There are many reasons why nationalising Thames Water would be a bad idea. Long gone are the days when governments had the competence to run major companies. The current examples are mostly dire. Nationalisation of the railways resulted in immediately conceding to the rail unions’ pay demands and freezing prices. CalMac, Scotland’s ferry company, is so badly run that it would be a challenge to make it worse. Northern Ireland Water and Scottish Water hardly encourage enthusiasm. While Thames Water is very badly run in the private sector, it is not at all clear it would get any better in the public sector.
Nationalisation often happens when there are no other alternatives. Steel is a case in point. But in the case of Thames Water, there is a failure regime already in place, designed to deal with the very circumstances Thames finds itself in. It is called Special Administration. The company is placed in a Special Administration Regime (SAR), and then others are invited to bid for it. Special Administration would bring forward bids, in whole or in part, with new investors and new managers.
It is because Special Administration is the obvious answer that the leading bondholders have put up such an expensive fight against it. They know that the Special Administrator has no duty to protect the excessive debt that has been built up. The leading bondholders understandably want to protect their debt, get their interest, get their capital back and, better still, make an additional return with an eventual sale of the equity. Apollo Global Management, Silver Point Capital and Elliott Management are masters of distressed-debt situations. They are not specialists in the water business, and they are not the obvious investors for a long-term rebuilding of Thames Water and its essential services.
To try to head off Special Administration, the leading bondholders have hired a plethora of lawyers, consultants and PR advisers, racking up tens of millions of pounds of costs. An obvious throw of the dice is to scare the Treasury into believing that an SAR would be very costly to the state. Step forward Teneo with the claim that an SAR process would cost around £4bn, notwithstanding that it is almost inconceivable that the Treasury would not recoup any and all the costs of an SAR in its privileged position as the first in the creditor leaderboard, and Thames Water is worth a considerable sum, given the revenues from customer bills and the central role of the regulatory asset base (RAB).
Contrary to arguments made, Thames Water is very different from Railtrack and its successor Network Rail. Network Rail is loss-making and requires long-term subsidies, the more so now that the fares are being politically interfered with and, hence, reducing revenues. Thames is a completely different case – hence others wanted to bid for it but were deliberately excluded by the Thames Water board in favour of KKR, which then walked away.
In putting forward the London & Valley Water (LVA) “offer” to OFWAT and the Treasury, Thames Water’s lead bondholders must have thought they had scared off Reeves and Starmer from an SAR. So confident were they that they actually asked for the environmental requirements and associated fines to be watered down and kicked into the longer grass. They want relaxed regulatory rules – relaxations that other bidders, when KKR was given the sole bidder rights, did not necessarily require. With the arrival of Burnham, a sense of panic must have set in. Burnham thinks the natural owners of Thames Water are public not private.
Trying to head off nationalisation appears to be the new priority of the lead bondholders. They could make the case that nationalisation is a bad idea because¸ they could argue, the company would be badly run, highly politicised, have customer bills manipulated for short-term political reasons, and be subject to public sector comprehensive spending reviews when it comes to access to finance. That would be a credible case to make.
But they didn’t. Instead they hired consultants to try to demonstrate that nationalisation would be costly, and even more costly than DEFRA had calculated.[1] Their consultants duly obliged, claiming that nationalisation of the whole water industry in England (of which Thames Water comprises over 25%) would cost £144bn, up from the £100bn that DEFRA had come up with. Presumably the lead bondholders preferred this line of attack on nationalisation because any focus on whether the state could run Thames Water well would inevitably put the spotlight on just how badly run it has been and arguably still is by the current owners and managers, and notwithstanding that the main managers have been paid retention payments and golden hellos. (More on this in a separate paper to follow.)
The trouble with the “big number” approach is that it is open to serious challenge. Put simply, whatever the correct number, neither £100bn nor £144bn is the right answer.
As was the case with DEFRA’s number, the consultants start with the RAB – and in particular the regulated capital value (RCV) – in the regulated accounts and then add on considerations of a premium to the RCV, on the basis of past transactions in water typically being at a premium of 10% to the RCV. Then they add on the costs of investment up until 2030, i.e. the envelope of investment in the current periodic review. It is therefore RCV + capital expenditure (CAPEX) (and the suggestion that this is an underestimate because of the past premia paid for water assets). There are no subtractions for the failures of this company: no company in a competitive market could expect its value not to be seriously dented by poor performance, and its failures to meet environmental law, leakage targets, together with ongoing poor maintenance. There are no benefits added back in, not least for the revenue streams from customer bills.
