June 13, 2008
Britain’s utility model is broken
Privatisation was one of the great achievements of the Thatcher era. But it is becoming increasingly evident that the transfer of monopolies into the hands of regulated companies that own, run and develop the assets is flawed. This is excessively costly to consumers. It is also an obstacle to investment in risky long-term assets such as airports, nuclear power , electricity and gas networks.
This is not to argue that privatisation is devoid of benefits. Where competition could be introduced into the newly privatised industry, as in the case of telecommunications, the gains were huge. Elsewhere, privatisation was the way to allow essential activities to escape from the dead hand of Treasury curbs on public investment. Private finance was more expensive, but investments were at least made.
Yet, as recent work by Oxford University’s Dieter Helm makes clear, it is time to review the model. The bundling together of different functions in one regulated entity, and the rules on costs, particularly of capital, need rethinking.
A regulated utility consists of a set of assets, an operating function and a co-ordinating function. The second, in turn, consists of two activities: running the business day to day and planning and implementing investment projects. Professor Helm argues, persuasively, that lumping all these together has led to inefficiency and a rip-off of consumers*.
The remainder of this column can be read here. Click through to read the debate from our expert panel.
Jon Stern: Martin Wolf, in his 13 June column, argues that Britain’s utility model is broken. Following Dieter Helm, he tries to persuade us (a) that the utilities earn excessive profits and (b) that the operation of existing assets should be debt financed. We are told (following Professor Helm’s May 2008 paper on the water industry) that the water companies have “excess earnings” through financial arbitrage of £600-1,000 million per year.
If this were true, it would, as Martin claims, indeed be intolerable. However, these are not excess earnings because of rents or similar. The “excess earnings” arise because utility companies earn an equity return on a significant part of their existing capital stock as embodied in the RAB (the Regulatory Asset Base). Hence, Dieter Helm’s long-standing arguments that at least some of the utilities, especially the water industry, should not have been privatised as equity based companies, but either been left in state ownership or been placed in a franchise model as largely used for water and energy distribution in France. But, that is entirely to ignore the much greater efficiency and other incentives in the UK privatisation plus price cap regulation model.
The argument of “excess earnings” only holds if the equity based model and the debt based model produced the same results on efficiency, cost reductions and technical progress as the debt based model across all the utilities. As Martin points out, for telecoms competition - including both network as well as service competition - has taken off. In the UK, we now have functional separation of BT which further entrenches competition in fixed line services as well as mobile and broadband competition. The result is rapid technical progress with significant quality improvements and falling prices for consumers. For telecoms, the state-owned monopoly model is dead. Light handed regulation with some price controls and equity based companies is the worldwide dominant and successful model.
But, telecoms has always been the pathfinder among the utility industries. Martin specifically exempts telecoms from his strictures but does include electricity and gas as well as water and airports. The key point to note here is that the positions of the various utilities are very different.
For railways, the privatise and regulate model failed – primarily because railways are chronically dependent on operating subsidies. For airports, there is general agreement that the current UK regulatory regime is seriously flawed, not least because there are no airport licences by which conditions can be placed on operators and their annual performance regulated via licence conditions. Hence, we now have the government commissioned review of airport regulation. Also, in both UK railways and airports, contracts are becoming increasingly important relative to regulation - including contracts between airports and airlines, train operators and network operators and similar. Hence, ‘regulation’ is increasingly being carried out by the regulator mediating and approving contractual agreements between operators and with much less reliance on standard regulation based on future operational expenditures, investment and RAB projections.
That leaves electricity, gas and water as the industries which Martin cites along with airports. These are also, not surprisingly, the areas on which Dieter Helm has, understandably, focused his arguments that we should have franchising of (at least network) operations plus debt financed asset base and public investment in new investment. That is a perfectly reasonable position to adopt and leads to a French type of franchising model where the companies (network or network plus service companies) operate under management contracts with the public sector (national or local government) financing new investment. However, with private ownership, there is the issue of how, moving forward, the new investment is counted into the debt-financed RAB and adequately remunerated. Maybe Dieter Helm’s City financial engineers will find a way.
