Information Paper. The Split Cost of Capital Concept

Transcription

1 Information Paper The Split Cost of Capital Concept February 2014

2 We wish to acknowledge the contribution of the following staff to this report: Michael S. Blake, Ralph Donnet, John Fallon, Dan Kelley and Kian Nam Loke We are also grateful for comments received from Professors Simon Cowan, Dieter Helm, Stephen King, and Martin Lally. They are not responsible for any remaining errors. Queensland Competition Authority 2014 The Queensland Competition Authority supports and encourages the dissemination and exchange of information. However, copyright protects this document. The Queensland Competition Authority has no objection to this material being reproduced, made available online or electronically ver 3 but only if it is recognised as the owner of the copyright 1 and this material remains unaltered. Draft as at 04/02/14 8:45

3 Table of contents Table of Contents EXECUTIVE SUMMARY Splitting the cost of capital The regulatory problem Current regulatory arrangements Issues and responses Application of the split cost of capital concept Conclusion and recommendations THE ROLE OF THE QCA TASK, TIMING AND CONTACTS I i ii ii iii v vi VII 1 INTRODUCTION 1 2 CURRENT REGULATORY ARRANGEMENTS The regulatory problem Price regulation of infrastructure businesses in Australia Key features Criticisms of the current approach 8 3 IMPLICATIONS OF THE SINGLE WACC APPROACH Investment signalling Financial engineering Evidence of above normal returns 11 4 THE HELM MODEL The split cost of capital The regulatory asset base (RAB) The operating and capacity expansion activities Other elements of the Helm Model 23 5 RELEVANT REGULATORY EXPERIENCE Australia Other jurisdictions 27 6 SPECIFIC ISSUES AND RESPONSES Regulatory risk Implementation issues Other issues Competitive neutrality Submissions 44 7 APPLICATION OF THE SPLIT COST OF CAPITAL Introduction Assumptions Price and overall rate of return results ver 6 i Draft as at 04/02/14 8:45

4 Table of contents 7.4 Summary 52 8 CONCLUSION AND RECOMMENDATIONS Conclusion Recommendations 54 APPENDIX A : OVERVIEW OF THE DORC METHODOLOGY 56 APPENDIX B : SUBMISSIONS 58 APPENDIX C : DETAILS OF THE MODELLING 59 Modelling 59 Annual Revenue Requirement 61 Final Price Determination 62 GLOSSARY OF ACRONYMS, TERMS AND CONDITIONS 63 REFERENCES ver 6 ii Draft as at 04/02/14 8:45

5 Executive summary EXECUTIVE SUMMARY The split cost of capital concept recognises that different components of investment may have materially different risks. A single weighted average cost of capital (WACC) potentially overcompensates investors for the risks associated with the regulatory asset base (RAB). At the same time a single WACC potentially undercompensates investors for major new capital projects. Separate returns for the RAB and major new capital projects could lead to a more efficient allocation of risks among investors, customers, and tax payers. In a series of papers and commentary articles Professor Dieter Helm (Oxford University) has explored the benefits of splitting the cost of capital. This paper presents the arguments for and against the split cost of capital. The most difficult issue for advancing the split cost of capital concept relates to implementation. In particular, agreement must be reached on the values of specific parameters. Until implementation issues are resolved it would be premature to adopt the concept. However, the split cost of capital is considered to be a useful tool for better understanding the amount, allocation and pricing of risk. Therefore, the Queensland Competition Authority (QCA) proposes to conduct further investigation of the split cost of capital concept in the course of ongoing price determinations. This will enable the assessment of relevant risks and development of reference point estimates for parameter values. Splitting the cost of capital The RAB simply represents the value of past investments that the firm s investors must recover from current and future users. Management activities cannot change the financial value of the RAB. Although management actions may reduce its service potential, this aspect can be addressed in the allowances for operating costs. Residual political and regulatory risks are effectively transferred to customers (and tax payers) by the regulator s duty to finance the firm s functions. The regulatory arrangements in effect provide a form of guarantee that the value of the RAB will be protected or preserved in real financial terms. Furthermore the nature of demand conditions for regulated infrastructure businesses and the nature of the regulatory arrangements collectively provide strong assurance that the realised rate of return will match or exceed the allowed or regulated rate of return. Accordingly, the appropriate compensation for investors in the RAB should reflect its low risk. By contrast expenditures on major new capital projects require active management of risks related to construction, engineering, and cost and schedule variability. In addition, major new projects may be subject to greater demand risk than existing infrastructure. Potential undercompensation can occur because the regulatory cost of capital is an average, while investors can require a significantly higher WACC for certain large investment expansion projects. Averaging the returns for existing assets and new investment does not provide accurate price signals to the regulated firm and can lead to less than optimal new investment. At the same time, averaging of returns can result in excess returns for the existing RAB that in effect transfers wealth from consumers to equity holders. A separate, higher WACC applied to major capital expenditure would align investment returns with risk. The higher rate would apply until the risk of new investment matches that of the regulatory asset base. To the extent that the RAB is guaranteed, a regulated return approaching the cost of debt would be appropriate. In addition, according to Helm, setting a regulatory cost of debt ex ante at the start of the ver 6 i Draft as at 04/02/14 8:45

