Sold Down the River: How Private Equity Loaded Britain's Water With Debt and the Regulator Did Nothing
In 2006, a consortium led by an Australian investment bank bought the company that supplies water and sewage services to sixteen million Londoners. They paid £8.5 billion for it - approximately £2.3 billion in equity and £6.2 billion of borrowed money. Over the following eleven years, they roughly tripled the company's debt, extracted billions in returns, and transferred a significant portion of their own acquisition borrowings onto the company's balance sheet through a subsidiary incorporated in the Cayman Islands. Then they sold their remaining stake in 2017, departing before the consequences became impossible to ignore. The company they left behind now carries statutory net debt of £17.6 billion. Its credit rating is junk. Its shareholders have written their stakes down to zero. The government has been preparing contingency plans for a statutory insolvency process that has never been used in the history of the privatised water sector. The water the company supplies has on occasion contained sewage. It discharged raw, untreated effluent into rivers and reservoirs for hundreds of thousands of hours last year alone. The regulator that was supposed to prevent all of this has just been abolished, its failure acknowledged by the government in terms that stopped only slightly short of a formal indictment.
This is not, fundamentally, a story about one rogue bank or one badly managed company. It is a story about what was always going to happen when a government sold regional monopolies to private investors, handed each of them a captive customer base that could not go elsewhere, and then relied on a regulator to substitute for the competition that can never exist when there is only one set of pipes. The Australian bank understood what it had bought. The customers had no choice. The regulator did not, in any meaningful sense, protect them. And the British government, across both parties and three decades, chose not to look too hard at the consequences until those consequences were visible in the rivers.
The Original Sin
The story begins, briefly, in November 1989 - because you cannot understand the private equity era without understanding the structural gift that made it possible.
Margaret Thatcher's government privatised the ten regional water authorities that year. The companies that floated on the stock market were sold debt-free. The government wrote off roughly five billion pounds in accumulated long-term liabilities, added what was known as a "green dowry" - estimated at around £1.5 billion - to help the new companies meet incoming European environmental standards, and provided capital tax allowances worth around £12 billion over the decade that followed. The total proceeds to the Treasury from the sale were approximately £7.6 billion, which sounds significant until you subtract the debts written off, at which point the net public benefit becomes rather harder to calculate. Within weeks of flotation, water company shares were trading at a twenty per cent premium to issue price. The analysts who had priced the sale had undervalued them.
The companies that entered the market in 1989 therefore possessed several structural advantages that no normal private business enjoys. They had no debt. They held a monopoly franchise over an essential service from which customers could not defect. They faced no competition, by definition - you cannot build a second set of water mains. And they faced a regulator - the Water Services Regulation Authority, known as Ofwat - whose job was to set the prices they were permitted to charge and oversee their investment obligations. The theory was that Ofwat, reviewing prices every five years, would act as a synthetic market. It would mimic the disciplines that competition would otherwise impose. What happens in practice when you ask a public body to do the work of a market it can never replicate is the subject of the rest of this article.
There was one feature of the original privatisation worth noting clearly. The companies were sold with legal obligations attached. They were expected to invest in upgrading the Victorian-era infrastructure they had inherited - the pipe networks, the treatment works, the sewage systems. Licences came with commitments. What those licences largely lacked was any mechanism to force companies to honour those commitments in preference to their other uses of cash. In time, this omission would matter enormously.
Enter the Infrastructure Fund
For roughly the first fifteen years, the water industry operated in a form its architects might have recognised. Companies were listed on the stock exchange, paid dividends to shareholders who were at least theoretically exposed to the risk of poor performance, and were subject to periodic price reviews that - whatever their other failings - operated in daylight. The system was not working well: investment was consistently below what the infrastructure required, sewage overflows were worsening, and Ofwat was already showing signs of the collaborative rather than adversarial relationship with the companies it was supposed to police. But the structure had a certain legibility.
What changed the sector fundamentally was the arrival of infrastructure funds in the mid-2000s. The insight driving this wave of acquisition was not complicated, and it was not secret. Regulated utilities with captive customers produce predictable, inflation-linked cash flows. Those cash flows can be used to service large amounts of debt. If you acquire the utility with borrowed money - secured against those future cash flows - you can achieve very high leveraged returns on whatever equity you actually contribute. The more debt you use, the higher the return on equity. The key to the whole model is that the cash flows are protected by the regulatory framework: customers cannot leave, Ofwat sets prices, and the government cannot permit water to stop flowing. The downside risk is carried, in the end, not by the owner but by the customer - who will pay whatever bills are necessary to sustain the service - and by the government, which cannot allow critical infrastructure to fail.