The fundamental question, though, is whether the French-style franchise model is better or worse for the UK and other countries in the energy water and maybe other industries. My answer is, in general, that it is not superior and, particularly for electricity and gas, the evidence suggests that the standard UK regulatory model has worked well and provided effective incentives for new investment and for both static and dynamic efficiency benefits. It is worth noting that, within the last month in Germany, Eon has announced that it is separating out its transmission network and RWE is proposing to separate out its gas transport network, as British Gas did a decade ago. This brings them much closer to the England and Wales model with price cap regulation of networks and competition monitoring of generation and supply markets.
In the UK, the evidence is that for electricity and gas, firstly, there have been major efficiency gains and abstracting from recent increases in world fuel prices, both cost and quality savings to consumers; and, secondly, that there is no evidence of under-investment. When Ofgem were concerned about under-investment in electricity networks, they introduced incentive mechanisms within RPI-X to handle this viz menu regulation for distribution networks. There have been suggestions (by Dieter and some others) of under-investment in generation but the case is not strong and, if there is a problem, it seems to be primarily due to planning process issues rather than to regulatory incentive problems. As for evidence of excess profits, the main argument here is the profits that generators have made on tradable emission permits. This is, of course, nothing to do with the financial issues raised by Martin in his column but derives from EU and Uk government decisions to allocate for free rather than auction the emission permits.
For water, the case is more debatable. This is presumably why Dieter Helm, understandably focuses his arguments there, arguments that Martin seems happily to accept. There is undoubtedly a question as to whether and how the water companies can finance their investment programmes. But, again, at leat up to now, insufficient investment seems not to be a great problem – Ofwat are making it clear that they are much more concerned with padded rather than insufficient investment programmes. But, is the best answer to the financing issue to go to French-style operating companies with public sector debt finance of new investment (as Martin and Dieter Helm suggest); or, as I and many others would argue, to separate water networks from service supply, actively foster competition in service supply and develop current generation models of price regulation for the networks?
One of the reasons why Dieter has advocated the water franchise model is that he clearly does not expect major efficiency gains from water company unbundling and the introduction of water supply competition. Effective competition in water supply is clearly a lot more difficult than in electricity or gas but I am very struck by the number of water industry specialists who claim that a lot more is possible and economic than one might imagine and that there may well be significant benefits. One of the tests of this will be the results of competition in the market in the Scottish water industry for industrial consumers.
This is not a new debate. As Dieter Helm points out in his May 2008 water paper, we had the same debate when Welsh Water as (effectively) a bankruptcy workout adopted a debt –only model in 2000. At that time, there was a major debate about whether the future of water companies (and energy network companies) was for debt-financed structures with operations contracted out to specialist operating companies. Not surprisingly, Dieter Helm has always supported this model and Martin follows him.
However, as Dieter has pointed out, the arguments for the Glas Cymru debt-financed model were only reluctantly accepted and both – particularly Ofgem made clear their strong opposition to any future moves in that direction. Indeed, Glas Cymru is obliged to hold £300m of reserves as quasi-equity. The arguments against the Glas Cymru franchising model were were set out at the time by Ofwat and Ofgem, with Ofgem being particularly hostile. The arguments behind the Ofgem view were set out in a 2001 Beesley Lecture by David Currie and the key points were:
(i) In a private debt-financed company, as the gearing level approaches 100%, the debt increasingly becomes like equity and is priced accordingly, so the debt-equity distinction disappears. (Note that Network Rail is dependent for its debt finance on government guarantees and that the value of these guarantees should be added to the pure interest cost.)
(ii) The severe and well-known co-ordination problems between management companies and their owners, particularly over the volume and cost of new investment.
(iii) The much weaker incentives on debt-only water or energy distribution companies over managing their assets and risks of operational failure – much of the risk of this effectively reverts to government and the taxpayers.
(iv) The difficulties over a long period of repeated long-term contracts of sustaining strong managerial involvement of the debt-owned company in the quality of service and, even more so, in continuing to improve efficiency. It is very difficult to provide effective long-term incentives for this.
Ofgem (and David Currie) placed particular emphasis on the last point. However, it is interesting to note that David Currie in his 2001 lecture raised some of the cost of capital and investment financing concerns that Dieter Helm has recently done in the water context – albeit with a clearly expressed view that they can be resolved within an equity based utility regulation model.