6 Executive summary regulatory cycle is inefficient. The regulated firm s cost of debt is largely beyond its control. Therefore, the cost of debt should be indexed on an annual basis through the regulatory period. The regulatory problem The appropriate rate of return for a business or an investment is fundamental to the regulation of businesses with market power. In order to develop an appropriate methodology for the determination of a rate of return for a business with market power, the starting point is to specify the problem that is being addressed. The conventional economic approach starts with identifying the nature of the relevant market failure or failures. It then considers whether and how best to address the market failure of concern. There are essentially three problems to address in designing and implementing regulation to address market power: (1) Market power may lead to prices that exceed the cost to society of providing the product, leading to excess profits and a lower level of production than is optimal. (2) The nature of regulation may change over time. This is the 'time inconsistency problem' and refers to the perception that the regulatory contract is not secure and may lead to ex post expropriation of investment capital, which in turn acts as a deterrent to invest in the first place. (3) The form of regulation may introduce incentives for the firm to operate and invest inefficiently. Regulators must address each of these issues to achieve the best outcome for society. It is important to recognise that regulators face significant information problems and trade-offs in doing so. The split cost of capital concept has implications for each of these three problems, but the concept is in effect intended to establish more credible, and hence sustainable, regulatory arrangements. Providing greater protection to sunk investments for large infrastructure projects while also gaining the support of customers provides a means to more effectively address the 'time inconsistency' problem. Current regulatory arrangements In Australia, regulators typically apply the building blocks model to compensate the regulated firm for its efficient costs, including a risk-adjusted cost of capital. The return on capital provided to investors compensates them for the time value of money and the risk associated with their investment in the initial (i.e. existing) assets and in current capital projects. In providing this compensation, a single, risk-adjusted cost of capital is applied to estimate the allowed return. The regulatory arrangements for many regulated infrastructure businesses in Australia, in effect, provide strong assurance that the RAB value will be protected and the allowed rate of return will match or exceed the allowed rate of return specified in the regulatory arrangements. There is evidence from the United Kingdom and Australia that the returns allowed by regulators for the purposes of setting prices or target revenues exceed the returns required to ensure efficient investment. Key evidence is the extent to which the market value of a regulated entity exceeds the value of the regulatory asset base, after allowing for the potential to realise efficiencies. Premia reflecting this excess value of per cent have been observed for 29 regulated airport, energy and water businesses in the United Kingdom (CEPA 2013). In Australia, premia averaging around 20 per cent have been documented for seven regulated energy and water businesses (Barry 2013). Thus there is a reasonable concern, based on both a first principles analysis and empirical evidence, that the current form of regulation for many regulated businesses in Australia overcompensates for the actual risk that investors face ver 6 ii Draft as at 04/02/14 8:45

7 Executive summary Issues and responses While criticisms of the split cost of capital concept have been raised, customers of infrastructure businesses have provided support for the concept. Among the points of criticism (and the responses by the QCA to them) are the following: (a) (b) (c) (d) (e) Criticism: Adopting the split cost of capital concept could lead to a substantially lower average allowed cost of capital (compared with the current single cost of capital concept). The result would be a realisation of regulatory or sovereign risk. Response: Rational investors will recognise that they do not have a right to expect above normal risk-adjusted returns to persist indefinitely. Therefore, reforming the regulatory arrangements will not necessarily reduce investor willingness to make future investments as long as the allowed rate of return is set at a reasonable level going forward. Moreover, investors may perceive the new arrangements as being more credible and sustainable and see the change as reducing future regulatory risk (Barry 2013, p. 31). Indeed, if substantial new investment in new infrastructure is required, the split cost of capital approach may increase investor returns. Finally, even if adoption of the split cost of capital concept increases regulatory risk perceived by investors, the costs must be assessed against the benefits of the split cost of capital approach in terms of sending correct investment signals and protecting consumers from excessive prices. If appropriate to address fairness concerns, a compromise is to grandfather existing debt and equity (Helm 2011, p. 19). Criticism: The RAB is not protected by law so a lower return is not justified. Response: The RAB might not be explicitly or fully guaranteed but in practice it is reasonable to conclude that the RAB is a very low risk asset in many regulatory contexts. This then conditions the determination of an appropriate return for the RAB. Economic fundamentals may also mean that some essential infrastructure faces very low demand risk. In addition, the implementation of the split cost of capital would represent a more explicit regulatory contract for protecting the value of the RAB. Criticism: There is no regulatory precedent for the split cost of capital. Response: Some aspects of the split cost of capital have in effect been implemented in Australia, the United Kingdom and the United States, such as protection of the RAB either implicitly (acceptance of line in the sand valuations) or explicitly (e.g. for rail in the UK). However, more importantly, if the concept has a sound economic rationale, and implementation and transitional issues can be reasonably addressed, then the concept can be justified. Criticism: There is no clear benchmark for implementing the split cost of capital. Response: Although there is no theoretical constraint to establishing separate WACCs for the RAB and new investment, there are implementation issues in defining appropriate WACCs for different assets. However, as demonstrated in Chapter 7, there is scope to use market benchmarks for gearing and to make conservative assumptions about the cost of debt and other parameters when implementing the split cost of capital. There is a risk of regulatory error in implementation but this risk should be assessed against the risk that the current arrangements have entrenched a material regulatory error. Criticism: Helm s proposal fails to consider the time horizon over which capital expenditure on expansion assets should earn a higher rate of return before it is rolled into the RAB. Response: Both the level of the return and the time horizon over which it is applied must be determined if the split cost of capital approach is adopted. Regulators must be prepared to authorise higher returns to some new projects than are applied under the current WACC approach ver 6 iii Draft as at 04/02/14 8:45