In October 2006, a consortium led by Macquarie Group - an Australian bank that had already applied this template to toll roads, airports, and utilities across several continents - acquired Thames Water from the German utility RWE. The enterprise value was £8.5 billion: approximately £2.3 billion in equity and £6.2 billion in third-party debt. Thames Water entered this new phase of its existence with a Regulated Asset Base of around £6.5 billion - the value Ofwat placed on its assets for price-setting purposes - and with net debt of roughly £3.4 billion. That gearing level was high, but arguably within the range that analysts considered manageable for an investment-grade utility.
Over the following eleven years, it would not remain there.
Following the Money
The mechanism Macquarie used was a whole-business securitisation. In essence: the operating company's future revenues were pledged as security against a large and growing pile of bonds, creating investment-grade debt at relatively low interest rates because the monopoly cash flows backing it were considered predictable and protected. Thames Water Utilities Limited - the regulated entity that actually operates the pipes and treatment works - sat inside this securitised ring-fence. Above it, Macquarie had erected a chain of holding companies, the most important being Kemble Water Holdings and its subsidiary Kemble Water Finance. The Kemble layer raised additional debt - riskier, more expensive, not covered by the regulatory protections of the operating company below - which was serviced exclusively by dividends passed upward from Thames Water Utilities. At each level of the structure, cash was flowing toward the owners. The question of what was flowing in the other direction - into the pipes, the treatment capacity, the sewage infrastructure - would become pressing much later.
This architecture is worth understanding in some detail because it explains precisely how what happened was possible. The securitised debt inside the ring-fence was packaged as safe, investment-grade paper, secured against the regulatory settlement and therefore, ultimately, against customer bills. Pension funds and institutional investors bought it. The Kemble-level debt was riskier but promised higher yields. And above both layers sat the equity investors - Macquarie's funds and their co-investors - who took dividends after servicing both debt layers, and who also charged management fees and arranged shareholder loans on which they received interest. Every layer was engineered to extract value. The infrastructure was the asset upon which the extraction depended.
A BBC investigation in 2017, drawing on analysis by the funding consultant Martin Blaiklock, established one of the more striking specific findings of the Macquarie period. When the consortium had bought Thames Water, they borrowed roughly the equity portion of the purchase price. Approximately £2 billion of that acquisition debt - the debt Macquarie's own funds had incurred to buy the company - was subsequently transferred onto Thames Water's own balance sheet, refinanced through bonds issued by a Cayman Islands subsidiary, with the interest thereafter paid not by the investors who had originally borrowed the money but by Thames Water's customers. The borrowings that Macquarie had used to buy the company became, in other words, the company's borrowings. The people who had not been consulted about the sale of their water service found themselves servicing the loan that funded it.
On dividends, the numbers are contested and the discrepancy is large enough to require explanation. Macquarie's own published account states that total distributions during its ownership amounted to £1.1 billion - £879 million in dividends and a further £277 million in interest on shareholder loans. The Financial Times, drawing on regulatory filings and company accounts, arrived at approximately £2.7 billion extracted over the Macquarie period, a figure supported by several independent analysts, including Frances Coppola, whose analysis described the Kemble structure as having "leeched an astonishing £2.7bn from the regulated utility." The gap between these figures reflects a methodological difference - Macquarie's account covers distributions to its own funds, while the larger figure captures group-level cashflows including the full scope of the holding-company structure - and it has never been resolved in a form the public can straightforwardly verify, in part because the corporate structures were specifically designed to make that kind of verification difficult. The BBC's analysis of Macquarie's returns estimated annualised gains for the firm and its investors of between 15.5 and 19 per cent - roughly twice what one would normally expect for a low-risk infrastructure asset. Macquarie disputed this, pointing to a 12.3 per cent internal rate of return for its European infrastructure fund overall.
What is not disputed is the trajectory of the balance sheet. When Macquarie arrived in 2006, Thames Water carried around £3.4 billion of net debt. When it sold its remaining stake in 2017, that figure had reached approximately £10.8 billion. The Regulated Asset Base had grown over the same period from £6.5 billion to £12.9 billion, partly reflecting genuine investment and partly the normal regulatory process by which asset values are uplifted. The pension deficit that stood at £18 million when Macquarie bought the company had grown to £380 million by the time they left. The buyers of the 2017 stake - a consortium including Canada's OMERS pension fund and the Kuwait Investment Authority - purchased a company that had been comprehensively re-engineered. The returns had flowed. The debt remained.
