Water DPC projects may require 9.2-9.6% Cost of Equity

In 2015, the Thames Tideway Tunnel cleared at roughly 4% real Cost of Equity (CoE); this month, fresh capital for Sizewell C was priced at 10.8%. The doubling of required returns in UK infrastructure is a market reality, not a regulatory choice—and it places Ofwat in a delicate position. Tasked with maintaining consumer affordability even as the water sector needs roughly £60 billion of new capital in the next five years^[i], the regulator must translate policy ambition into an allowed return that can attract capital.

That challenge crystallises in the Beckton Water Recycling Plant, one of the first big tests of Ofwat’s Direct Procurement for Customers (DPC) scheme and a bellwether for the £40 billion pipeline^[ii] that follows. A DPC is a competitive auction in which a water company tenders a large, discrete project to a third-party provider, with the aim of securing lower lifetime costs for customers than building in-house. Vallorii’s analysis puts Beckton’s real CoE at 9.2–9.6%, about four points above the 5.1% allowed in PR24. That higher required return translates into larger annual payments demanded by the provider—costs that ultimately flow into household bills.

The premium is rooted in evidence. Vallorii’s asset-pricing engine generated about a million scenarios, then tied each to hard data. The model found 3 key insights.

1. A study of 1,093 UK water-infrastructure construction projects shows fat tailed cost overrun risk^[iii]; feeding that distribution into Beckton’s capital expenditure budget adds 240–290 basis points to the Cost of Equity.
2. Counterparty risk contributes another 150–190 bps: Thames Water’s CCC junk-grade credit rating implies substantial risk of payment default. DPC projects rely solely on the Appointee water company to deliver annual payments, and the risk of a Thames Water bankruptcy heightens cash flow uncertainty.
3. By contrast, asset-health is not a significant risk. Performance penalties are capped at about a tenth of annual revenue^[iv] and the project’s terminal value will be heavily depreciated, so even extreme asset-condition shocks barely register against construction or counterparty volatility. As a result, asset-health contributes near-zero incremental CoE for a new-build plant (though it can dominate brownfield appraisals once construction risk subsides).

Vallorii’s modelling shows that allocating these risks explicitly can soften investors’ demands. One option is a fixed-price construction contract: the construction contractor, not the investor, eats any cost overruns, trimming the required return by more than a percentage point. The snag is that most contractors run on razor thin margins; a major overspend can push them into insolvency, throwing the bill back onto investors. A sturdier—though politically sensitive—remedy is for consumers or government to underwrite the worst outcomes, from extreme overruns to a Thames Water default. Vallorii’s modelling can stack these safeguards one by one, quantifying how each cushion nudges Beckton’s CoE closer to levels acceptable to both investors and bill-payers.

Going a step further—ring-fencing the scheme as a Thames Tideway-style standalone provider (a ‘SIPR’)—could push the CoE lower still, a route we will unpack in an upcoming post. In an era of dear capital, such calibrated cushions offer Ofwat the best shot at pairing policy ambition with affordable bills.

In an era of dear capital, such calibrated cushions offer Ofwat the best shot at pairing policy ambition with affordable bills.