The arguments above are all based in theory rather than on empirical observation. But, French experience, particularly in water does show that the traditional franchise model is not a settled ideal. Firstly, there has been a significant growth over the last 10 years or so in the use of lease contracts for water concessions. These essentially operate as PPPs; they bring together management and operations and provide a significant element of equity return to the lease-holders. Secondly, the Conseil d’Etat, the French court which acts as the monitor and enforcer of the contracts increasingly looks like a UK style regulator with the powers to review and modify contracts, including awarding tariff increases or reductions for causes similar to those used by Ofwat.
However, the most important point to note for French water (or energy) is that there are very few operating companies willing to tender for the management contracts - only two or three major ones - and that the contracts are of long duration. Hence, competition for the contracts is very limited and the potential – as well as the incentives – for anti-competitive and even corrupt behaviour are very high. It is far from a transparent regime and it is also one with weak efficiency incentives. In consequence, more market-oriented French infrastructure economists are looking at ways in which market disciplines can be enhanced, including more equity style funding as well as the kind of tendering of construction, investment and some operations that Martin Wolf, Dieter Helm and I would all support.
Jon Stern
P.S. A final question. Royal Mail is a utility industry. It operates as state-owned entity but one with substantial wholesale competition at the sorting level and little retail competition and it is regulated by Postcomm, who set a 5-year price cap. For many years, Royal Mail was a stagnant nationalised industry with low efficiency levels. It still has sole operational responsibility for the postal Universal Service Obligation and is also trying to compete with the other postal operators, while carrying major legacy problems. The current regime is in crisis and there is a government review commissioned by BERR. The one thing on which Royal Mail and Postcomm are agreed is that Royal Mail should be able to attract private investment (including at least potentially equity or equity partners) in some businesses.
Political concerns may prevent full-blown privatisation but wouldn’t that - with regulation and maybe with functional separation - be (and have been) a better solution? Yes, it would give equity returns within the equity element of the weighted cost of capital but it would also have given hugely increased efficiency incentives, including strong pressures finally to resolve legacy pensions and labour relations issues.
Note that various utility economists have advocated ‘standard’ privatisation and regulation of Royal Mail … as did Dan Corry, currently Head of the Number 10 Policy Unit, in a 2003 IPPR pamphlet.
Jon Stern is Senior Research Associate, Centre for Competition and Regulatory Policy (CCRP), Department of Economics, City University, London
Posted by: Jon Stern | June 16th, 2008 at 1:47 pm | Report this commentDavid Starkie: Dieter Helm puts forward some interesting ideas but goes too far in claiming that the current system provides tendencies for utilities to underinvest. It is not one I recognise in relation to the regulated part of the airports industry, and for good reason.
First, the utilities have some market power. They are responsible for formulating their investment programmes which the regulators have, in many cases, been reluctant to second-guess and thus constrain. Consequently, the utilities have the opportunity to leaverage their market power through their capex programmes.
Second, they have an incentive to do so. Managers prefer to manage big rather than small companies (and are remunerated accordingly). There is, therefore, a combination of both opportunity and incentive to expropriated monopoly rents, not through prices, but through excessive investment. For an example, consider Stansted airport which has struggled for more than a decade to make a decent return on capital.
The writer is an associate of the Institute for Fiscal Studies, London and Visiting Scholar at the Sauder Business School, UBC Canada. He is also a co-editor of the Journal of Transport Economics and Policy.
Posted by: David Starkie | June 17th, 2008 at 2:09 pm | Report this commentMichel Rocher: Your article arguing that the UK’s utility model was “broken” was very thought-provoking but I disagree profoundly with the assumptions on which the article is based.
In a rational world, RAB can only be financed with 100 per cent senior debt if it carries absolutely no risk at all. The contractual structure that Prof Helm described in the note which is attached to your article does not (and cannot) eliminate all the risk that the owners of RAB have to bear. These residual risks include the risks: that the contract creates the wrong incentives; that material adverse change clauses mean that unexpected costs are incurred by the holder of the RAB; that there are changes in taxation; that the operator goes bankrupt having received significant payments which then have to be written off and perhaps most serious of all that there is a mismatch between the optimal asset replacement cycle and the length of the contract. Close examination of the water contracts in France does not support Prof Helm’s statement that the concession structures there are risk-free to the municipalities that theoretically own the assets.