This is not, it should be said, a straightforwardly illegal account. The structures were legal. The dividends were permitted under the licence conditions as Ofwat had defined them. The debt was issued on market terms. Ofwat was aware of the financial structures; it raised concerns periodically and was, periodically, satisfied by the responses it received. This is precisely what makes the story important. What happened at Thames Water was not a fraud. It was the predictable consequence of a model in which the sole discipline on private owners of a monopoly was a regulator that lacked either the adversarial instinct or the tools to impose the public interest over the interests of owners who understood the structure considerably better than the body supposed to police it.
The Regulator That Wasn't
To understand how Ofwat failed, you need to understand what it was asked to do and why the task was, in the form it was set, essentially impossible.
Every five years, Ofwat conducts a price review. It sets the revenues water companies are allowed to collect, the return on capital they are permitted to earn, and the investment obligations they are required to meet. The process is elaborate, formally adversarial, and produces thousands of pages of technical documentation. Companies submit business plans. Ofwat scrutinises them. There are multiple rounds of challenge, draft determinations, and rights of appeal to the Competition and Markets Authority. The framework looks rigorous. Its fundamental weakness is that Ofwat's ability to assess company plans depends critically on the information companies provide, and the companies providing that information are the same entities Ofwat is trying to regulate. The information asymmetry is not a correctable detail of the system. It is the architecture of the system.
Water companies operate complex underground infrastructure networks that only they can fully observe. They know what their pipes look like, what their actual investment costs are, what their genuine environmental performance has been, and what they have done versus what they have told the regulator they have done. Ofwat, operating on an annual budget of around £30 million against an industry generating approximately £12 billion in revenues, was structurally dependent on company-submitted data to perform almost every function. The National Audit Office, in its damning April 2025 report on the sector, found that the Environment Agency had independently inspected only one per cent of the 8,780 environmental actions that companies had reported completing over five years. The gap between what companies stated they were doing and what any external body could verify was not narrow. It was nearly total.
The consequences were systematic and predictable. Researchers at the University of Greenwich documented what they described as companies overpromising and underdelivering on investment through successive price-review cycles, claiming "capital efficiency" savings - purporting to deliver contracted investment for less than Ofwat had allowed in its price settlement - and then using those claimed savings to justify enhanced dividend payments. Ofwat's response was to incorporate expected efficiency gains into the following review's allowed expenditure, rewarding companies for demonstrating they could do things cheaply and penalising honesty about actual costs. Over time, the incentive structure selected for sophisticated regulatory management rather than infrastructure maintenance. The companies that performed best in the price review process were not necessarily the ones that invested most. They were the ones that best managed Ofwat's information.
The enforcement record tells its own story. Between 2010 and 2024, Thames Water had been sanctioned ninety-eight times, accumulating fines of £175 million - a sum that sounds significant until set against the billions extracted in the same period. Yorkshire Water: ninety-four sanctions, £109 million in fines. Anglian Water: seventy-four sanctions, £6.2 million. Severn Trent: eighty-two sanctions, £8 million. These are the penalties accumulated by operators of some of the most profitable monopolies in the country, and they are, measured against the cashflows involved, not far above rounding errors. Ofwat's largest-ever single penalty was the £122.7 million levied on Thames Water in May 2025 - £104.5 million for wastewater breaches and £18.2 million specifically for dividend payments made in violation of licence conditions. Announcing it, Ofwat's chief executive David Black described it as "the first time we have used these [dividend] powers." The first time. In 2025. On a company whose dividend extraction and licence non-compliance had been visible to competent analysis for the better part of a decade.
There is also a revolving-door question worth naming directly. Cathryn Ross served as Ofwat's chief executive from 2013 to 2018. She subsequently became Thames Water's director of strategy and regulatory affairs, and was later appointed its interim co-chief executive. When she appeared before the Environment, Food and Rural Affairs Committee in July 2023 - after Thames Water's financial crisis had become fully public - she declined to apologise for the regulatory settlements reached during her tenure at Ofwat, the settlements that critics argue had permitted the debt accumulation she was now overseeing from the other side of the desk. This single fact does not constitute proof of systemic regulatory capture. It does illustrate, with uncomfortable clarity, the professional culture in which the boundary between the regulator and the regulated was always somewhat porous.