The consultants do not mention the words “Thames Water” anywhere in their report, other than to acknowledge that they are being paid for by Thames Water’s lead creditors. Presumably these creditors will claim that none of the costs of their report fall on customers. Throughout this paper I shall refer to Thames as the example through which to understand the £100bn and the £144bn.
The consultants do not separate out the economic costs and benefits, and the economic costs to the government, from the costs as they show up in the national income accounts. But they should.
Let’s start with the economic costs and benefits. These should be capitalised at zero if the company is properly priced, and after the full costs of the SAR have first been recovered. To be clear, the economic number is not £100bn or £144bn, but £0.
The reason why the answer is zero is that, as a regulated monopoly, Thames Water (and the other water companies) are economically viable. They are indeed guaranteed, through the regulator’s duty to finance functions, to recover all their efficient costs, including a return on their capital – and, indeed, an excess return on their RABs via the application of the weighted average cost of capital (WACC), which by definition over-rewards debt, being an average over both debt and equity. The point of privatisation was to embed the concept of customers paying water and sewerage bills sufficient to meet the duty to finance functions.
I will come back to the “efficient” bit when considering the claim that the premium in past financial transactions has been 10% and that a fair value would therefore mark up the RCV by 10% for Thames Water. This is presumably an argument that the lawyers Thames has hired to oppose nationalisation and gain the maximum price would pursue.
Suppose for a moment that the RCV for the industry is £100bn, and for Thames Water its RCV is as recorded in its regulatory accounts. Suppose the government pays this amount in full for the company’s RCV. Under the current arrangements, the interest on the RCV (actually the WACC) is paid on the total amount, and there is a gradual repayment of the debt itself over time. What happens is that the government swaps an economic asset – the RAB, as measured by the RCV – for gilts, and gains a profit on the gilts that is equal to the difference between the interest on the gilts and the interest on the RAB. The former is bound to be lower than the latter (except possibly in an extreme economic crisis, when bond markets might lose faith in the government’s ability to honour its debts).
Buying the RABs is, in economic terms, a profitable opportunity for the government. Even if the government then decides to pass on to customers the difference between private and public debt costs, the net result would still be zero or above.
How, then, can a profitable acquisition for the government be described as a cost of £100bn? (I will come back to the CAPEX and the extra £44bn in the consultants’ report.) The answer is that the consultants are applying the government’s accounting rules. In essence, the government accounts are really all about cash: cash-in and cash-out. When the water companies were privatised, the government of the day counted the sale proceeds as a cash boost that reduced the overall government cash deficit. They did the same for North Sea oil revenues and for selling council houses. It was what Harold Macmillan famously described as “selling the family silver” to fund day-to-day spending.
This was, at the time, not a good way of presenting the government’s longer-term fiscal position, and so it has proved. The sale of assets reduced ongoing income. Now the reverse is happening. Buying an asset is treated as a liability, full stop. There is no corresponding benefit from owning the asset in the accounts. Instead of a proper balance sheet, where the asset “Thames Water”, with its revenue stream, is set against the liability, gilts replacing private debt, there is simply a ledger of cash-in and cash-out.
To see how absurd this is, imagine that the government nationalises Thames Water at its full RCV, and then a few years later (or even the next day) it re-privatises it. Suddenly the government would have, as at the original privatisation, a cash inflow from the sale. Over the two transactions, the nationalisation and then the re-privatisation, the net effect should be zero (ignoring the transaction costs and the temporary reduction in the state sector of interest payments on the debt).
The next questions are what the correct value of the RABs is, whether the RCVs recorded in the accounts properly represent the true RABs, and whether the current debt of Thames Water is a correct mapping of the RCV value. In favour of the Thames argument, and the one the consultants have used, OFWAT has deliberately encouraged this view. As OFWAT states in its final determination of the 2024 periodic review (PR24), ): “These arrangements include: […] the regulatory commitment to the RCV, which has been used to calculate the net stock of investment contributed by investors over successive price controls, and commitment to remunerate cost over- and under-spends subject to cost sharing rates.”[2] There is, however, no formal statement of the link between the RCV and the duty to finance functions, which is the mainstay of protecting investors from ex post expropriation and, in particular, from the time-inconsistency problem. The RABs are not set in law and they are not in the licences. Financing the functions is however explicitly linked to the functions, and from this must follow their efficient delivery. It is an important consideration to ask whether Thames Water has in fact carried out its functions, and in particular met its licence conditions. It is not unreasonable to argue that the obligation to honour the RCV is conditional on checking that the company has indeed met these conditions, and it is not hard to argue that it has failed to do so. It admits as much.