8 Executive summary for some period of time after project completion. Competitive tendering for expansion projects is one possibility for dealing with this difficult issue. Bidders would reflect project risk in their bids, which would be passed through to the RAB. However, this would not address demand risk for a new project. (f) (g) (h) (i) (j) (k) Criticism: The RAB cannot be financed with 100 per cent debt. Response: Implementation of the split cost of capital concept does not require 100 per cent debt financing. In any event, there are examples of firms securitising RAB-like investments. Criticism: The overall risk for a regulated entity reflects the nature of the business, market conditions and the regulatory arrangements. Splitting capital into two components might allocate risk in a different way but cannot change the overall risk of the entity. Response: The main rationale for the split cost of capital concept is not to reduce risk but to identify and price risk more accurately. The point is that the compensation to investors for the real risk they face has been more than is necessary to ensure efficient investment. The more accurate pricing of risk occurs, in applying the split cost of capital concept, by specifically recognising the low risk of the RAB and the higher risk of non-rab activities. The disaggregation of risk provides better information to identify risk and to assist in defining appropriate compensation for risk more accurately. (Note that it is possible for the split cost of capital concept to lead to a higher average cost of capital, depending on the size of new investment relative to the RAB and the cost of equity for that investment.) Criticism: The operation of the RAB and its cost recovery cannot be separated from non-rab activities without causing economic inefficiencies. Response: The RAB does not need to be physically separated from non-rab activities. The split cost of capital concept is about the pricing of risk not necessarily about the separation of RAB and non-rab functions. Criticism: Physical capital could fail, and this risk is not recognised by the split cost of capital concept. Response: Protection of the RAB effectively means that the financial capital that is represented by the RAB is protected. So even if associated physical capital should fail to provide a service or be underutilised, the financial capital is still protected and will be recovered in capital charges. Criticism: A split WACC is not necessary to encourage investment. Response: It is possible that an average WACC can be calculated to provide an appropriate lifetime return to investments, such that the WACC recognises different risks associated with different phases of the investment. However, where an entity undertakes a mix of investments over different time horizons the calculation of an appropriate single WACC could well be more difficult to implement than the split cost of capital. In addition, although users of an investment may face a single averaged price the key issue here is to ensure that the returns to the marginal investment match marginal costs. Furthermore, when investment market pressures specify a target rate of return for an entity the implementation of a split cost of capital may provide the incentive for a regulated firm to undertake new investment sooner rather than holding out for a higher return for new investment. Criticism: It is not the value of the RAB that is at issue but rather recovery of the cash flows generated by those assets; i.e. the prospects for realising the allowed rate of return on capital as well as the return of capital must be considered ver 6 iv Draft as at 04/02/14 8:45