The National Audit Office's April 2025 report placed the structural failure plainly on the record. Regulators had failed, it concluded, to deliver a trusted and resilient water sector. Ofwat was concerned about the financial resilience of ten of the sixteen major companies it was supposed to be overseeing. The rate at which water mains were being replaced had fallen to 0.14 per cent per year over the four years of the most recent price-review period - a rate that, if maintained, would mean the entire network was renewed once every seven hundred years. The pipe network inherited from the Victorians is, in the most literal sense, being left to decay while the companies licensed to operate it extract dividends and service debt.
What the Sewage Tells You
In March 2025, the Environment Agency published its Event Duration Monitoring data for the preceding year. Untreated sewage had been discharged into England's rivers and coastal waters for a total of 3,614,427 hours during 2024 - across 450,398 separate spill events. That is the equivalent of more than four hundred rivers running with sewage every hour of every day for an entire year. The figure was a new record, though only in a technically narrow sense: 2024 was the first year in which all of England's approximately fourteen thousand storm overflow points were monitored, compared with around ten per cent a decade earlier. As monitoring coverage has expanded, the apparent scale of the problem has grown - not necessarily because conditions are deteriorating faster, but because more of the deterioration is now visible.
Thames Water, serving sixteen million customers across London and the Thames Valley, recorded 523 pollution incidents in 2024 and was rated the worst-performing company in England in the Environment Agency's Environmental Performance Assessment published in October the following year. It received the lowest possible rating - one star - among the nine major providers assessed. Its serious pollution incidents more than doubled in a single year, rising from fourteen in 2023 to thirty-three. In January 2024, it discharged raw sewage into a chalk stream feeding the River Ver near St Albans for more than one thousand consecutive hours. The River Thames itself - on whose banks the Houses of Parliament stand and whose condition is in some sense a statement about how Britain manages its essential infrastructure - has been contaminated by sewage at a rate that would have constituted a public emergency in any earlier era of public administration.
The relationship between the financial engineering and the sewage is not subtle. Infrastructure for managing overflow events - the storage capacity that holds untreated effluent when rainfall exceeds system capacity rather than releasing it into rivers - requires sustained capital investment. Companies that have maximised leverage and optimised their capital structure for dividend extraction have less headroom for that investment, and face stronger incentives to claim investment that has not been made, or to defer investment that was supposed to have been made, or to assert efficiency improvements that exist only in the data submitted to the regulator. The NAO found mains are being replaced at a seven-hundred-year rate. The sewage data shows what an infrastructure network looks like when it has been managed as a financial product rather than maintained as a public service.
This should not have come as a surprise to those responsible for overseeing it. Sewage overflows have been a documented problem for decades. The monitoring technology that could measure their duration and frequency was available, and was at various points required, under Ofwat's regulatory framework. Companies then argued successfully for monitoring requirements to be scaled back. The data that now makes the scale of the problem legible was data that the industry worked for years to avoid producing, until the political pressure became irresistible. What the full monitoring record reveals, in other words, is not something that suddenly got worse. It is something that was always there, and is now, for the first time, visible.
Not Just Thames
It would be convenient for the industry if the Thames Water story were exceptional. It is not.
Southern Water serves around 4.6 million people across the south and south-east of England. In July 2021, it pleaded guilty to 6,971 illegal sewage discharges, totalling 61,704 hours of untreated effluent released between 2010 and 2015 across seventeen sites in Hampshire, Kent, and West Sussex. The judge found not merely negligence but deliberate concealment - that Southern had structured its environmental reporting to hide the scale of the violations from the regulator. The criminal fine was £90 million, the largest ever levied on a water company in England, and it followed a £126 million Ofwat penalty package in 2019 for deliberately misreporting performance data. Southern had submitted false environmental returns to its regulator while simultaneously paying dividends that the regulator, receiving those returns, had no grounds to prevent. Who owns Southern Water since 2021? Macquarie Group, which acquired a roughly 62 per cent majority stake and has since injected approximately £1.65 billion in equity, with a further tranche arranged in 2025. The company carries around £6.2 billion in net debt, has been downgraded to junk by the credit rating agencies, and has new debt priced at close to ten per cent - a cost that flows directly, and in full, into the bills of the 4.6 million customers who, as always, have no alternative.