In Thames Water’s defence, it can reply that it is up to the regulator to check at each periodic review that any additions to the RCV are validated, and the functions have been fully delivered, and OFWAT carried out a periodic review very recently. Though the company has broken, in spirit and possibly in letter, the obligation to deliver the functions, OFWAT is open to the challenge that it has failed to fully enforce this on Thames Water.
As originally intended, the RAB was designed to represent investment that is being paid for by future and not current customers. It is possible to calculate what the RAB value today would be on a pro-forma basis, if genuine asset enhancements only are added, and from which the run-off is deducted. The answer almost certainly would not be the current value of the RCV. On the contrary, what has in fact been happening is two things: capital maintenance has been called investment when it should be an operating cost; and the assets themselves have not been fully maintained. The pro-forma RAB would not equal the RCV – it would be substantially less. Hence the cost of nationalisation should be less than the RCV for Thames Water.
If Thames Water was nationalised, there is therefore an argument to say that what the government should compensate investors for is not the RCV as stated but the “true” RAB. The current owners should be paid the RCV minus the costs of making good on their failures. By its own admission, Thames Water has not been well managed (it is not an “efficient” company) and its assets have not been adequately maintained. It has engaged in polluting activities through spills outside its allowed parameters, and indeed it did not even bother properly measuring what its sewers have been discharging. It has not therefore met the requirement of efficiency in delivering its functions, and hence the obligation of the regulator is not to finance its actual costs, but what its costs would have been had it been properly managed.
It is difficult to ascertain how far the company has fallen short of its obligations under its licence and more generally, and any calculation is hindered by the very vague and poor drafting of its (and other water companies’) licences. There are two ways this could be handled. The first is for OFWAT and the government to try to calculate the shortfalls, and then deduct these from the RCV; or, under Special Administration, it could oblige new buyers to make good on these requirements, including the environmental ones, and then the bidders will value the company at the RCV minus the costs of the remedial works. Other bidders did previously indeed indicate that they would meet the environmental and other obligations, but they were not allowed by the Thames Water Board to proceed with their bids. OFWAT should now be considering again these early “offers” against the LVA document, and canvassing potential buyers generally about these conditions in adjudicating on the LVA offer document and comparing with the SAR option.
Assuming that the shortfalls in asset maintenance and in environmental performance are non-trivial, the nationalisation cost to the government would be RCV minus the costs of remediation. That would value Thames Water at potentially a substantial discount to that implied in the industry calculations of £100bn and £144bn.
The next question is whether the bondholders in Thames Water should receive a premium (the 10%) or a reduction to reflect a damaged reputation over and above the costs of remediation, as would be the case in a private sector failing company. Thames Water is a failing company. Its reputation is extremely low. Potential buyers will know this and, in a normal competitive market, they would want a discount. This might have been one of the factors that persuaded KKR to withdraw. Reputational damage must have been a consideration.
It is hard to think of any failing company in a competitive market that argues that no account of its failures should be taken in arriving at a “fair market value”. On any private sector comparison, there should be a discount to the RCV for Thames Water’s reputation, even if the past average for the industry is a premium of 10%.
To this correction should be added one more. Ask yourself why companies have in fact traded at a premium to RCVs. The answer is that their returns have been expected to exceed those necessary to remunerate the RCVs and provide a normal return on the day job of doing the operating expenditure (OPEX) and CAPEX. Looking backwards, the vast bulk of “outperformance” has come from the overstatement of expected interest rates at periodic reviews by OFWAT for the successive five-year periods up to 2020, the excess of the WACC over the cost of debt, and financial engineering. These premia are therefore substantially explained by OFWAT’s failures to regulate the businesses properly.
