The Dividend Puzzle: what should utilities pay out?

The Dividend Puzzle: what should utilities pay out?

Dieter Helm

4th December 2018

Why does Centrica yield 9.3% and SSE 8.4%, whilst DRAX yields 3.2%? Why do United Utilities, Severn Trent and Pennon yield respectively 5.0%, 4.8% and 4.4%, and National Grid 5.8%, whilst BT yields 6.0% (all at the end of November 2018)? Why are they all different, but most much higher than the yield on the FT 100?

The answers to these questions turn out to tell us a lot about both corporate strategy and its limitations, about what privatisation does, and about the effectiveness or otherwise of regulation.

Why pay dividends at all?

Let’s start with the age-old corporate finance problem - the “dividend puzzle”. The challenge is to explain why investors want to receive dividends at all. Investors invest in the belief that the management of the company can find profitable projects to invest in. The managers are the agents of the principals, the investors, in seeking ways to invest their savings. Why hand back some of the profits instead of investing them in the business? Paying dividends looks like an admission of failure: the managers can’t do better than the average, and many do worse.

Tax makes the puzzle of why companies pay dividends even stranger. Dividends are treated as income, and taxed at marginal income tax rates. Capital gains from share price increases are treated differently, taxed much lower and subject exemptions and offsets. As a result investors are better off with capital gains than dividends, typically by a lot.

Equity and management strategies

So what is going on?  Let’s go back to the wide disparity of yields lists above. Centrica and DRAX share many things in common. Both have faced declining core businesses. DRAX’s problem is that it had one big asset – the coal-fired power station from which, it got its name. It is aging and it is horribly polluting. Its days are numbered.

What should DRAX have done? It had two option options. It could have run the asset into the ground, and paid out all the profits, holding capital and operating expenditures to a minimum and then eventually closed the business down, maybe selling off the site itself.  Alternatively it could keep the profits as retained earnings, and instead invest them in new projects to give the business a future. In effect it could say to investors “let us reinvest the retained earnings in opportunities that will earn you above average returns”.

No management likes the harvest-and-exit option. Running down the business reduces scale and usually executive salaries too. Few top executives have made their careers this way. Yet from an investor perspective it is perfectly sensible and, as set out in my book Burn Out: the endgame for fossil fuels, it is a very credible strategic option for both large oil companies and the conventional vertically integrated energy utilities.  It has been forced on several of the latter. E.ON and RWE are shadows of their former selves. But the key word here is “forced”. The management, faced with significant losses could not hold their businesses together. For big oil, the catalyst has been crises – the combination of the collapse of oil prices at the end of 2014, together with the Gulf of Mexico disaster for BP, and for Shell the sheer scale of the cost of buying up BG just before the oil price falls. Both as a result had yields of around 6% plus, and these have persisted ever since.

Centrica has faced its challenges from the decline of its core customer base, and its post privatisation inheritance. It has tried all sorts of ways of getting out of the slow death of its core business. It tried to vertically and horizontally integrate, buying up North Sea assets, buying a stake in British Energy and its nuclear power stations, and getting into gas CCGTs. It tried home insulation, and lobbied hard and successfully to make sure smart meters were in supply, shoring up its core customer base. It has tried boiler services and home energy management.

For both DRAX and Centrica, it is understandable – and predictable – that they would look to diversify through reinvesting some of the profits. But as the yields reveal, and even after Centrica cut its dividend when it announced its latest strategy towards energy services, investors have been broadly supportive of DRAX and unforgiving when it comes to Centrica. The shareholders appear simply to not fully trust Centrica to re-invest profits profitably.

The story carries over to SSE, which unlike DRAX and Centrica started off with a better and more diversified hand. SSE has distribution and transmission and generation and supply. For SSE, it has many more options to tilt its strategy towards the more profitable bits of its portfolio. It can’t be accused of not trying in all its business areas. Yet investors do not buy the results. Why? Despite its wider optionality than Centrica, it is hard to resist the conclusion that investors do not have a great deal of faith in its management. Recent “surprises” have not helped.