Anglian Water, which serves six million customers across East Anglia and the East Midlands, was taken private in 2006 by the Osprey consortium. The largest single shareholder is the Canada Pension Plan Investment Board, holding approximately 32.9 per cent, alongside IFM Investors of Australia. The pension savings of Canadian public sector workers are, in this respect, directly invested in the monopoly cash flows of English water customers - a transaction that has been legal, profitable, and entirely consistent with the model. Whether those Canadian savers would consider the sewage data, the infrastructure degradation, and the decades of extraction an acceptable trade-off for the returns being generated on their behalf is a question that nobody has thought to put to them.
Across the sector as a whole, the Financial Times calculated that the sixteen regional water monopolies paid out approximately £78 billion in dividends between 1991 and March 2023, adjusted for inflation - representing roughly forty-one per cent of the £190 billion spent on infrastructure in the same period. The companies were privatised with no debt. They now carry approximately £64 billion of it. The debt accumulation, in other words, has exceeded £64 billion in a sector handed debt-free monopoly franchises by the public in 1989. The £78 billion figure is sometimes rounded to £72 billion in political debate and cited at higher numbers still by campaign groups, depending on the methodology, the time period covered, and whether inflation adjustment is applied. Ofwat's own statutory dividend dataset, covering 1992/93 to 2023/24, records £52.7 billion in nominal terms. Any of these figures, set against a sector that began with no debt and a regulatory settlement designed to ensure sustained infrastructure investment, represents a remarkable transfer of value from customers to owners over thirty-five years.
What They Did Instead
The mechanism by which water company cash was directed toward dividends and debt service rather than infrastructure was not, in most cases, conspiracy. It was the rational response of private owners to the incentive structure they faced. Ofwat's price-setting system allowed companies to earn a defined return on their Regulated Asset Base. If they could deliver the required investment for less than Ofwat had allowed - or if they could persuade Ofwat that they had delivered it, which is not the same thing - the savings accrued to the company. Dividends could be paid from those savings. The review process rewarded companies for claiming efficiency without requiring rigorous independent verification that the claimed efficiencies translated into functioning infrastructure. Over five price-review cycles, the cumulative effect of this structure is a water network whose physical condition the regulator cannot adequately assess, maintained at a rate implying seven hundred years to complete a single cycle of renewal.
This is the heart of what went wrong, and it is worth being precise about it, because the wrong conclusion is easy to draw. The wrong conclusion is that private ownership of public services is simply bad, full stop, and that the answer is self-evidently to return everything to public hands. The right conclusion is something more specific and more damning: that private ownership of a natural monopoly - where competition is structurally impossible and the customer has no exit - requires a regulator capable of substituting for the discipline that competition would otherwise impose. That means a regulator with genuine enforcement powers, independent access to operational data rather than company-submitted self-reporting, adequate resourcing relative to the industry it oversees, and an institutional culture of adversarial scrutiny rather than collaborative negotiation. Ofwat was none of these things.
It was a body running on approximately £30 million a year set against a £12 billion industry, reliant on that industry's own data, conducting enforcement that was modest relative to the cashflows it was supposed to constrain, and staffed at senior levels by people who moved between regulator and regulated with a regularity that is not generally associated with adversarial oversight. The water companies were not competing with anything. The only discipline available was Ofwat's willingness to impose it, and that willingness was, over three decades, demonstrably insufficient to the task.
The Comparison That Should Embarrass Everyone
In Scotland, a publicly owned company called Scottish Water delivers drinking water and waste water services to 2.6 million households. It pays no dividends because it has no shareholders. It charges bills averaging roughly £107 per year less than those paid by customers of comparable English companies. Researchers at the University of Greenwich found that between 2002 and 2018, Scottish Water invested nearly thirty-five per cent more per household than the English private water companies - more investment, lower bills, no dividends, no infrastructure fund overhead, no Cayman Islands holding structures servicing acquisition debt. Scottish Water has reduced its debt since 2009. It is publicly accountable to the Scottish Parliament. Its performance is not perfect - no large utility's is - but the basic arithmetic of what it spends on infrastructure relative to what it extracts from customers is, in comparison to the English model, a straightforward illustration of what the absence of shareholder extraction looks like.
In Wales, Dwr Cymru - Welsh Water - has been operated since 2001 by Glas Cymru, a company limited by guarantee with no shareholders. Surpluses are reinvested rather than distributed. The structure has its critics - Welsh Water has faced enforcement action from Ofwat for misreporting leakage data, and its executive pay is not modest - but it pays no dividends, and what would otherwise be shareholder distributions is channelled instead into support for customers struggling with their bills. Not a perfect model. A demonstrably better one.