9 Executive summary Response: Realisation of the allowed rate of return on capital will depend on the nature of the economic circumstances, regulatory arrangements and the efficiency of the firm. However, assuming the firm is efficient, the economic circumstances facing regulated firms typically imply that there is low risk that allowed revenues or returns will not be realised. Regulated firms face little or no competition, demand for their services is often inelastic, and in the case of many infrastructure businesses, demand for their services is not highly sensitive to overall economic fluctuations. The nature of the regulatory arrangements reinforces this proposition. This is especially the case where revenue caps or tariff structures effectively eliminate demand risk, and cost pass through mechanisms eliminate or greatly diminish cost risks. (l) (m) Criticism: The split cost of capital concept conflicts with the cost recovery principles of the National Water Initiative that recognise the need to provide a return to equity consistent with its risk. Response: The split cost of capital concept is not inconsistent with the National Water Initiative pricing principles requiring full cost pricing. Under the split cost of capital approach, at the end of an asset's life, the regulated entity will have received the full return of the capital invested. Moreover, 100 per cent debt financing of the RAB is not likely to be feasible and is not required to implement a split cost of capital concept. If the regulatory arrangements support the adoption of a cost of equity that is so low it is near the cost of debt for the RAB, the capital structure will have a negligible impact on the cost of capital. In these circumstances the cost of equity and the cost of debt will both contain low risk margins above a risk-free rate. Criticism: The application of competitive neutrality principles to regulated infrastructure businesses does not cause distortions in, or an excess of, investment. Response: The concern about the specific application of competitive neutrality is that it may be having unintended, adverse resource allocation effects. In addition, the competitive neutrality issue provides an additional consideration of the extent to which the RAB is protected by the regulatory arrangements for government-owned businesses. Strict application of the principle of competitive neutrality would require the government to make an explicit commitment about the degree of protection of the RAB for both government-owned and privately-owned businesses. However, given the wide range of views about the application and implications of competitive neutrality, this paper does not advocate a change to the current arrangements in relation to competitive neutrality. Application of the split cost of capital concept This paper applies the split cost of capital concept to a hypothetical, regulated urban water business. The analysis determines and compares regulated revenues and prices under the QCA s standard single WACC approach and under a split cost of capital approach. The base case scenario depicts the application of a single WACC to the RAB and to all types of capital expenditure (CAPEX). In addition, the base case scenario includes the standard operating expenditure (OPEX) allowance. Four scenarios are analysed: Two scenarios for capital structure of the RAB (and replacement CAPEX) 75 and 100 per cent debt as a proportion of the total value of debt and equity Two scenarios for the cost of equity applied to expansion CAPEX 12 and 15 per cent. Results for combinations of the four scenarios for the split cost of capital approach are generated and compared with the QCA base case approach. Details of the WACC parameters applied are explained in ver 6 v Draft as at 04/02/14 8:45

10 Executive summary Chapter 7. The parameters are based on information on market transactions and conservative assumptions. Based on the assumptions adopted, the modelling suggests that the split cost of capital approach can result in material reductions in the average bill per customer and the rate of return as compared to the base case. The magnitude of the net effect depends on a combination of the level of gearing applied to the RAB, the cost of equity applied to expansion CAPEX, the relative values of the RAB and CAPEX, and whether the disaggregation of the OPEX function decreases or increases the cost of that function. Conclusion and recommendations The regulatory arrangements in Australia typically provide strong likelihood of revenue recovery through the effective guarantee that regulators give to the value of the RAB and to the realisation of the allowed return on capital. However, there is a reasonable case for recognising that in some circumstances substantial capacity expansion entails greater risk than for existing infrastructure. The split cost of capital concept provides a means of recognising this demarcation. It is a potentially important and useful tool, as it is another method for helping to identify, disaggregate, understand and quantify relevant risks of the firm. The approach makes use of additional information to enable a more informed assessment of a firm s risks and their magnitude. At a minimum, applying the method provides a useful check on whether the existing approach to setting an allowed cost of capital is broadly reasonable. While regulators have not formally adopted the split cost of capital to date, many have adopted elements that are consistent with the split cost of capital concept. As noted above, it would be premature to adopt the concept unless implementation issues can be resolved. However, the split cost of capital approach is considered to be a useful tool for better understanding the amount, allocation and pricing of risk. Recognising these considerations the QCA proposes to explore the split cost of capital concept in the context of relevant price determinations. The determination of appropriate cost of capital parameters used in a single WACC approach would be informed by reference estimates of allowed rates of return and hence prices and revenues derived from the split cost of capital concept. The most relevant determinations would relate to sectors where the regulatory asset base is considered to be virtually guaranteed by the regulatory arrangements. The QCA also proposes to investigate use of the trailing average cost of debt concept for setting the debt component of the return on the RAB as part of its cost of capital review. The trailing cost of debt approach entails estimating a benchmark cost of debt that is a trailing average of debt costs over an efficient benchmark time period. Recommendations for further research are summarised below. Recommendation application of the split cost of capital The split cost of capital will be further investigated in relevant price determinations, where practicable, to assess relevant risks and provide reference point estimates. Recommendation trailing average cost of debt The trailing average cost of debt methodology will be further investigated as a separate matter as part of the QCA's comprehensive cost of capital review ver 6 vi Draft as at 04/02/14 8:45

11 The Role of the QCA Task, Timing and Contacts THE ROLE OF THE QCA TASK, TIMING AND CONTACTS The Queensland Competition Authority (QCA) is an independent statutory authority that promotes competition as the basis for enhancing efficiency and growth in the Queensland economy. The QCA s primary role is to ensure that monopoly businesses operating in Queensland, particularly in the provision of key infrastructure, do not abuse their market power through unfair pricing or restrictive access arrangements. In 2012, that role was expanded to allow the QCA to be directed to investigate, and report on, any matter relating to competition, industry, productivity or best practice regulation. Task The QCA is currently undertaking a comprehensive review of its cost of capital methodology for regulated businesses. A series of discussion papers covering various aspects of the cost of capital have been released for public comment. This information paper explains the split cost of capital concept, its rationale and its potential for application in the Australian regulatory context. The split cost of capital refers to the application of differentiated rates of return for RAB and non-rab components of a regulated business. The differentiation in the rates of return reflects different allowances for risk. This information paper was prepared taking account of submissions received on the Discussion Paper published on 3 April Further comments on this paper can be forwarded to the contact below. Key dates 3 April 2013 Release of Discussion Paper 31 July 2013 Closing date for submissions. Contacts Enquiries regarding this project should be directed to: ATTN: Dr Michael Blake Tel (07) research@qca.org.au ver 6 vii Draft as at 04/02/14 8:45