Interestingly none of the above has been acquired. Though quoted, and though in a context in which most of the electricity distribution companies, and most of the water companies have fallen to M&A, all three of the above have survived. No raider thinks they should pay a premium for any of them.

Yield driven investors

In the cases of Centrica and SSE, investors expect the cash to be paid out to them. They appear not to trust the management to use it better than they can.  Faced with this, what can incoming new managers do?  They have two options. They can tell an equity investment story, and come up with a strategy that they try to sell to shareholders of a future of capital growth. That is what Centrica has been trying to do. It has a perfectly sensible strategy to develop energy services, and it cut the dividend in part to pay for it. The other option is to accept the lack of credibility in future investment and manage the business to pay the dividends. That broadly is one interpretation of what SSE is doing.

It is however not that simple. Once a utility has a high yield, it attracts investors who like high yields. They buy it for the yield and what they expect from the managers is that they will run the business to maintain and enhance it. The managers then are in a bind. They want to grow and expand, to reinvest profits, but their owners do not want them to do so. Indeed they express their preferences by hammering anyone who dares to cut the dividend.

This might sound strange, given the preferential treatment of capital gains over dividends noted above. Yet there is an obvious explanation. Those who prefer dividends to capital growth, and hence high yields, tend to be financial institutions, which combine tax protection for dividends, fixed liabilities, and legal constraints. These are pension and life funds. It turns out that these are indeed major ways in which world savings are channelled, and it is no accident that Britain, as one of the few countries in the world with privatised utilities, has attracted pension surpluses, from as far afield as Australia and Canada. In most other countries where utilities are owned by governments and municipalities, the only other options are government bonds and corporate debt. Unsurprisingly these are where they also put the savers’ money.

These companies are as result trapped. They cannot just pursue the opportunities in front of them. This matters not just for the obvious wealth opportunities forgone by investors and for the managers themselves, but also for the wider economy. They struggle to invest in infrastructure, and indeed more generally privatisation has not led to the leading exponents having the best infrastructure in the world. Quite the contrary. Few would say the competitive advantage in the US and Britain lies in the quality of its infrastructure, though some sectors may look better than others.

This brings us to BT, and its managerial problem. BT has in front of it a cornucopia of opportunities. Unlike the other utilities, it has not just the potential of a growing market, but fast changing technologies. The market has, in the space of just three decades gone from trying to install fixed line telephones in houses and offices to the world of smart phones, the Internet, the internet of things and the digitalisation of the economy, with big data and AI. It has literally seen the market transform in front of it, and the opportunities keep on multiplying.

The obvious expectation is that BT would have transformed itself from a boring utility stock at privatisation into a growth and equity-driven business. Yet it has not realised this full potential. It yields 6%, more than water or National Grid. Why? Because it has the pension investors incentives right inside the business. It is not Canadian or Australian pensioners who are looking to its yield. It is its own current and past employees. These act as a milestone around the management’s neck. And past management mistakes do not help its reputation when it comes to deploying profits for investments.

The result has not been good for BT’s shareholders. But what really hurts is the damage to the economy. Britain has gone from a leader in communications, following liberalisation back in the 1980s and 1990s, to a laggard. Britain needs a fibre roll out – and fast. The National Infrastructure Commission has appointed out the benefits, and it costs less than half that of HS2 to deliver. BT is central to this, and yet it has to pay out 6% to shareholders rather than invest.


Regulators play to the pension fund tune too. From an economy-wide perspective, the utilities matter to every household and company. They provide the basic framework for an economy to function. Electricity, heating, roads, railways, water, sewerage and communications networks are the building blocks for everything else. Only health and education matter more.

Why do the regulators allow regulated utilities in water, for example to earn 5% yields? The obvious answer is that is what the investors expect, and they can always go elsewhere. But this constraint on the regulators is not quite what it seems. For the wider economy, we want well maintained and appropriately enhanced system. We want clean water, safe disposal of sewerage, trains that work and roads that are not too congested, and we want our broadband to enable us to participate in society and enable companies to flourish.

