On 9 September 2026 we held the inaugural Vallorii Infrastructure Forum at Rhodes House, Oxford. 40 senior leader guests from government, regulators, utilities, investors and banks spent the day on one question:

What will it take to make UK infrastructure fit for 2035, when the capital to build it exists but the certainty investors need before they commit it does not?

The day had four panel sessions – macro risk, delivery, maintenance and valuation – as well as keynotes by Minister for Pensions, Torsten Bell MP, and EBRD President Odile Renaud-Basso, and a conversation with Ofgem CEO Tim Jarvis.  Each panel session opened with a short Vallorii case study drawn from our analyses and products; the figures below are ours.

The discussion around infrastructures valuation

The macro environment has turned against long-duration assets. Real five-year government bond yields across the US, euro area and UK went from about minus 1.5% in late 2021 to about 1.9% by mid-2026 (Figure 1). Debt is no longer nearly free, so equity matters again. Higher-for-longer interest rates and inflation coupled with low domestic growth in UK net financial assets, which fell from 5.4% in 2021 to minus 0.7% in 2024, the weakest of the six economies we compared.

Figure 1. Higher-for-longer interest rates and inflation, coupled with low domestic growth, pressure infrastructure financially and politically.

But what impact does today’s macro-environment have on the value of a utility’s Regulatory Asset Base (RAB)? To answer this question, we looked at Heathrow’s third runway. For a notional company, the asset is worth about £49bn once built. With regulatory risk priced in, it is worth under £45bn, and with regulatory certainty over £55bn (Figure 2). The gap of about £10bn is what certainty is worth on a single asset.

Figure 2. Regulatory certainty is worth about £10bn on the third runway.

A key takeaway followed from this: while capital is available, certainty is scarce. Investors said they look for predictable cash flows and a clear path for how prices will be set, and capital that does not find these goes to another jurisdiction. Several participants said a five-year price control is too short a horizon to underwrite against. The broad view overall was that the capital will come, at a higher price.

Delivery risk is priced before anything is built. After having looked into the macro environment, we moved onto the construction one. The NISTA pipeline puts UK infrastructure capex at £460bn for 2025 to 2029, against £303bn for 2020 to 2024, an increase of 52%. A further £258bn is identified for 2030 to 2035, with more in early planning.

Delivery carries a known cost. In the Flyvbjerg database of international nuclear projects, about 85% overran their sanctioned budget, 55% overran by more than half and about a fifth by 200% or more. The average overrun was about 120% (Figure 3).

Figure 3. Capital delivery is costly and risky: nuclear projects average cost overrun doubles the planned costs.

For Sizewell C, our Valorii Price of Risk tool (VAPRI) estimate for the cost of equity to start from a 2.5% risk-free rate (Figure 4). It adds 9.5 percentage points for cost overrun and 3 for the restriction on selling the stake, then takes off 4 for the totex sharing in the Government Support Package. That gives 11% real, against a 10.8% outcome. On our estimate the support package adds about £1.4bn to the value of the equity.

Figure 4. The average nuclear project overruns its budget by about 120%.

Participants were less convinced that finance is the binding constraint. Planning and political will came up more often. Contractors were said to run on thin margins, which leaves little efficiency to gain from construction itself. The concept of repeatability came up in two senses: building the same design many times, and reusing RAB and Public-Private Partnership Models (PPP) that debt and equity already know how to price. Several participants said there would be more appetite if there were an agreed framework for assessing delivery risk.

Maintenance is being financed as though it were growth. After discussing construction risks, we brainstormed on resilience. By looking across water, broadband and electricity, we deconstructed the idea that higher household bills always mean better networks. However important, this misalignment does not show up in the Regulatory Capital Value (RCV), which records what was spent on building, not the condition it is in. For the four UK network groups we examined, the Modern Equivalent Asset Value (MEAV) is between 5.4 and 6.7 times the RCV (Figure 5).

Figure 5. The regulatory capital value is a fraction of what the networks would cost to replace.

Looking into a specific distributor, like UK Power Networks, about 70% of capex is maintenance, and the real regulatory capital value grows by about 1.1% a year. We value the company at £12.6bn, against £9.1bn for a notional company. The £3.5bn difference comes from financing outperformance (£0.8bn), operational outperformance (£1.6bn) and growth in the regulatory capital value (£1.2bn) (Figure 6). We asked the room whether replacing an asset should count the same as building a new one.

Figure 6. About 70% of UK Power Networks capex is maintenance.

Participants agreed that the relevance of resilience goes beyond isolated networks. It is a property of the whole system, which is worth more than the sum of its assets. The question that remained open is what the system should be resilient to. Aviation was offered as a model. Reporting is mandatory there, and one incident teaches every operator of that aircraft type. Energy and water have no equivalent, and asset lives set for financial purposes bear little relation to physical condition. Participants also agreed that relying on the next price review to correct under-investment has not worked.

Valuation has not caught up with the asset. The conclusions drawn by the room naturally led to the final section. Infrastructure returns now range from double digits to write-offs, which is far from the low-risk profile the sector is sold on. Cash-flow valuations can miss the state of the physical asset base. At Thames Water, the compliance obligations currently missed by more than 5% would cost £5.6bn to meet. That reduces a gross regulatory capital value of about £21bn to about £15.3bn (Figure 7). Renewing potable water mains (£2.0bn) and rising mains (£1.2bn) are the largest items. We left the room with the question of who pays.

Figure 7. Compliance obligations reduce Thames Water's regulatory capital value by £5.6bn.

The panel was frank about method. Several participants said valuation has changed little in thirty years: it is still a spreadsheet with cash flows and a discount rate. CAPM leaves out risk specific to the company, and that over the medium term the supply of and demand for capital drives the cost of capital more than any model does. More data gives more precision, but precision is not accuracy. On affordability, the view was that a contract that is out of the money for the customer will not hold, so the equity market needs a project to be affordable as much as the bill payer does. Asked what should change first, the panel's answer was to put physical asset data and performance data into valuations.

Stable institutions matter more than any single decision. In the regulatory conversation, the argument was that investors look for stable institutions and predictable rules, and that the energy framework came through review broadly intact. Participants accepted that regulators duplicate one another and that continued growth in regulation is not credible. AI was put forward as a way to reduce the burden on regulated companies. The same logic was applied to emerging markets. Capital is not the shortage there; what is short is bankable projects of enough scale. The gap between perceived and actual risk keeps investors away, and the appetite for risk exists, but the risk is not perceived as it is.

The conclusion on lack of credibility and long-term commitment as key inhibitor for infrastructure delivery

Senior leaders broadly agreed on three things. Certainty is the binding constraint, not capital. There is money and appetite. What is missing is a credible long-term project pipeline and regulatory commitments that run closer to the life of the asset than a five-year price control. Maintenance is being financed as though it were growth. When replacement work is booked as capital expansion, financial asset lives drift away from physical condition. Resilience has to be assessed across the whole system. Adding up individually sound assets does not show whether the system holds. The first question is what the system should withstand.

The day ended with the conclusion that delivery following the £725bn of infrastructure investment by the Government over the next ten years depends on three design choices: how long a regulatory commitment runs, how maintenance is funded, and how the system is valued. Government, regulators and investors can make all three. At Vallorii, we are building the evidence these choices need on asset condition, delivery risk and the risks owners bear, and we will keep bringing regulators, investors and utilities together to test it.