12 Introduction 1 INTRODUCTION The Queensland Competition Authority (QCA) is currently undertaking a review of its cost of capital methodology. The cost of capital referred to here is the weighted average cost of capital (WACC) applicable to the assets of a regulated business. The WACC is a weighted average of the cost of equity and the cost of debt, with the respective weights representing the shares of equity and debt in the capital structure of the firm. The cost of equity and cost of debt components of the WACC need to be set at a level that ensures investment can be financed at economically efficient levels. The QCA s practice to date has been to apply a single WACC to the firm s existing regulatory asset base (RAB) and capital expenditure in estimating the annual return on capital. As regulated utilities tend to have sizeable asset bases, their return on capital represents a significant proportion of the revenue requirement for these firms. Accordingly, the allowed WACC has a substantial impact on regulated revenues and prices. The allowed regulatory WACC is based on a benchmark firm's rate of return or cost of capital, rather than on the regulated firm's actual rate of return. There is no theoretical constraint on specifying separate WACCs for different aspects of a regulated business. Professor Dieter Helm (Oxford University) has authored a series of published papers and commentary articles, suggesting that applying a single WACC provides too high a return to the RAB and too low a return to capital expenditure. Specifically, Helm argues that the RAB is an accounting construct that reflects the value of past investments and once set, there is no scope for management activities to change that number. As the residual risk applicable to the RAB is political and regulatory in nature, Helm argues that it is best borne by customers and possibly tax payers. Providing a guarantee (whether explicit or implicit) to investors of recovery of past expenditure effectively transfers risk to customers or taxpayers. At the same time, capital expenditure can involve substantial financial risk. Large infrastructure projects tend to be complex and require active management to ensure, for example, that project costs remain on budget, and delivery is on time. Demand risk for major new infrastructure investments is also likely to be higher than for existing infrastructure. Thus, Helm argues that applying a single WACC over compensates the RAB and under compensates capital expenditure. Accordingly, Helm reasons that the RAB should earn at, or just above, the cost of debt; otherwise, customers are bearing the residual equity risk while paying investors a return on equity as if the investors bore the risk. In contrast, capital expenditure, particularly in the context of a major infrastructure expansion, is likely to entail higher risk than that for the RAB, and Helm contends it should therefore earn a WACC with a significant equity return component to take account of the higher risk. This approach would send a clearer signal for efficient investment to occur. This proposal is known as the split cost of capital concept, as it involves splitting the cost of capital into two distinct components to reflect the different risk profiles for the regulated firm s RAB and non-rab activities. The purpose of this paper is to explore this concept in more detail and to assess its potential for application by the QCA. This paper includes: a detailed discussion of the economic efficiency rationale for the split cost of capital; consideration of the various criticisms of the concept; and presentation of a stylised application of the concept to a hypothetical regulated urban water ver 6 1 Draft as at 04/02/14 8:45

13 Introduction business. The analysis determines and compares regulated revenues and prices under the QCA s current single WACC approach and under the split cost of capital approach. The QCA is unaware of any other regulator that has investigated this issue in as much detail, or developed a stylised example that applies the split cost of capital ver 6 2 Draft as at 04/02/14 8:45

14 Current regulatory arrangements 2 CURRENT REGULATORY ARRANGEMENTS 2.1 The regulatory problem Utilities and other businesses that own 'essential' infrastructure, where there is limited scope for competition, are often subject to some form of price regulation. Establishing the appropriate rate of return for a business or an investment is fundamental to the regulation of these businesses. This in turn requires balancing the need for efficient investment and the impact of prices on customers. The starting point for developing an appropriate methodology for the determination of a rate of return for a regulated business is to specify the problem that is being addressed. The conventional economic approach begins with identifying the nature of the relevant market failure or failures. Whether and how best to address market failure is considered next. Market failure refers to a situation where economic efficiency is not achieved. An economically efficient outcome is obtained where no feasible changes in prices, production or consumption can benefit society as a whole; that is, lead to a net public benefit. Regulation may be necessary where there are significant market failures. The specific form of regulation that is adopted ideally needs to lead to the largest improvement in economic efficiency, after taking into account the costs of the regulation or other intervention. In the case of public utilities and other businesses providing services requiring substantial essential infrastructure, the market failure of concern is market power. Market power is the ability to raise and maintain prices above levels that would be consistent with an effectively competitive market. However, there are three important aspects of market power and government or regulatory policy that need to be taken account of when designing an appropriate regulatory solution: (1) Market power may lead to prices that exceed the cost to society of providing the product, leading to excess profits and a lower level of production than is optimal. (2) The nature of regulation may change over time. This is the 'time inconsistency problem' and refers to the perception that the regulatory contract is not secure and may lead to ex post expropriation of investment capital, which in turn acts as a deterrent to invest in the first place. (3) The form of regulation may introduce incentives for the firm to operate and invest inefficiently. In order to achieve the best outcome for society each of these issues must be addressed when designing the form of regulation. It is also important to recognise that there are significant information problems and trade-offs in addressing these issues. The literature on regulation of market power has tended to focus on problems (1) and (3). However, there is an important strand of the literature that addresses (2), referred to as the transactions cost approach to regulation. The seminal paper was by Goldberg (1976) who described regulation of monopoly as a form of 'administered contract'. Gomez-Ibanez (2003) draws on this literature, presenting the problem of dealing with monopoly as a contracting problem in a situation where both the entity with market power and its customers are vulnerable to opportunistic behaviour (reneging on the rules by the government and exercise of market power by the entity respectively). The vulnerability stems essentially ver 6 3 Draft as at 04/02/14 8:45