For the core networks, these assets are best thought of as assets-in-perpetuity. Hence they should have capital maintenance to ensure the assets do not deteriorate, and they should be subject to current cost accounting. If the regulator then, having allowed for capital maintenance, gets the cost of capital right, the value of the companies should be equal to the regulated asset bases the regulators de facto guarantees. Enhancements, over and above capital maintenance, should be financed through a mixture of debt and equity, and for investors they will expect to get a return for the genuine equity they are exposed too, notably construction risk, and a yield on debt a bit above the return on government bonds.

But this is not what has broadly happened.  First, most utilities use historic cost accounting and include depreciation. That means that the yield incorporates a return of past investment without account being fully taken of the capital maintenance. The worst case is in the public sector – roads. The potholes are not filled in, so capital maintenance is neglected, and the Treasury does not have to incorporate an allowance for this. Governments national accounts and the national budget look artificially better as a result.

Second, investment has come from debt rather than equity. In water dividends roughly equal profits since privatisation. There has therefore been no net reinvestment by shareholders. Indeed, worse, the water companies (and the electricity distribution companies too) have engaged in massive financial engineering to take out equity and leverage up the businesses.

How has this been allowed to happen? The answer is bad regulation. Good regulation would have ensured two things. The balance sheets would have been used for investment and not financially engineered. Regulators should have stopped the great financial engineering game. And regulators should have imposed proper CCA accounts, and focussed on capital maintenance. In water and rail, the accounts are in principle CCA, but the practice is not all that it should have been.


Corbyn’s Labour Party argue that the right answer for many of these utilities is 0% yield, that nationalisation would enable the dividends instead to have been partly reinvested in the utilities and infrastructure, and partly returned to customers thorough lower bills – a public rather than a private return to investors.

Despite the outcry and what the water regulator calls the “Project Fear” campaign by a number of the water companies, there is nothing particularly revolutionary about the Labour proposals. They are what happened in Britain after the Second World War and through the great golden age of British economic growth, and for water at least the Corbyn model is the norm rather than the exception globally. Most water companies are municipally or national owned in the world.

What Labour is proposing is essentially the confiscation of the profits – the surplus of the revenues over costs, and a state directed allocation of those profits. It is the route by which the expansion of the electricity industry in the post war period was achieved, matching 3% economic growth with 7% electricity generation growth. Broadly, and whatever the critics say, it worked. Whatever Britain’s’ relative economic problems in the period 1945-70, it is hard to argue that it was infrastructure investment that held Britain back.

Yet this easy argument is not quite what it seems. What matters is less ownership per se, but rather how the utilities are managed and regulated once nationalised. For obvious political reasons, Labour has not explained how all this is going to work. Yet it is in the details that the future performance lies.

In the state sector, the first problem is that the cost of capital is not abolished. Scarce savings need to find a home, and pension funds have lots of options. Why buy government bonds in the Labour world? Project Fear would have it that nationalisation will lead to ballooning public debt and drive up the cost of government borrowing. On its own, this is nonsense. The cost of government borrowing under Labour might well rise, but the reasons lie largely elsewhere. To pretend that the RABs are purely private from the perspective of the economy as a whole is not credible. The RABs are a claim on the population, whether as consumers or taxpayers. Since the utilities will continue to charge customers, that claim remains to service the cost of debt.

What really matters in the Labour model is the equity bit. That after all is what privatisation is all about. Publicly-owned companies, like TfL, can borrow in private or public markets. What they cannot do is issue shares. Investors are replaced in nationalisation by ministers and civil servants.

Management by ministers and civil servants has a mixed record, and recently it has been challenging. Network Rail is now nationalised. Its performance is not stellar as a result. TfL is more difficult to call. The Channel Tunnel in the private sector was not a great advert, but then Cross rail is over-running and few think HS2 is going well. The key equity questions are about whether the public or the private sector is more efficient in capital maintenance, enhancements and operations. On this, contrary to Project Fear, the jury is out.