Should Thames Water, or others, be rewarded with an additional mark-up of 10% on its RCV in nationalisation? The answer is unambiguously “no”. There is no requirement for OFWAT to carry on with regulatory failures in the future because it has done so in the past, and nothing in the duty to finance efficient functions dictates that it should.
The argument points in the other direction. There should be a negative adjustment for Thames Water’s reputational failings and for its remediation requirements. This is actually recognised by Thames Water in the LVA document. The leading bondholders paid a discount for the debt that had become distressed and are offering write-downs in the LVA document. They have conceded the point that the value of their investments, their debt, is a variable and not locked into the cement of the RCV.
It is not only Thames Water’s bondholders who implicitly accept a below-RCV value. There are several transactions that the consultants avoid mentioning which have been below the value of RCVs. These include Welsh Water, Southern Water and, very recently, Kelda, and the previous equity owners of Thames Water wrote off their entire stakes. A 10% premium would have been an equity return. They valued the equity at zero.
The reality is that Thames Water’s debt trades at a significant discount. The distressed-debt investors are reported to have bought in at around 75p in the pound. They are now offering to take a cut from the 100p face value towards or even slightly over what they actually paid, though the exact write-down relative to the purchase cost is hard to fathom.
This is further complicated by the fact that the bondholders were, as noted, in effective control when Thames Water’s board accepted a short-term £3bn loan, upon which they set a headline interest rate of 9.75%. There are also fees attached to the loan, so it is hard to work out how much higher the implied interest rate is.
This raises an additional question as to how the costs and fees that the bondholders have accrued are being paid for and by whom. There is a strong protestation that “customers will not pay a penny”. This may be correct, but it will only be known if there is full transparency of the accounts of Thames Water and what goes into the various headings, including operating costs. Press reports have suggested that the costs of lawyers, accountants and consultants may be as high as £300m, and some have claimed that they are, or will be, even higher.
The final question is whether the consultants are correct in adding the investment to the RCV to get from £100bn to £144bn. This is what the private companies are expected to deliver and make at least their cost of capital on delivering. In the company accounts, this investment is funded from customer bills. Some is immediately charged (the Thames Tideway costs are an example) and some is delayed on the pay-when-delivered basis. It is the latter for which finance has to be raised. If the companies are nationalised, then the state will presumably issue gilts equal to the amount not paid for by current customers, and then receive a run-off of the capital sums (paying off the gilts) plus a return from investors (paying the interest on the gilts). There is no economic cost to the Treasury, only a good business opportunity to finance fully funded investments.
There is a question as to whether the government, for general and short-term fiscal reasons, fails to do the necessary investment. There are good reasons to believe that it might, and then the £44bn is not only temporary, but also reduced. As noted at the outset, this is one reason why nationalisation is a bad idea, but it has nothing to do with the cost of nationalisation.
In conclusion, the consultants’ report and the various lobbying and PR exercises do not fully represent the economic costs of nationalising the water companies in England. There is a good case for adjusting the RCV total for a discount rather than a market premium in some cases, and discounting for the need for remedial maintenance and investment. This arrives at a number for RCV minus X, where X varies from company to company.
In the case of Thames Water, nationalisation could be achieved at a sharply lower valuation than the RCV, and that too could be adjusted to a lower number on grounds of “efficient functions”. The economic cost of this would be zero. In the muddle of government accounting, it is nevertheless likely that the government could even make a net profit through nationalisation of Thames Water, since the bills should exceed the costs of the business activities over time and gilts are cheaper than commercial loans.
No wonder the consultants focused on the costs of nationalising the industry in England as a whole, and left out any consideration of Thames Water, despite being paid by the company’s lead bondholders and Thames Water being 25% of the total England industry.
None of this suggests nationalisation is a good idea. It isn’t. But perhaps Thames Water has realised that any focus on its own value in nationalisation would highlight its numerous failings and why it is not worth its RCV. Special Administration would reveal that value in bids, and perhaps that is why the bondholders are so hostile to Special Administration. They fear the consequence would be a much greater write-off of their debt, and this may be why they have spent so much money on consultants and advisers to try to head this off. If Thames Water was bought from the Special Administrator in a competitive process, then there would be a market test of its value.
[1] Frontier Economics (2026), “Estimating the cost to the UK government of nationalising the English water companies”, report commissioned by Thames Water’s creditor group.
[2] Ofwat (2024), “PR24 final determinations: Aligning risk and return”, p. 15.
