National Grid - Gilts Plus a Margin

National Grid has sold the Isle of Grain, the UK's largest liquefied-natural-gas import terminal — capable on its own of meeting close to a third of national gas demand — to a joint venture of Centrica and the American infrastructure fund Energy Capital Partners, for an enterprise value of around £1.5 billion, most of it the buyer's own borrowing. It cleared the National Security and Investment Act and the competition authorities, and completed without much fuss.

The coverage asked the obvious questions: should a strategic energy asset pass into private-equity hands, was national security weighed? The answers were reassuring — a British lead buyer, supply obligations that bind any owner, the Act applied and the sale cleared. All true, and all beside the point. The sale's real significance is the reverse of the headlines: National Grid is not retreating from critical infrastructure but concentrating into the most lucrative form of it. It sold the competitive piece and kept the monopoly. To see why that is rational — and what it costs everyone who pays a bill — you have to look at the one number that never appears in a press release: the regulated cost of capital.

The puzzle nobody restates

Privatising a competitive business has a clean logic: competition disciplines price and cost, private ownership harnesses that discipline, the public benefits. Privatising a natural monopoly has no such logic, because there is no competition to do the disciplining. One firm serves the whole market because duplicating the network would be wasteful — which is exactly why no rival exists to hold it honest. Hand it to shareholders and you keep monopoly pricing power and add a profit motive, with nothing between them.

The usual answer is regulation as a substitute for competition. Britain's first such regime, RPI-X, capped a utility's prices at inflation minus an efficiency factor the company had to find each year; its successor, the framework the regulator Ofgem uses today, is called RIIO — Revenue equals Incentives, Innovation and Outputs. Both work the same way: cap allowed revenue, let the firm keep what it saves by beating the cap, reset the cap every few years to pass savings to consumers. The other, less stated, reason is the balance sheet: a privatised monopoly's investment doesn't count as government borrowing, so moving tens of billions off the public books is the real attraction, dressed as efficiency. Both rest on a single number.

Gilts plus a margin

The regulator sets an allowed return — a cost of capital the company earns on its regulated asset base, the accumulated value of its pipes, wires and substations. That return is not paid from the company's pocket. It is collected from consumers, a line in every bill, and a dominant one, because a network is overwhelmingly capital rather than labour. Part is interest on the company's debt; part is pure profit for shareholders. The bill-payer funds both.

From April 2026, under the current transmission price control, the allowed return for electricity transmission is a real cost of capital averaging about 5.6 per cent, with the shareholders' slice — the allowed return on equity — quoted by National Grid at 6.12 per cent. The tell is the comparison with the previous control, where it was 2.81 per cent. It has roughly doubled, though nothing about the business doubled in risk; what changed is that real interest rates rose, and the allowance tracked them. The regulated return is, in substance, the risk-free rate plus a broadly constant margin — a tracker on the government bond curve with a spread bolted on. Which raises the question the whole settlement turns on: if the return is gilts plus a margin, what is the margin buying?

Is the margin risk, or rent?

The defence is that it compensates equity for real risk — construction overruns, the buffer that absorbs failures before they reach taxpayers. That is testable: if equity bears delivery risk, realised returns should fall below the allowance in bad years. They don't.

Take the price control that ran from 2013 to 2021. The allowed equity return was around 7 per cent real; the realised return for the transmission networks ran between nine and eleven. The National Audit Office found the networks delivered shareholders about 9 per cent against a wider corporate average of five to six, and concluded that targets had been set too low, budgets too high, the cost of equity overestimated — and the lax settlement left to run for eight years rather than five. That is as close to an official finding of structural over-reward as the record offers. For the 2010s, the margin was rent.

To its credit, Ofgem responded, cutting the baseline equity return by roughly a third for the next control. And here the story turns subtler than profiteering. The headline realised return stayed near 9 per cent, but its source changed entirely. National Grid's transmission arm beat its allowance of 4.8 per cent by a little over a point on operations — then leapt to 9.1 on a single line: financing. That was not delivery or efficiency but a treasury bet — the company had chosen not to index much of its debt, and high inflation eroded the real cost of its fixed-rate borrowing. Roughly four-fifths of the apparent outperformance was an inflation windfall, and it has faded as inflation cooled. The same pattern appears, near-identically, at the other transmission owners, which tells you it is a property of the regime, not anyone's brilliance. The effect is symmetric — it would have been a loss had inflation undershot — and the regulator is removing it; Scottish Hydro even publishes the figure stripped of inflation, and its return falls from 9.1 to about 5.5, barely above the allowance.

That clears the noise, and what it leaves is the point. Across every owner the cost line came in positive in every year — none bore a delivery loss at the equity level. Scottish Hydro built the most, in the hardest terrain, energising the Shetland subsea link ahead of schedule, and still lost nothing. The delivery risk the margin supposedly prices did not materialise.

The honest number

So set aside the over-reward the regulator corrected, the inflation windfall it is removing, and the modest operational outperformance. What survives?

To find it you change the metric — and the choice of metric is itself revealing. Return on equity, the figure the companies headline, measures only the slice shareholders finance. But consumers fund the whole base, debt and equity, through their bills, and the return on the whole base is the return on the regulated asset value. Only Scottish Hydro reports it: an allowed 3.3 per cent real, a realised 3.8. National Grid and SP Transmission disclose the shareholder number and withhold the one that bears on the bill.

Set that 3.3 per cent against the government's own real borrowing cost — long index-linked gilts yielding perhaps 1.5 to 2 per cent. The public funds the entire asset base at around 3.3 per cent real when the state could fund it at half that. The gap — call it a point and a half — sits on the whole base, owes nothing to inflation or skill or any controversy the regulator has tidied up, and has survived every reform. On a transmission asset base now well north of £30 billion and heading toward £60, that point and a half is several hundred million pounds a year, rising toward a billion.