A mixed public-private model

A golden rule of finance is that risk should be allocated where it can be best managed. Political and regulatory risk should lie with government. Project and operational risk should lie with managers. The cost of capital depends on how this risk is in fact allocated.

An obvious conclusion is that there should be a clear division in the structure of these utilities. Operations and project delivery should be private; decisions about the design and functions of the systems should be public. This is the system operator model.

This division already exists in parts of the electricity industry. Investment decisions have been renationalised. All – or almost all – new investments are on the basis of state-backed contracts, and the networks are determined by regulators. In neither case is it the final customer. The state is the customer. This is the largest renationalisation since privatisation kicked off in the 1980s. Not even railways have been subject to such a change – mainly because the state was always the prime customer and not the hapless rail travellers.

The system operator model can and should be extended further into water catchment system operators, to rail and to fibre.  Again this is an evolution from what has already been happening, not something new. The result is that the delivery of new enhancements and the day-to-day running of the businesses is private, whilst control is public. Nationalisation of the building of power stations, water treatment works, and the building of HS2 would be a thoroughly bad idea, and one which would not have featured much even in the old nationalisation days. Running water, broadband and rail services is similarly best left to the private sector.

But what about the finance? Labour argues that the cost of capital to the government is less than that to the utilities. This is both right and wrong. It is right when it applies to the political and regulatory risk. Since these are in the hands of government and regulators, that is where they are best assigned. It is wrong when it comes to the equity risk in day-to-day management and delivering new projects. We can see this vividly by looking at a host of state-run activities. It is far from obvious that the government’s core state day-to-day functions – collecting taxes, running local services and the delivery of health and education - are exemplars of efficiency. There are other reasons why some of these should nevertheless be in the state sector, and especially for health and education, but they are not obviously better run.

When it comes to project delivery, the state has long given up trying. It turned to the private sector for almost all its construction works and capital maintenance.

The cost of capital in this system operator model then follows. The system operators auction the projects and the running of the networks. The private sector bids, and the cost of capital to the private sector is embedded in the bids. The market reveals the lowest cost for all the elements together. The private company than has to face its equity investors. They will expect a combination of capital growth and some dividends.

What is different in the system operator world is that the dividend yield is lower, and the equity return higher, because the contracts carry inherent equity risks.

All of this does however depend upon the structure of the contracts. The equity risk could be still with the state, if there is an element of pay-as-you-go, rather than pay-when-delivered in the contracts. Why would the state allow private contractors to have money for assets-in-the-course-of-construction? The main case is where the government might want to keep changing the spec, or where the fundamental risk remains political and regulatory. Nuclear might fall into this category, but not most of the utilities.


The dividends puzzle in the utilities derives from the peculiar political and regulatory characteristics of the British model that has evolved over the last 30 years. It has been inherently inefficient, and the yields reflect this. Every customer individually, and the economy more generally, pays a price for this, as a deadweight welfare loss. It is no accident that in water profits equal dividends. They should not, and this should never have been allowed. Specific regulators who had choices in the 1990s and failed to take them have a lot to answer for, but so too does government also for not putting a stop to this financial engineering.

The fact that companies remain trapped in paying high dividends, and hence that retained earnings are not financing as much investment as they should be, opens to the door to more radical arguments about nationalisation. In itself, changing ownership solves very little. What is required is to go back to basics and sort out which equity risks should be allocated where. The system operator model does this – and does it much better than any alternative model. It is vastly superior to the original privatisation model, and much better than outright nationalisation. It sorts out ownership and control.

The remaining problem is the tax system and the distortions between the taxation of capital gains versus dividends. A very high price is paid for this distortion. Instead of pension funds financing the investment upon which the future economy that is supposed to pay the pensions should be built, the tax distortions are doing the opposite. They undermine investment. It is no accident that Britain has such poor overall infrastructure, that we are faffing about with fibre, that our transport system creaks, and that water management and catchments are less good than they should be.

© Dieter Helm. All rights reserved.Copyright & Terms