15 Current regulatory arrangements from the sunk cost nature of the infrastructure investment and the transactions costs of developing contractual solutions. Biggar (2010, 2011) also draws on this approach to emphasise that regulation of public utilities in practice is not focused on the standard allocative efficiency aspects of the monopoly problem but rather on designing arrangements that focus on protecting the sunk investments of both the monopoly and its customers. The QCA (2013a) Statement of Regulatory Pricing Principles highlights the relevance of these perspectives including recognition of higher order governance principles that relate to overall credibility and sustainability of the regulatory arrangements. The split cost of capital concept has implications for each of the three problems noted above, but in particular represents a step towards more credible and sustainable contracting to help ensure the 'time inconsistency' problem is addressed effectively. An explicit or implicit guarantee of the RAB addresses the problem of how to ensure that sunk costs will be recovered and hence that efficient investment will not be deterred by the perception of potential regulatory expropriation of an investment ex post. An allowance for a higher return on certain new capital expenditure recognises that a single, 'average' cost of capital may not be sufficient at the margin to encourage new investment, particularly if some risks are not effectively addressed prior to the capital expenditure being included in the RAB. In addition, to the extent that the WACC for the RAB is lowered to better reflect the actual amount and allocation of risk, prices would be lower than otherwise while the return to capital should be sufficient to ensure that efficient investment occurs. These outcomes should have more economic credibility and, in particular, reduce the threat of arbitrary changes to the regulated cost of capital in future regulatory determinations. Furthermore, the extent to which there would truly be scope for a material, adverse impact on investment incentives depends on the extent to which investment risk is appropriately compensated for, under the current arrangements; that is, the extent to which above normal profit or economic rent exists. Another prominent criticism focuses on implementation issues. However, implementation issues are not necessarily any more difficult than the issues that arise with applying a single WACC. Considerable professional judgement is required when implementing either a single WACC or the split cost of capital concept. 2.2 Price regulation of infrastructure businesses in Australia Regulation of the output prices of firms with market power is a common feature of many developed market economies, including those of Australia, New Zealand, the United Kingdom, and the United States. The general form of this regulation involves setting the output price for a certain period (i.e. the regulatory cycle ), at the end of which it is subject to review and reset. Regulatory cycles are often around five years. A fundamental proposition in regulatory economics is the 'Net Present Value(NPV) = 0' principle, which requires that the present value of the regulated firm s expected net revenue stream equals the value of the initial investment, when those cash flows are discounted at the risk-adjusted, regulatory cost of capital. This principle is also known as the financial capital maintenance (FCM) principle, which in effect allows economic depreciation (the reduction in asset values between two periods) to be recovered in capital cost charges. Following directly from this principle, the regulated price should compensate the firm for its expected efficient ver 6 4 Draft as at 04/02/14 8:45

