The Thames Water saga has run on for several years.[1] Almost all the forms of conduct regulation have been explored, with ever-greater interventions by Ofwat and the Environment Agency (EA). There is little evidence that these have made much difference: the performance remains on a plateau; the investors are reluctant to put more money in; and the large debt pile casts a shadow over the business. Thames Water is environmentally unsustainable and it is probably uninvestable.
The classic response is to think up yet more interventions, to regulate more and more of the decisions that would otherwise be taken by the company’s board, and to try to fudge the outcomes with a whole packet of sticky plasters. It is not hard to see why: incremental regulatory creep staves off making hard decisions; it kicks the problems down the road and leaves them for someone else to solve. In the meantime, there is always the hope that “something will come up”, and that the latest “turnaround” plan will work.
However tempting this option might be, the unsustainable nature of the business and the multiple problems throughout the Thames catchment mean that it will not be work. Whether now or in a few months’ time, or at best one or two years’ time, Thames Water, as it is currently set up, will probably go under. The day of reckoning is fast approaching. Instead of sticky plasters, it would be better now to grasp the opportunity and allow a fundamental re-set to place the business in a position to deliver what is required over the next couple of decades. Patching up the twentieth-century model will not be good enough for the twenty-first century.[2]
Conduct regulation has already gone too far; structural solutions are required
Conduct regulation grows as each problem emerges. There is a demand for the regulator to “do something”. Something is then done, and this creates unintended consequences, and typically what is done does not resolve the underlying problems. Not always, but mostly this is what happens. More intervention is piled on top. Conduct regulation keeps growing and growing.
In the case of water, the sheer scale of the documentation for periodic reviews is now enormous. What once filled a filing cabinet is now in large digital files. Armies of consultants are employed by the companies and their investors, and Ofwat struggles to come up with reasonable answers. Periodic reviews are now major business undertakings. Almost all of this activity is a deadweight welfare loss.
This is not just an Ofwat problem, though it started out from day one with a more detailed and intrusive approach. Across the utilities, the company boards find that they are essentially asking the regulators to make decisions for them. In recognition that the “customer” is at Ofwat rather than the household and business users, utilities engage in lots of lobbying, and try to work out what answer the regulator wants, rather than what their customers want and the wider environment needs.
This has played out in the periodic reviews. Utilities start by trying to guess the answer the regulator (and the government of the day) might want, and then shape their business plans around them. That is why the prices have followed a fairly flat profile despite the investment needs. The usual answer has been to start with RPI – 0 or CPI – 0 and work backwards. Ofwat has also developed the incentive structure and the companies have mostly worked out how to play games to get the best periodic review deal and beat the efficiency targets.
This regulatory and political dance has been facilitated by the extremely vague contents of the licences. Ask a water company what precisely its functions are and it is hard to get a good answer. This means that violating licence conditions remains hard to do and harder to prove. Having extended the licences to 25 years, threatening to take a licence away has remained an empty threat. Not even Southern Water was put into special administration. Instead, the functions are defined by the law, and on the environmental side this is also less clear. In any event, the result is limited fines. As yet, failing to meet an environmental consent or other legal requirement has not triggered any process for taking away the licences.
More of this sort of conduct regulation is not going to sort out Thames Water or the rest of the water sector. Instead, something more fundamental is required: structural regulation, accompanied by a retreat and re-set of conduct regulation.
Why might structural regulation be better? The purpose of structural regulation is to separate out clear entities, with well-defined licence obligations, appropriate balance sheets and accounting frameworks. These entities are charged with specific tasks, and are transparent in their delivery. They are of a size that gives the management a reasonable span of action, without distractions driven by wider corporate empire-building and mergers and acquisitions (M&A). The role of these entities is defined so they can “stick to the knitting”. They do not need “global leaders” paid “global salaries” and bonuses. Very high executive pay packages have not transformed Thames Water, nor are they likely to do so.
In the case of water, the appropriate domain to start with is the catchment, and this requires a catchment regulatory approach.[3] It needs one entity to set out the catchment plan, but it does not require one monopoly to carry out all the catchment functions.
There is a clear structural case for having well-defined and stand-alone utilities. There is little or no structural case for having complex ownership entities, including holding companies. In the case of Thames Water, its holding companies in Kemble, in all its myriad of related entities, is neither necessary nor desirable as the owner and controller of Thames Water. At the very minimum, Thames Water should be completely separated out from Kemble. Owners should be required to invest in core utilities directly, and not via intermediaries.
The greater the structural clarity, the less the necessary conduct regulation.