The contrast with Paris adds a further dimension. The French capital privatised its water in the 1980s, contracting Suez and Veolia to operate the system. By 2010, after decades in which costs had risen while investment had lagged, the city chose to remunicipalise. It created Eau de Paris, a publicly owned operator, and took back control of the system. Within a year, tariffs had been cut by eight per cent. A decade later, bills remained among the lowest in the region. The Transnational Institute documented 235 cases of water remunicipalisation across thirty-seven countries between 2000 and 2015. England and Wales - not Scotland, not Wales in the same way - are a genuine international outlier: the only major developed economy to have fully privatised its water and sewage systems and to have maintained that privatisation through thirty-five years of accumulating evidence about what it produces.
The standard defence of the English model, offered periodically by the industry and occasionally by ministers of both parties, is that private ownership has attracted investment that public ownership would not have funded. The total is approximately £190 billion since privatisation. This is true as far as it goes, and genuine improvements in water quality and leakage management over the period are real. What the defence omits is that the same customers who funded this investment also funded the £78 billion in dividends and the £64 billion in debt accumulation that accompanied it - that the investment was funded by bills, not by equity, and that customers also paid the entire financial overlay required to make private ownership profitable on top of the investment itself. Scottish Water, without any of that overlay, invested more per household. The defence, on inspection, is that privatisation delivered what public ownership would have delivered, at greater total cost to customers, while generating substantial returns to owners. As commercial propositions go, it is an extraordinary one.
The Reckoning
By the spring of 2025, Thames Water's situation had reached a point where the language of financial analysis and the language of crisis had become indistinguishable. The shareholders of the consortium assembled after the 2017 Macquarie exit had written their equity down to zero. OMERS, the Ontario Municipal Employees' Retirement System, formally valued its stake at nil in May 2024. The company had drawn on a £3 billion emergency liquidity loan, approved by the High Court in February 2025 against vigorous bondholder negotiation, secured against the company's assets at rates considerably above anything the original whole-business securitisation had contemplated, simply to remain solvent while attempting to agree its future. KKR, the American private equity firm, had been named preferred bidder for a roughly £4 billion equity injection that would have constituted a genuine recapitalisation. On 3 June 2025, KKR withdrew.
The government, through UK Government Investments, had engaged the consultancy FTI to plan for the contingency of special administration - the statutory process, established by the Water Industry Act 1991, by which a court-appointed administrator takes over the running of a water company while its future is determined. Special administration has never been triggered in the history of the privatised water sector. It exists because Parliament understood, even in 1991, that a private operator of an essential monopoly might eventually fail, and that when it did, the water still had to flow. The Water (Special Measures) Act 2025 strengthened the regime and added a rescue option that stops short of full government acquisition. What special administration would actually cost the public is genuinely contested: Dieter Helm, the utility economist at Oxford, has argued that the net fiscal cost could be close to zero, since the government would take on assets as well as liabilities, and that the political presentation of it as a vast public expenditure overstates the real position. The bondholders, who stand to absorb haircuts in any restructuring, have the strongest financial interest in making the alternative look expensive.
At the holding-company level, the consequences of the Kemble structure have been playing out in the courts and in the bond markets. Kemble Water Finance defaulted after the existing shareholders - unwilling to inject the £500 million equity they had committed, having already described Thames Water as "uninvestable" given Ofwat's initial resistance to the dramatic bill increase the company's financial position required - declined to honour that commitment. Ofwat's PR24 final determination, published in December 2024, allowed average bill increases of £157 - approximately thirty-six per cent - over the 2025 to 2030 period. Several companies appealed to the Competition and Markets Authority, which provisionally determined in October 2025 that only around £556 million in additional revenue across all appellants was justified - about a fifth of what they had sought. The gap between what the owners needed to make the model viable and what the regulator was prepared to allow customers to pay became, in the end, unbridgeable.