If RAB is not risk free, then 100% debt financing can only be accomplished if someone else bears the residual equity risk. If the debt structure were tiered, then the subordinated debt holders would in effect have equity risk and ought to charge for this in the same way as equity holders. The risks could be insured through the private sector but any insurer would determine the insurance charges using the same cost of equity model that the Regulators use i.e.WACC, which after all is simply drawn from generally-accepted financial economics. If the public sector bears the residual or equity risk then it rationally it should charge a fee to compensate it for taking the residual risk. This is the opportunity cost to society of the taxation that would have to be imposed. This cost is high because taxation distorts market prices and reduces economic activity. There is considerable evidence to suggest that the marginal cost of taxation to society is greater than the cost of private sector financing for most goods and service (i.e. greater than WACC). On top of this is the problem, which many others have observed, that public sector managers are less efficient at managing these kinds of residual risks than private sector owners of RAB because they have less to gain (and lose) from success and failure. Of course the residual risks could simply be passed to the customers but then customers would be left with volatile prices and there would be no incentive for anyone to control these residual risks and they would most likely escalate.
Of course rationality does not prevail at all times and there are doubtless some who argue that 100 per cent financing is possible but it ought to be obvious now that low interest rates through the world have created an illusion that many businesses have more debt capacity than they really do. The illusion has been exposed in the financial sector and in the housing sector. It would be madness to extend it to the utilities sector.
The author is an economist at Milan-based Cabinet Azurro
Posted by: Michel Rocher | June 19th, 2008 at 5:13 pm | Report this commentDieter Helm: Martin Wolf set out his views on Britain’s utility model drawing heavily on articles I have written on the subject. Jon Stern has responded setting out his interpretation of my arguments and his comments. In respect of my views, he misinterprets them. For the avoidance of doubt, the main thrust of my argument is that:
(i) in the setting of a weighted average cost of capital (WACC), regulators create an arbitrage opportunity between the marginal cost of debt and the WACC
(ii) in respect of the RAB, the duty to finance functions in effect transfers equity risk from shareholders to customers - but only in respect of the RAB
(iii) it is this RAB and the duty which together distinguishes utilities from other conventional commercial activities from a finance perspective
(iv) the WACC is also below the marginal cost of equity, and hence shareholders do not earn a full return on OPEX and CAPEX
None of this should be controversial. I next argue that:
(v) because the RAB risk is transferred to customers (or taxpayers in the Network Rail case), it can be financed by debt
(vi) recent M&A activity in utilities (and also share buy-backs and gearing measures by those not taken over) have resulted in a substantial financial arbitrage
(vii) since the equity risk is being carried by customers, they should not pay in respect of the RAB the WACC, but rather the cost of debt
(viii) customer should also pay the cost of equity in respect of the OPEX and CAPEX
(ix) this in essence is the split cost of capital model which I have advanced for a number of years, and the RAB/split cost of capital model can be generalised across a range of activities where society commits not to expropriate investors making long term sunk capital investments, where the ex post incentive to opportunism by politicians and regulators is considerable - because once the investments are sunk, the marginal cost is well below the average costs.
Jon Stern does not dispute this argument in his comment. He does however quite incorrectly suggest that I therefore argue that “some of the utilities…. should not have been privatised as equity based companies, but either left in state ownership or placed in a franchise model as largely for water and energy distribution in France”. Let me stress I have never argued either the premise or the conclusion Stern asserts. On the contrary, I am strongly in favour of privatising these assets and have not advocated the French franchise model - particularly since it leaves the assets in the public sector. I can only conclude he has not read what I have written and published! And my views on equity are as above: equity is critical in respect of the efficiency incentives for the OPEX and CAPEX.
Stern then goes on to criticise the French model, and by implications me. He makes some good points - but none of this has much to do with my views.
More extraordinarily, Stern then asserts that I have “always” supported the Welsh model. This is simply wrong - I actively argued against it, supporting Ofgem in its criticisms. For the avoidance of any doubt, amongst the many flaws in the Welsh model is that there is no genuine equity to motivate efficiencies in the OPEX and CAPEX of the business - and the quasi-equity to which Stern refers is just customers’ money the management have chosen to withhold.