16 Current regulatory arrangements costs, including its risk-adjusted opportunity cost of capital and a return of capital (Schmalensee 1989; Lally 2012). 1 Importantly, the regulatory (i.e. allowed) revenues should normally reflect ex ante compensation to the regulated firm. In other words, the firm might under- or over recover the allowed revenues ex post. 2 As a result, the regulatory arrangements enable investors to have an ex ante belief that they will be able to earn a specified rate of return consistent with the risks of the investment, taking into account the impact of the regulatory arrangements on risk. The expectation is characterised as ex ante since an ex post arrangement would in effect be a guarantee of a specified rate of return. An ex post guarantee for all actual investment would amount to a guarantee of earning a risk-free rate of return, which would reduce the incentive to invest and operate efficiently. However, a RAB that is in effect risk-free or nearly risk-free, given the nature of the regulatory arrangements, has implications for the WACC that should apply to it. In addition to protecting the RAB, the form of the regulatory arrangements can also provide a strong guarantee that an ex post return on capital will be very close to, or more likely exceed, the ex ante allowed rate of return on capital. These features have not been explicitly recognised in regulatory arrangements in Australia. This model for regulating natural monopolies is commonly referred to in Australia as the building blocks model, and it is consistent with the NPV = 0 principle. The model specifies the regulated firm s ex ante allowed revenue requirement in terms of constituent components, or building blocks. These include a risk-adjusted return on capital, a return of capital (i.e. depreciation), and expected efficient operating costs, where the latter includes an appropriate tax allowance. An important building block is the return on capital, which compensates investors for the opportunity cost of the capital invested in the regulated firm. This component provides investors with a return based on what they would earn by investing in an alternative asset of equivalent risk. Australian and overseas regulators typically apply a single WACC to the entire business when determining the return on capital, despite apparent differences in risks associated with different classes of assets or business segments. Relevantly, unless the regulator has perfect information, the regulatory, or allowed, cost of capital will differ from the firm s actual expected return on capital. The WACC reflects the risk-adjusted rate of return to investors for the opportunity cost of their capital. Consistent with the general approach for estimating a range of regulatory parameters, the regulator estimates key WACC parameters by referencing comparable, private sector firms on the basis that they are efficient benchmarks. Parameters that are specific to the firm or investment include the asset beta, the debt premium, and the capital structure. Market-wide parameters include the risk-free rate, the market risk premium, the statutory tax rate and the rate of inflation. Allowances for tax 1 This principle is fundamental to regulation and uncontroversial. Any revenue lower than that revenue which satisfies this principle will be insufficient for investors to invest, and any revenue higher will result in the excess profit that regulation seeks to constrain. 2 The ex ante allowed revenue is not necessarily an upper limit on the regulated firm s revenue. For example, some Australian electricity distributors have been subject to a weighted average price cap, which allows them to over-recover the target revenue (and retain the difference) if realised demand is more than forecast demand. Such an outcome is possible because the regulatory price is based on forecast demand while the actual revenue earned is based on realised demand (AER, 2012b: 54-56; ) ver 6 5 Draft as at 04/02/14 8:45

17 Current regulatory arrangements imputation (i.e. franking credits) can be based on market-wide, industry-specific or firm-specific estimates. The return on capital in dollar value terms depends on the value of the firm s RAB and the regulatory-determined WACC. 3 The regulator applies the resulting WACC to the RAB to calculate the firm s allowed return on capital. The RAB represents the current regulatory value of the regulated firm s assets. It comprises the total value of assets attributable to a regulated firm at the time of regulatory establishment, plus the allowed value of replacement and additional assets acquired or constructed since establishment, less allowed depreciation (return of capital) and disposals since establishment. In Australia the asset base is also typically indexed by the Consumer Price Index (CPI) so that it is expressed in current, depreciated value terms. Indexation of the RAB by the CPI is consistent with maintaining financial capital from a general purchasing power perspective; that is, in real general inflation-adjusted terms. The regulatory arrangements for many regulated infrastructure businesses in Australia effectively provide strong assurance that the RAB value will be protected. Specifically, investors can be confident that there is minimal, and in most cases no material, risk that the financial value recognised in the RAB will not be fully recovered. However, aspects of the regulatory arrangements can also extend beyond providing significant protection or guarantee of the RAB to, in effect, a similar treatment for the return on capital. For example, revenue caps and price caps with a tariff structure that ensures the recovery of fixed costs, cost pass-through mechanisms and unders and overs accounts can, in effect, guarantee the return on capital. The RAB regulatory arrangements can protect its value and the regulatory arrangements in relation to the return on capital can substantially protect the expected or ex ante allowed return on capital. Regulated natural monopolies usually provide goods or services that require significant and long-lived, network-type infrastructure assets. As these assets reflect a significant element of the firm s total costs, the return on capital component in dollar terms is typically a significant proportion of a regulated firm s allowed revenues. For example, in its April 2012 final decision on the Queensland transmission network service provider, Powerlink, the Australian Energy Regulator s (AER s) allowed return on capital of $553.3 million and regulatory depreciation of $41.0 million for (account for 71.0 per cent of the total allowed annual revenue of $835.0 million). Over the entire five-year regulatory period ( to ), the return on capital and return of capital together account for 77.0 per cent on average of the total allowed annual revenue (AER 2012a, p. 3). 2.3 Key features Regulation of natural monopoly firms across Australia shares a common approach. In broad terms, this approach has the following key features: (a) (b) a single firm is responsible for the safe and reliable supply of infrastructure services to both existing and new customers a regulator ensures that the firm s allowed revenues compensate it for its expected efficient costs, including a single return on the RAB to compensate investors for the risk that they bear for investing in the firm. 3 Capital expenditure during the regulatory cycle also earns a return on capital ver 6 6 Draft as at 04/02/14 8:45

