only be estimated on the basis of information available”. However, he went on to state that:
“In my opinion, all regulatory WACCs should be determined with reference to the error involved in estimating the parameters and hence the WACC. Further, the regulatory WACCs should be set above the best estimates of WACC to reflect the asymmetry of the social consequences of errors in setting WACC. This should be done as a matter of principle.”
(Emphasis added)
If the statistical estimating procedure is unbiased, then any resulting best estimate will be either right or wrong. There is no way to tell which outcome has occurred, and whether the estimate is too low or too high, as the true parameter is unknown. But the nature of the estimating process is that, on average, over the long run, the process will deliver the true value of the variable.
The problem, as we see it, is that little firm and testable evidence was provided to us as to whether the Commission has in the past set the WACC too high or too low. Importantly, we also lack convincing evidence as to how the principle of adjustment for asymmetry advocated by Professor Bowman should be applied. Professor Bowman was refreshingly candid on this latter issue. Thus, he said:
“164In my opinion the best approach for the ACCC would be to first determine statistically valid ranges for the parameters considered in estimating WACC. In my view the range interval should approximate one standard deviation of the distribution. Although it would generally be necessary to make informed judgments as to the one standard deviation ranges, rather than precise measurements, the objective of the ranges should be clear. The ACCC could then simulate the likely one standard deviation range on WACC based on these parameters.
165The process of estimating WACC is full of estimation and uncertainty at every single step, including the very foundational principles and precepts. At least in principle, every parameter could have a distribution. There are also issues of uncertainty and estimation with respect to the CAPM [capital asset pricing model] and WACC models that are inducing estimation error. An additional allowance in WACC could be made for the models themselves.” (Emphasis added)
No further guidance was forthcoming from Professor Bowman that was of assistance to us in determining whether the particular approach advocated by Telstra to allow for asymmetric risks was reasonable. Professor Bowman admitted that he had not fully developed and defended the ranges for each of the parameters discussed in his report, and that what he had presented in his report were a discussion of all of the parameters and his “preliminary estimates of appropriate ranges to reflect one standard deviation.” (Emphasis added). It was not surprising therefore that the Commission submitted that his range of parameter estimates were “undeveloped, subjective, and … unjustified”.
Further, Professor Bowman stated that:
“Determining the range for my WACC estimates from the ranges for individual parameters has some complexities. Ideally the parameters and the ranges would be modeled using Monte Carlo simulations. I have not conducted those simulations with … [my set of estimates] of parameter values and ranges.”
(Emphasis added)
The Commission submitted that its Monte Carlo analysis gave weight to a range of opinions (including Professor Bowman’s) regarding various WACC parameter values, whereas Professor Bowman’s method was highly subjective and assumed implicitly that he had correctly estimated the mean values of the component WACC parameters.
The Commission presented to us the outcomes of one set of its Monte Carlo simulations (calculated for a different set of undertakings lodged on 13 December 2004) and showed that all of Professor Bowman’s WACC estimates were above the upper bound of the 95% confidence interval for the true WACC value, and, tellingly, his best WACC estimates with a one standard deviation uplift were substantially higher than this upper bound figure. While the Commission’s analysis was for a different set of undertakings, it submitted that most of the input parameters would have changed little, if at all, and so the fact that Professor Bowman’s current set of estimates were so high brought into serious question the reliability of his estimates.
Telstra accepted that “determination of an uplift from the estimate of WACC is a matter of judgement”, but went on to say:
“However, the application of this judgment is not arbitrary. Rather, the basis for Professor Bowman’s recommendations is that it will avoid the decision being arbitrary. Professor Bowman notes that ‘setting ranges to one standard deviation permits statements to be made about the confidence level of WACC.’ Furthermore, Professor Bowman states that adopting a one‑standard deviation range provides ‘83% confidence that the WACC was not going to lead to the adverse social consequences of economic inefficiency such as under provision of service, maintenance and investment.’ Placing the process in a statistical context, such as Professor Bowman suggests, provides structure and transparency and allows statements about the confidence that can be placed upon the resulting WACC”.