A second comment has been made by David Starkie, focusing on the investment incentives. David will know from careful reading of my papers that I refer to the investment incentive in respect of the difference between the WACC and the marginal cost of equity - which he does not dispute in his comment. On the wider investment incentives, it is true that given the arbitrage available between the marginal cost of debt and the WACC (once the investment is completed and has been placed in the RAB) may make utilities into “RAB generators”, but this is hardly a stable basis for these industries - not least because regulators may well in due course catch up with the sheer scale of this arbitrage.
On the specifics of airports which David refers to, I cannot think that those who have to use Heathrow Airport will conclude with him that it has been subject to either sufficient or excessive investment.
In conclusion, neither set of comments explains why it is appropriate for customers to pay for the (massive) arbitrage between the marginal cost of debt and the WACC on the RAB, given the equity risk has been placed upon customers by the duty to finance functions. Nobody has to do much to earn the RAB return. And going forward, given the UK economy faces a wall of CAPEX in infrastructure industries (and to address climate change) there is a world of difference between financing an important part of this from the debt *for that component which is the sunk and completed capital investment*. This can be achieved without damaging efficiency incentives - and this is better met by providing the marginal cost of equity for the OPEX and CAPEX, rather than the (too low) WACC. Making the RABs tradeable - which is what I have also suggested - would bring further financial efficiency to the market for what are in effect utility-RAB bonds. I do not understand how anyone could think that this model either is equivalent to advocating keeping the assets in public ownership or reducing efficiency incentives. It is also notable that the logic of my arguments are precisely what the capital markets have been doing in the large scale financial engineering that has been taking place, and in the way in which competitive markets have been brought to bear in bidding for components of the OPEX and CAPEX programmes.
Posted by: Dieter Helm | June 20th, 2008 at 9:14 am | Report this commentDieter Helm: Michel Rocher makes the obvious point that the RAB can only be 100% debt financed if it carried no equity risk (he actually writes “absolutely no risk at all” which is not strictly correct, since debt is not risk-free). He however fails to make any reference to a core component of my argument - that the specific duty on regulators to ensure that “functions can be financed” transfers equity risk from shareholders to customers in respect of the RAB. The RAB is an accounting number in UK regulation representing the sunk funds of investors, and the guarantee is in effect that regulators (and politicians) will not behave opportunistically ex post to exploit the difference between the marginal and average costs. The list of risks that Rocher sets out arise in respect of the CAPEX and OPEX - and here equity has a strong role to play provided that the regulation is ex ante not ex post. Thus Rocher “disagrees profoundly” with me, whilst entirely and completely ignoring the financing functions duty and its implications.
I argue that the regulation in the UK transfers equity risk from shareholders to customers in respect of the RAB, but that by providing a WACC customers are paying a premium in respect of the WACC over the marginal cost of debt and bearing the equity risk in respect of the RAB as well. The scale of this arbitrage is potentially around £1billion per annum - and Rocher does not explain why this considerable premium should be both paid by customers whilst they bear the risk too. This is not a trivial sum, but once the scale of the future investment requirements for UK infrastructure investments are taken into account, represents an enormous difference in the future financing burden for the RAB component.
Shareholders - and managers as their agents - do not have to do very much to earn the return on the RAB. They can make or lose money on the business of the utility, but even in the event of a business failure, the special administrator function takes over, and the business is sold on to someone else - with the associated RAB.
The tax issue can be - and indeed is being - sterilised in the UK model.
On the equity risks, my argument is that the WACC under provides, since the marginal cost f equity is significantly in excess of the WACC. These equity risks in respect oft eh OPEX and CAPEX should be better rewarded - as my split cost of capital approach would provide.
Finally, I do not advocate the French model, which may have many strengths and weaknesses. Mine is a private sector RAB-based model, not a public franchising model.