18 Current regulatory arrangements (c) If the firm is government owned, it is treated as if it raises debt and equity to finance its functions, where these include administration, operations, maintenance, and capacity expansion (both incremental and major). This model, therefore, involves a single corporate entity that takes responsibility for providing the essential network good or service by undertaking several key activities. These activities can be broadly categorised into three different functional areas: (a) (b) (c) capital recovery operations capacity expansion. The capital recovery function involves administration, financing and refinancing the completed and approved network assets (i.e. the RAB) until the capital value of historical investments is fully recovered from customers. The risks associated with this function principally relate to investors ability to recover the value of the RAB and their ability to raise finance at a reasonable cost. Given the monopoly nature of the assets (with limited scope for substitutes and relatively stable demand) and the nature of the regulatory arrangements in protecting both the return of and on capital, the risk of capital recovery and not meeting investor expectations with respect to returns is low. As a result, there is a minimal need for equity to finance RABrelated activities. As such, ongoing low-risk debt capital is the primary capital required. The network function involves operations, maintenance (e.g. asset renewals), and incremental capital projects (e.g. asset replacement). The principal risks with this function relate to efficient output and performance, including delivery of the output to a specified performance standard or level of reliability. The financing requirement for this function is moderate, with some risktaking equity capital required. The capacity expansion function refers to major capital expenditure programs that can involve lumpy increments of new capacity and long construction periods. Key risks with this function include planning and design, construction, cost and schedule variability, and political and environmental risk. Equity financing requirements for this activity are usually moderate to high, with significant risk-taking capital required for project construction and related activities. Table 1 summarises these functional areas and their characteristics. Table 1 Function / business Capital recovery Operations Capacity expansion Network business stylised model Adapted from First Economics (2010, pp. 7-8). Cash outflows Level of equity risk Financing requirement Capital servicing and Nil / very low Low-risk capital on an repayment ongoing basis Ongoing expenditure on Moderate Moderate risk-taking OPEX and incremental capital on an ongoing CAPEX basis Project-related CAPEX Moderate to high Moderate to high risktaking capital required during project construction ver 6 7 Draft as at 04/02/14 8:45

19 Current regulatory arrangements 2.4 Criticisms of the current approach A number of criticisms have been levelled at the current (building blocks) regulatory model, both within and outside of Australia. While these criticisms are covered in other research work, they are briefly discussed below, as they have a nexus with the split cost of capital topic Government-owned utilities, competitive neutrality and risk The first criticism relates to whether the model is appropriate in the context of governmentowned utilities and the competitive neutrality requirements that apply to them. Specifically, under the current model, the regulator allows the firm to earn a return as if it is owned by private equity holders. This gives government-owned businesses the same (regulatory) cost of capital as would be given to an otherwise identical, regulated firm in the private sector. There are valid reasons to expect that some regulated, government-owned firms have an actual cost of capital that is materially lower than that of a private sector equivalent. First, government-owned firms typically benefit from a government guarantee of their debt; that is, the firm s debt is backed by the government s credit rating. Biggar (2011) notes that this guarantee allows these firms to borrow at lower rates than private sector equivalent firms. Competitive neutrality fees are levied on the cost of debt incurred by State-government-owned firms in order that these firms do not have a competitive advantage over private sector firms in the marketplace due to their ability to borrow at a lower cost. 5 However, anecdotal evidence (see Section ) suggests that these fees are not necessarily high enough to reduce the incentive to borrow. 6 Second, tax equivalent payments, along with the competitive neutrality fees, are paid to the owners of the firms (i.e. the government) (Biggar 2011, p. 14). 7 The size of these revenues (i.e. the tax equivalent payments and the competitive neutrality fees) directly depends on the level of investment. Third, government-owned firms that are subject to economic regulation typically do not face competition from other firms for their services. Given these arrangements, the absence of any meaningful competition aggravates the problem and likely provides a commercial incentive for government-owned, regulated firms to overspend on investment projects: It is well-known that if a regulator over-estimates the cost of capital of a firm, other things equal that firm has an incentive to inflate its regulatory asset base. If government-owned firms in practice have a lower cost of capital than privately-owned firms, the policy of allowing these 4 For example, see Mountain and Littlechild (2010) and Biggar (2011). 5 King (2013) observes that regulators determine revenues for government-owned firms and that any benefit from using private sector benchmarks relates primarily to corporate governance and ensuring efficient operations. King argues that, unless the government gives regulators explicit guidance on revenue-setting rules, regulators should set prices for government-owned monopolies to reflect their actual costs. Otherwise, the absence of complete transparency in the revenue-setting rules conflates regulatory policy with tax policy (King 2013, p. 2). 6 In addition, the transactions costs of sourcing debt for these businesses might be lower than comparable firms in the private sector, given the involvement of state-level treasury corporations in assisting with debt management. 7 Tax-equivalent payments are equal to the corporate income tax a privately-owned business would pay in the same financial position. State government-owned firms make tax-equivalent payments to ensure competitive neutrality. Payments by these firms are made to the relevant State Treasury, not to the Australian Tax Office ver 6 8 Draft as at 04/02/14 8:45