Thames Water is too big, and it needs breaking up
It is widely recognised that Thames Water is too big to be effectively managed – too big in scale and too big in the multiplicity of functions. It is the largest water and sewerage utility in Europe, spanning not just the capital city but almost all the activities in the catchment, with around 16 million customers.
Failure by Thames Water would be (and arguably already is) an economy-wide event. Its board, even if filled with the best possible managers, cannot hope to grasp the detail of this vast operational domain.
The easiest split is into two separate entities: “London Water”; and “Greater Thames Water”. Both would be stand-alone, with no holding company above them.
A single London-only water company would have not just the clear focus on the capital city, but be able to interact with other London-defined regulatory and government structures, and other London-defined utilities, like Transport for London and the successor to London Electricity. London’s Tideway Tunnel fits into this area definition.
As a stand-alone London utility, “London Water” would not be independent of the catchment, whilst Thames Water in its current form is described over the full catchment. Catchment integration would remain important, but this could and should be transferred to a catchment regulator, with system duties and obligations and the catchment system plan.
Note that the Rhine and the Danube are not covered by single monopoly utility companies. The European model is very much a city-based utility model.
The sewerage capital expenditure programme should be taken outside the normal five-year periodic review framework
The new challenge for Thames Water and the other water companies is the sheer scale of the investment programmes over the coming decades. There needs to be a major overhaul of the sewers and the sewerage systems, and an overhaul of water supply provision, given the changing climate and population characteristics, and the emergence of new challenges, like forever chemicals.
These large capital expenditure (CAPEX) programmes will require access to funding and finance and appropriately focused management.
The current way of paying for all this – and hence the funding – comes from a single billing by the incumbent utility (or two bills where there are separate water and sewerage utilities). Ofwat agrees the business plans of the companies at the periodic reviews, and then sets the allowed maximum charges to customers: the funding.
Current affordability considerations, and political and customer resistance to higher bills, limit the amount of funding, and there is little possibility of additional government funding from taxpayers.
Thames Water, with 80% gearing, will have to raise equity to finance the sewerage requirement (and the proposed new Abingdon Reservoir). To expand its balance sheet with equity, Thames Water will face a loss whilst Ofwat continues to set the cost of capital as the weighted average cost of capital (WACC) – and hence below the marginal cost of equity. It is a very unattractive proposition for existing owners. It is hard to blame them for being reluctant to invest more under these terms.
With these considerations in mind, one option is to separate out the sewerage CAPEX from the day-to-day businesses of Thames Water, and to provide a new special-purpose vehicle (SPV) to carry out this programme over the next decade or more. There would be a special regulatory framework, with a new and separate balance sheet and separate regulatory asset base (RAB) mechanism for the gradual recovery of the costs over the full period. This would greatly lighten the finance role of the Thames Water balance sheet.
Sewerage and water can and should be separated
There are two current models in the water industry: combined sewerage and water; and separate sewerage and water. There is no compelling evidence that the integrated model is better than separated services, and the synergies are not that large. More importantly in the Thames Water case is that it already has a massive agenda for its board and its balance sheet, and a separation of the two functions would increase focus.
A separate sewerage business would have the challenge of the major upgrade, and the reconstruction of a system fit for the twenty-first century.
The full break-up model would see London separated from the surrounding greater Thames catchment area, and then for both “London Water” and “Greater Thames Water” to be split into two. The sewerage upgrades would be separated further through the SPVs proposed above.
The new successor companies should be floated so that there is listed equity
The experiment with highly geared utilities over the last 30 years has been at best mixed. In water there is some evidence that the three remaining quoted companies have performed better than some of the privately held companies, and some of the biggest failures over the period – in Yorkshire Water, Northumbrian Water, Southern Water, South East Water and Thames Water – have all been private.
Listing brings with it the more transparent corporate governance and accounting, plus the daily evaluation of the companies by the equity markets. Rights issues are more transparent and open to all investors. The remaining listed companies do not have elaborate and complex holding companies (although Pennon owns a series of water businesses).
Listing provides a clean prospect of exit for investors.
The only credible route to this outcome is special administration
Breaking up Thames Water could be achieved by the existing owners. They would start from the perspective of their particular interests and the debt structures and covenants in place. Separating out the interest between the parties would be at best a long and protracted process, especially in the case of Thames Water where there are losses to be allocated, notably amongst bond holders with the multiple and complex covenants on the debt which may be in place. Hold-ups would be an obvious strategy for some of the parties to gain individual advantage. It is therefore not surprising that there are few, if any, examples to draw upon of this bottom-up negotiated approach.