Posted by: Dieter Helm | June 23rd, 2008 at 9:16 am | Report this commentI think Dieter Helm has provided adequate replies to his critics. Whether he has provided an adequate reply to MY critics is less clear, since I have clearly diverged from him on some important points. In particular, I have failed to see any great merit in privatising the assets of utilities that cannot be made subject to competition. Far better, I argued, to make firms compete for the right to manage and develop those assets.
I do not wish to go through all the many issues discussed above. But there are certain fundamental points on which, I believe, Dieter and I agree. In particular, the aim of any sensible regulatory regime is to align risk-taking with incentives.
As Dieter has noted, regulators have a duty to ensure that the regulatory asset base is financed. In this case, therefore, risk is borne by customers, not owners. There is no reason, therefore, for the latter to earn an equity return on their ownership of the RAB. But operating the business and building new capacity are inherently risky activities. Furthermore, it is desirable to create a regime in which those who make these decisions both bear the risk and are rewarded for doing so. They need to be given an ex ante equity return.
This, then, is the case for the split-capital cost regime that Dieter recommends, instead of the weighted-average cost of capital regime actually used, in which the return on the RAB is too high and the return on the risky activities is too low. This, argues Dieter - persuasively, to my mind – provides inadequate incentive to undertake new investment and excessive rewards to financial engineering of the return on the RAB.
While this theoretical position seems to me to be convincing, it does raise practical problems. One is how to incorporate the capital expenditure in the regulatory asset base, without undermining expected returns on this risky activity. Another is where to locate the co-ordination function. Yet another is how to create competition for the right to operate and expand assets. I have some answers to these questions, but hope to persuade Dieter to add his comments on these points, first.
Posted by: Martin Wolf | July 19th, 2008 at 1:36 pm | Report this commentDieter Helm: The theory behind the split cost of capital is based upon the allocation of equity risk between taxpayers, customers and shareholders. For much utility and infrastructure investment, the regulated asset bases represent the sunk capital which investors have provided and customers (or in some cases, such as Network Rail, the taxpayers) commit to financing. In effect this is akin to a long term take-or-pay-contract: given the risks that governments and regulators will behave opportunistically by imposing marginal cost pricing ex post (the time inconsistency problem), customers need to commit to paying these sunk costs to induce the investment. This commitment is formalised through the duty on regulators to finance functions, which has the effect of transferring the equity risk in the RAB to customers (or taxpayers). Hence the RAB can be financed through debt - at a cost much below the WACC.
The fact that shareholders currently receive the WACC on the RAB provides a major opportunity for financial engineering, with customers both taking the equity risk and paying a premium. Conversely, since the WACC is below the marginal cost of equity, in the creation of new assets (the CAPEX) and the OPEX - both of which carry equity risk and should be exposed to high-powered incentives, shareholders are provided a return which is too low.
Martin - and a number of commentators - then raise a series of practical questions about implementation. These include: the translation of CAPEX into the RAB; the location of the coordination function; and the competition for the CAPEX and OPEX delivery. Each of these have relatively straightforward solutions, but it is important to point out before addressing them that if the theoretical arguments above are accepted (and there has been no convincing rebuttal to my knowledge) then even if there are practical compromises and imperfections in the implementation, it is better to be approximately right than certainly wrong - especially given the sums involved in the error between the WACC and the cost of debt.
Taking each briefly in turn:
(i) updating the RAB
This is already an integral part of the regulatory regime at periodic reviews
(ii) the coordination function
In principle this could be done by government, regulators or companies. Governments are increasingly involved in determining the shape of the CAPEX (for example in airports. rail and roan, and through the water framework directive in water). Regulators have a duty to ensure that the ex ante costs of CAPEX and OPEX meet the efficiency requirements, However there is much to be said for the utility itself organising the competition for these activities - provided of course that it is not itself one of the competitors. Which route is taken is a matter of pragmatically weighing up the industry characteristics. In rail, it is very hard for the government not to play this role. In water and energy it is far less obvious.
(iii) competition in delivery
There has been a strong growth in the development of both general and specific service companies who are increasingly involved in doing CAPEX and OPEX for the utilities. and this market provides a very good and rich source for competitive bidding.
I am currently preparing an extensive and detailed paper which will elaborate these issues in much greater detail, to be published in the autumn.
Posted by: Dieter Helm | July 23rd, 2008 at 12:05 pm | Report this comment