(Telstra’s emphasis)
We note that neither Professor Bowman nor any other evidence presented by Telstra provided any empirical economic analysis that demonstrated the economic loss associated with a WACC that was set too low, or, that the future loss expected to be associated with a too‑low WACC would in fact be ameliorated by a WACC that was uplifted by one standard deviation beyond the best estimate determined by the Commission.
As a matter of statistical reality, the estimate of any parameter is subject to estimating error. But these errors may not always be present in estimating each component of the WACC, and the degree of error may differ greatly as between the estimates of each separate cost component. Optus submitted that a blanket one‑standard deviation uplift to allow for the individual error involved in calculating each component parameter was arbitrary, in the absence of convincing evidence that an all‑inclusive summary error calculation was justified. We agree with this submission. In our view, it would have been reasonable to consider each parameter estimate on its own, and argue for and adjust for any estimation error accordingly.
Professor Bowman based his assessment of asymmetry in terms of long‑run consequences. As the Commission pointed out, such an asymmetry was not been proved as a fact by Professor Bowman, and the undertakings under consideration are short‑term instruments after the expiry of which the WACC will be reset. A period as offered in the undertakings would not, in most circumstances, and certainly not for a long‑lived asset like the CAN, be normally considered as representing the long run. Given an unbiased estimating procedure by the regulator, the WACC will on average be correctly estimated over time. In contrast, if the best point estimate of the WACC is systematically increased by some arbitrary amount like one standard deviation, then in the long run cost over‑recovery will occur with all of its negative social consequences.
In this sense we regard an estimate of the true WACC value, if it has been arrived at through a statistically‑unbiased estimating process, as representing a figure that, on average, in the long-run probabilistic sense in which all such estimates should be considered, would yield the true expected value of the variable in question. To add an amount artificially to such an estimate would in this correct statistical sense result in too high an estimate of the true average of the variable in question, in this case the WACC. We are not satisfied that such a procedure is, in any statistical or economic sense, reasonable in the present circumstances.
23.5 Conclusion on a correction for WACC asymmetry
The guarded language used by Professor Bowman suggests to us that caution is necessary in determining whether an allowance should be made for the asserted asymmetric consequences of estimation error in the course of measuring the WACC. Professor Bowman’s estimated WACC values were the product of many caveats and qualified assessments, and he essentially conceded that his work was what might be colloquially called an educated guess that was not the product of a rigorous statistical inquiry and measurement process, albeit his estimates being informed from many years of relevant experience. He performed no Monte Carlo that would provide us with some sort of statistical comfort that his best estimate of the WACC was statistically defensible.
Clearly, in any statistical estimating process for an unknown parameter, errors will exist. This applies for each of the many components of the WACC. In the absence of compelling economic and statistical evidence, we do not consider that it is reasonable to account for such errors using a single (and arbitrary) error calculation. In our opinion, the reasonable approach would be to consider the error involved in estimating each of the individual component parameters and from this derive a reasonable estimate of the range in which the true overall WACC value would be accepted to lie.
While Professor Bowman’s indicated preference for adding one standard deviation to the best estimate of the WACC could be said by some to be conservative in allowing for estimation error and the asymmetric consequences of such error, Professor Bowman provided no empirical justification or support for his preferred approach, nor did he cite any literature that confirmed the acceptability or reasonableness of his advocacy for setting the WACC at a level one standard deviation above the Commission’s best point estimate. To that extent we regard his report as lacking the convincing demonstration necessary to satisfy us that his method was reasonable in the current circumstances.
This is not to say that we reject the possibility that the approach advocated by Professor Bowman might be able usefully to inform the Tribunal in the future on the procedures necessary to provide a commercially, and socially, realistic estimated WACC value. But a more robust demonstration in terms of empirical justification and acceptability both commercially and in terms of rigorous academic support would be necessary. In the present matter, such a demonstration has not occurred and therefore we cannot be satisfied that the process advocated by Telstra to compensate for asymmetric errors (if they do in fact exist) in estimating WACC values is reasonable.
Because we have found against Telstra on issues of principle with respect to the two disputed matters in relation to the WACC, it is unnecessary for us to consider the submissions relating to the measurement of the values of each of the specific components of the WACC. Before doing this, we would need to have been satisfied at a higher level that the overriding principles used to arrive at these empirical estimates were reasonable, and that the process was therefore capable of producing commercially reasonable values for the WACC. We have not been able to reach that level of satisfaction.