The alternative is to impose a break-up on Thames Water from above and from the outside. This could, in principle, be done via an inquiry by the Competition and Mergers Authority (CMA), but again the way this works would be long and complex, and the CMA has a narrow competition remit as its starting point.
The better option is to bring in the special administrator. In placing Thames Water into administration, and with the “special” dimension being guided by the overall public interest, rather than narrowly those of the existing owners and creditors, the special administrator could start with a blanker piece of paper.
There are two objections to using the special administration process. The first is to demonstrate that Thames Water has violated its licence obligations. Since, as noted, these obligations are very broad and ill-defined, this is not straightforward. Nevertheless, the existing owners will have an interest in getting the maximum subsequent sale value for the successor companies, and the bond holders will have an interest in their positions being protected, even at the cost to the equity holders.
It would be a brave set of owners to contest special administration, given the history, the performance and the likely unwillingness to put new equity into the business if and when Ofwat declines to grant Thames Water the 40% price increase it is demanding. Not even Railtrack resisted being put into administration.
The Railtrack example may be used nevertheless as an argument against special administration. It turned out to be complicated and very expensive, especially as suppliers used the period of administration to put in invoices for costs that could be argued to have been widely excessive. But this is not the right conclusion to draw, for several reasons. The administrative procedure for Railtrack was distinct, the management of the supply chain gives many lessons about how not to do this dimension of administration, and the continuing revenues from customers were distinct, less solid and more complicated than in the water industry. Railtrack is in fact a really useful case study on how not to handle administration, rather than an argument against special administration of Thames Water.
The interests of foreign investors
The reluctance to press the special administration button in the Thames Water case comes from a fear of upsetting foreign investors. This is a very serious concern not just for Thames Water and the wider water sector, but for investment across the infrastructures and in net zero.
The UK has got itself into a position of relying on the kindness of strangers to allow its citizens to live beyond their means, and for almost all financing of investment across the economy. There are two reasons for this exposed state of affairs: net saving (after capital depreciation) has been negative for over a decade; and the current account of the balance of payments is in serious deficit, with imports exceeding exports. UK investment is largely foreign investment; and the trade deficit requires a capital inflow to offset the current-account deficit. That capital inflow comes in a variety of forms, including the selling-off of many British companies to foreigners – selling the family silver to live beyond our means. The government is a net dis-saver, for much of industry profits equal dividends, with the consequence of few retained earnings, and households do not save enough for their pensions. It is not just in the water industry that profits equal dividends; it is quite general across industry.
Bearing in mind the sheer scale of the net zero energy investments in new generation, new grids and networks, the required investments in communications and transport, to which water is additional, upsetting foreign investors is a risky business. For foreigners, investment is a voluntary activity, and there are plenty of other places to invest in other than the post-BREXIT UK.
In the case of net zero, the situation is now one in which most of the companies in the energy sector are foreign-owned (directly or indirectly), investment is from foreigners, and since the UK lacks the relevant supply chains, the foreign-owned companies borrow from foreigners in order to pay for the necessary equipment and other supplies from foreigners. This is the reality of the situation the UK now finds itself in.
The question is: would putting Thames Water into special administration and splitting it up lead to less and more expensive investment by foreigners in the UK generally, and the infrastructures in particular?
In answering this, it is important to recognise just how much damage ever-greater conduct regulation has caused already and how far the valuations of the owners’ stakes in Thames Water have already been written down. It is highly likely that further conduct regulation of dividends would make matters much worse. Yet more conduct regulation is what is being demanded by politicians and increasingly by regulators. Both government ministers and their opposition shadows demand more conduct regulation of executive salaries, dividends, more fines and even imprisonment for directors, and ever-more-detailed interventions.
Given that the status quo is not acceptable, those who favour more and more conduct regulation should explain why this does much other than to raise the cost of capital and encourage foreign investors to look elsewhere. Less rather than more conduct regulation is needed. Structural regulation offers a way out of this mess.
The critical feature for regulatory credibility of interventions in Thames is that the rules that have been established for failing utilities should be followed. These “rules of the game” are known to all investors, and these rules include those for utility failures. The only uncertainty has been about whether they will ever be applied. Not putting Southern Water into administration may have misled investors to believe that this mechanism is never going to be applied. It would be very useful all round to clarify this now. If Thames Water is not now put into special administration, it would be reasonable for all investors to assume none will ever be placed into this. The end result is more and more conduct regulation. It is not clear how this could possibly be in the interests of the investors, unless they assume that they can capture the regulators.