In July 2025, the Independent Water Commission chaired by Sir Jon Cunliffe published its final report: 450 pages, 88 recommendations, and the most thorough examination of the sector since privatisation. It concluded that the regulatory framework had comprehensively failed and required not adjustment but a fundamental reset. Responding the same day, the Environment Secretary Steve Reed made an announcement whose directness was unusual in a policy area characterised for three decades by studied ambiguity: Ofwat would be abolished. Its functions would be merged with those of the Environment Agency, the Drinking Water Inspectorate, and Natural England into a single consolidated regulator. Reed said, without qualification, that Ofwat had failed to protect customers from the mismanagement of their money and had failed to protect the country's rivers from record pollution. That Ofwat had failed is no longer, in any official account, a matter of serious dispute.
What remains disputed is whether abolishing it and reorganising the nameplate changes the underlying problem. Campaigners from We Own It and the water monitoring group Windrush Against Sewage Pollution pointed out with some force that a new regulator facing the same resourcing constraints, the same information disadvantages, and the same structural relationship to the companies it is supposed to police will face exactly the same pressures as the one it replaces. The problem was never the name on the door.
What This Is Actually About
There is a version of this story in which the moral is simply that privatisation of public services is wrong, and that the obvious answer is public ownership. That version is not without force, but it is incomplete in ways that a serious account of what happened cannot ignore.
The question England's water crisis poses is not "public or private?" It is a prior question: how do you regulate a private monopoly with enough adversarial rigour to protect the public interest when there is no market to do it for you? In almost every sector of the economy, the discipline that keeps private companies honest is competition. If a company charges too much, performs too poorly, or neglects its infrastructure, customers leave, competitors gain share, and the market corrects. In a natural monopoly, this mechanism is structurally unavailable. The customer of Thames Water cannot switch to another provider. They cannot punish poor performance by withdrawing their custom. The only available discipline is regulatory, and the only way a regulator can substitute for a market is if it is designed, resourced, and institutionally committed to being genuinely adversarial toward the companies it oversees.
Britain did not build this kind of regulator. It built Ofwat. And then it relied on Ofwat to constrain owners whose entire investment thesis was premised on the regulator being manageable - owners who were sophisticated, well-resourced, and entirely rational in extracting the maximum possible return from the monopoly cash flows they had acquired. The water companies were not competing with anything. The only discipline available was Ofwat's willingness to impose it, and that willingness was structurally and systematically insufficient.
Three things need to follow from an honest accounting of how this happened.
The first is that any replacement for Ofwat must be designed as an adversarial regulator from the ground up: resourced at a level commensurate with the industry it oversees, empowered to commission independent infrastructure assessments rather than relying on company self-reporting, and governed in a way that ends the revolving-door culture that makes genuine adversarialism institutionally impossible. A merged body with the same budget, the same information constraints, and the same professional culture as the body it replaces will produce the same results.
The second is that the financing model must change. No water company should be permitted to pay dividends until its investment obligations have been independently verified as met. The whole-business securitisation model - which converts monopoly customer cash flows into leverage for private equity, and which allowed Thames Water to operate at gearing levels of nearly eighty-six per cent of its Regulated Asset Base - should be subject to enforceable caps that are meaningfully policed rather than negotiated around. The structure that allowed Macquarie to transfer its acquisition debt onto Thames Water's balance sheet through a Cayman Islands subsidiary should not, in retrospect, have been possible under the terms of a monopoly licence granted in the public interest.
The third is the hardest. Thames Water's trajectory toward special administration is not an aberration from the privatised water model. It is the logical destination of it. The owners loaded the company with debt. The regulator did not stop them. The equity is now priced at zero. The infrastructure is degraded. The sewage is in the rivers. And the government is performing considerable contortions to avoid the word "nationalisation" while preparing, in operational reality, for exactly what nationalisation means. When that moment arrives - and the probability that it does not arrive is, at this point, not high - the government should consider seriously whether an entity that has been asset-stripped by its owners and whose equity has been written to nothing by those same owners is best reconstituted as another private company, handed to another infrastructure fund, and subjected to another three decades of the same model. The market has already provided its answer. The equity is worth nothing. The question of what should replace it is a political one, and it is overdue.
Somewhere under London, a network of Victorian pipes is leaking roughly a quarter of the water that passes through it. Storm overflows are releasing untreated sewage into rivers at rates the monitoring data now makes visible for the first time. And the company responsible for those pipes - the company that sixteen million people have no choice but to rely on - is being kept solvent by a three-billion-pound emergency loan negotiated in the High Court while its shareholders and bondholders argue over who bears the loss.
This is what it looks like when you sell a natural monopoly, call the regulator Ofwat, and assume the rest will sort itself out. It did not sort itself out. The money was followed, and this is where it went.