24. CONCLUSION
We have found and determined, having regard to the matters set out in s 152AH and the objectives in s 152AB, that we are not satisfied that:
·the averaged charge for access to the ULLS of $30 per line per month in the undertakings is reasonable;
·the Network Modernisation Provisions in the undertakings are reasonable;
·Telstra’s estimated network costs set out in par [59] represent a reasonable estimate of its projected network costs for the periods covered by the undertakings;
·Telstra’s allocation of its ULLS specific costs across ULLS lines only is reasonable with the consequence that we are not satisfied that Telstra’s estimates of its ULLS specific costs set out in par [59] are reasonable;
·the use by Telstra of two WACCs, one for the assets in the CAN, and one for the ULLS specific assets, is reasonable; and
·the calculation and estimate of the WACC by the addition of an amount equal to one standard deviation in order to allow for the alleged asymmetry and the social consequences of errors made in estimating the WACC, for the purpose of determining and estimating Telstra’s specific ULLS costs, is reasonable.
It follows, therefore, that we are not satisfied that the terms and conditions specified in the undertakings in the respects to which we have referred are reasonable. The consequence is that we must not accept the undertakings. The decision of the Commission on 25 August 2006 rejecting the two undertakings given to it by Telstra on 23 December 2005 is affirmed.
I certify that the preceding four hundred and seventy‑five (475) numbered paragraphs are a true copy of the Decision herein of Justice Goldberg (President), Mr R Davey & Professor D Round. Associate:
Dated: 17 May 2007
Counsel for Telstra Corporation Ltd: Mr A Archibald QC, Dr J Griffiths S.C. and Mr M Connock Solicitor for Telstra Corporation Ltd: Mallesons Stephen Jacques Counsel for the Australian Competition and Consumer Commission: Ms M Sloss S.C. and Mr A McLelland Solicitor for the Australian Competition and Consumer Commission: DLA Phillips Fox Counsel for Optus Networks Pty Limited: Mr S Gageler S.C. and Mr S Free Solicitor for Optus Networks Pty Limited: Clayton Utz Counsel for Agile Pty Ltd, Chime Communications Pty Ltd, Primus Telecommunications Pty Ltd, Macquarie Telecom Pty Ltd and PowerTel Limited: Mr M Nicholls Solicitor for Agile Pty Ltd, Chime Communications Pty Ltd, Primus Telecommunications Pty Ltd, Macquarie Telecom Pty Ltd and PowerTel Limited: Nicholls Legal Counsel for the Commonwealth of Australia: Mr A Southall S.C. Solicitor for the Commonwealth of Australia: Australian Government Solicitors Date of Hearing: 4-8, 11-14 December 2006 Date of Decision: 17 May 2007 GLOSSARY AND ABBREVIATIONS
ADSL Asymmetric digital subscriber line
CAN Customer Access Network
CAPM Capital asset pricing model
CBD Central business district
CMUX Customer multiplexer
COLR Carrier of last resort
DSLAM Digital subscriber line access multiplexer
FTTN Fibre‑to‑the‑node
HSDPA High Speed Down Link Packet Access
IEN Inter‑Exchange Network
IES Information Equivalence Strategy
IRIM Integrated – Remote Integrated Multiplexer (RIM)
LCS Local Carriage Service
LSS Line Sharing Service
O&M Operational and Maintenance
OSP Operational Separation Plan
OTA Originating and Terminating access
POI Point of Interconnection
PSTN Public switched telephone network
PSTN OTA PSTN originating and terminating access
RAF Regulatory Accounting Framework
RKR Record Keeping Rules
RIM Remote Integrated Multiplexer
RPPO Retail price parity obligation
RSS Remote Subscriber Stage
RSU Remote Subscriber Unit
SAO Standard access obligation
SIO Services in operation
SSS Spectrum Sharing Service
STS Standard Telephone Services
TCAM Telstra Customer Access Module
TELRIC Total Element Long‑Run Incremental Cost
TSLRIC Total Service Long‑Run Incremental Cost
ULLS Unconditioned Local Loop Service
USF Universal Service Fund
USO Universal Service Obligation
WACC Weighted average cost of capital
WLR Wholesale line rental
7
3
0