The transition requires guarantees but should have little or no net costs
It has been claimed that a collapse of Thames Water, with or without special administration, would impose significant costs on the Treasury and hence on taxpayers. A figure of £5 billion has been widely quoted.
The Treasury comes into play for several reasons. The first is that the state stands behind all the utilities, both for the practical reasons that no government can allow the services not to be provided, and second via the duty to finance functions. Whilst Thames Water is being sorted out, the business needs to be continued. In this continuation, the supply chains need to be kept in play and this will require contractors to know that their contracts are guaranteed. Finally, there are costs associated with the special administration process and the sale of the subsequent businesses, plus any legal costs from challenges by the existing owners, debt holders and possibly also other interested parties, notably on the environmental and land-use sides.
Much of this is exaggerated. The key thing to remember is that the special administrator has an immediate source of cash – the bills from customers. As in all utilities, the marginal costs are way below the average costs, right down to the point where the RAB is worth just £1. That will cover the day-to-day running costs. There are also the costs of the capital programmes under way. However, provided that these programmes are efficiently conducted, the costs go in due course into the RAB, and so any costs are interim and then reclaimed via the subsequent RAB pass-through. Finally, in making good the failures of Thames Water, and in particular the shortfalls in capital maintenance and CAPEX committed in PR19 (the 2019 periodic review), this is a claim against the existing investors, and will be reflected in a discount to the RAB in the eventual sale value.
Taken together, the residual costs are very limited, and the guarantees to contractors and the supply chain can be recovered and then passed on to the new successor companies. The new owners after special administration get the RABs apportioned out to them, but they will pay less than the RABs because there will be the impairment costs and the obligations to make good on the failures on capital maintenance and the PR19 required CAPEX. The difference between the value of the RAB and the sale price will be the losses imposed on the existing owners, and out of this difference any costs for temporary guarantees to the supply chain and other administration costs can also be claimed.
The net costs over time to the Treasury should therefore be close to zero.
The environmental improvement CAPEX can be paid for over 10–15 years
The scale of the necessary expenditure on the sewerage systems across all the catchments is significant and, as noted above, this will not easily fit into PR24 and a five-year framework. Thames Water’s demand for a 40% price increase and a cap on environmental fines reveals both the perilous state it has allowed itself to get into and the lack of reality in thinking that current customers are willing and able to absorb this in the second half of this decade.
It also reflects a lack of realism about how long the improvements are going to take. Investing on the scale necessary to clean up the rivers will require a step change in the supply chain, at a time when there are also planned major increases in housebuilding, expansion of the electricity network and generation infrastructure for net zero, and increases in home insulation and home heating. There are also the costs of developing a credible and significant electric car charging network. Suppliers will be in great demand.
The supply chain will need several years to ramp up and it will depend upon the credibility of a well-designed and long-term investment horizon, without the risk of political stop–starts. It needs to be everything that HS2 has not been. Design and implementation plans need to be set and fixed, and not subject to repeated change.
This requires a new approach to the system planning of CAPEX over a decade and more, and a radically different regulatory control mechanism. This has been managed for the Thames Tideway, although the broader sewerage investment will have rather different project characteristics.
There should be a 15-year CAPEX programme, taken out of PR24.
Moving in the medium term to a system regulation model
In the medium term, splitting up Thames Water introduces greater opportunity to move towards a system regulatory model.[4] With a system regulator, and a system plan, many more of the activities of the incumbent water companies can be opened up to competitive bidding, allowing a whole new set of players to contribute to improving our rivers, flood controls and water supplies. After special administration and break-up and a retreat from conduct regulation comes the greatest opportunity of all – a modern twenty-first century reworking of the contract between the state and the market, with competition promoted through the catchment, coordinating the companies, reducing costs, and offering lots and lots of innovation.
But for the meantime, Thames Water should be put into special administration, broken up, and a long-term ring-fenced capital programme over a 15-year timescale should be put in place asap.
[1] See Dieter Helm, “What to do about Thames Water”, August 2023.
[2] See Dieter Helm, “Thirty years after water privatization—is the English model the envy of the world?”, The Oxford Review of Economic Policy, 36: 1, January 2020, eds D.E. Garrick, M. Hanemann, and C. Hepburn
[3] For more on this, see Dieter Helm, “Floods, water company regulation and catchments: time for a fundamental rethink”, March 2020; and “Water: A New Start”, October 2022.
[4] See Dieter Helm, “The Systems Regulation Model”, February 2019.