Be plain about what that money is, because the language is built to obscure it. It is not payment for wires or maintenance. It is the extra the public pays purely because the network is financed by private shareholders rather than the state — the difference between the return investors demand and the rate the government can borrow at. Finance the identical assets with government debt and that line shrinks toward the gilt rate: the wires, the work and the engineers stay the same; the only thing that disappears is the shareholder.

The standard reply is that the premium buys risk transfer — the public pays more but sleeps safe, because if the network fails it is shareholders who are wiped out. But the premise is false. A national grid cannot be allowed to fail; if an owner collapsed the state would step in, as it has had to circle Thames Water. The catastrophic risk already sits with the public. Consumers pay an equity premium to insure a risk they still carry, from a counterparty who, in the only scenario that matters, could not pay the claim. The risk was never transferred. Only the profit was.

The merchant mirror

If the networks show private ownership extracting a premium from an asset the public cannot leave, the Isle of Grain shows the same extraction in reverse — selling off a profit rather than overcharging on a monopoly. The terminal is genuinely cash-generative: £176 million of earnings before interest, tax and depreciation in the year to March 2025, on long-term take-or-pay contracts.

How the buyers expect to profit is the whole thesis in miniature. They project an unlevered return of about 9 per cent — and an equity return of 14 or more. The gap is manufactured entirely by debt: roughly £1.1 billion loaded onto the terminal, so the equity holders pocket the spread between what the asset yields and what their borrowing costs. The state could have taken the unlevered nine per cent against a near-gilt cost of capital and kept the surplus for consumers or the Treasury. It is the same argument in different clothes: the public surrenders an ongoing yield for a lump sum, and private capital banks the difference.

A fair critic notes that Grain, unlike the networks, is a merchant business whose higher return reflects real risk — gas in decline, market swings — and that the state may have sold near the top, passing on the transition risk. Not a foolish trade. But it understates what changed hands: one of only three LNG terminals in the country, a third of gas demand, a matter of supply security as much as commerce. The public did not shed a liability; it converted a strategically vital, cash-generating asset into a private profit stream, took a one-off cheque, and surrendered the income and control of part of the nation's gas lifeline.

The same toll at every gate

And the wedge is not charged once. Electricity reaches a house through two monopolies, not one — the national grid, then a regional distribution network, the local wires owned by a company with no competitor in its patch. Distribution earns the same kind of allowed return: a real cost of equity around 5.2 per cent, a cost of capital near 3.9, again well above what the state could borrow at, again at every regional network. Gas arrives the same way. Step outside energy and it repeats: the water companies are regional monopolies of identical construction, allowed around 4 per cent real for 2025–30 — several appealing for more — as household water bills rise by over a third on a capital base approaching a hundred billion pounds.

So the wedge stacks: transmission, distribution, gas, water, each a separate monopoly charging a return above the cost at which the same assets could have been publicly financed, the premiums compounding down the chain. This is the real substance of the sense that we are made to enrich too many parties for things we cannot opt out of. A wholly state-financed system would carry none of it.

Two honest qualifications. The retail suppliers — the companies that send the bill — are the one layer where profiteering does not hold: their margins are thin and capped, and many collapsed in 2021 and 2022. The charge that lands against them is not profit but waste — marketing, churn, and failed suppliers mutualised back onto everyone's bills. And the stacked cost of capital is not the main reason British power is dear: the UK has the highest industrial electricity prices in the developed world and among the highest for households, but its gas is cheaper than most of Europe's, and the chief causes lie in a market that prices all electricity off the costliest gas unit running, and in the levies loaded onto power rather than gas. The financing wedge is a real, avoidable, compounding cost — one among several, not the whole. Reason to name it precisely, not to inflate it.

The system working as designed

Now the Isle of Grain makes sense. National Grid sold a volatile merchant terminal and is pouring the proceeds, and a capital programme in the tens of billions, into the regulated networks, where every pound invested enlarges the base on which it earns its gilts-plus-a-margin return, underwritten by the regulator and paid by people who cannot shop elsewhere. Not a retreat from critical infrastructure — a concentration into its most dependable form. The model working as intended.

The arguments we usually have have largely been settled. The regulator tightened. The windfall is going. The operators deliver roughly to allowance. Conceding all of it costs nothing, because none of it reaches the point. There is one real counter: that private operators, under threat of underperformance, run and build more cheaply than an insulated state body — and the public sector has its own over-budget projects to answer for. But grant them every point on efficiency, which the record puts at around a single percentage point, and it still cannot recover the financing premium. Cost of capital is a separate and larger question than cost of operation, and on it the answer is not seriously in doubt: the state borrows more cheaply than any company, and on a near-guaranteed asset base whose risk the public underwrites anyway, the equity premium buys nothing it does not already own. A publicly financed grid would mean a lower cost of capital and lower bills, for the plain reason that there would be no shareholders to pay. That is not ideology. It is arithmetic.

So the conclusion is harder than the comfortable middle allows. We chose, and go on choosing, to route the financing of essential infrastructure through investors demanding a return we could have avoided, and to hand them, through every bill, a margin over the government's own borrowing cost, indefinitely, on assets too critical ever to be allowed to fail. The premium is not the price of protection. It is the price of ownership we gave away.

We privatised the monopolies, told ourselves a regulator would stand in for the competition that by definition could not exist, and stopped looking. The Isle of Grain changed hands for £1.5 billion this year, and the only question anyone asked was who was buying. The better one is what the rest of us still pay, year after year, to rent back the things we used to own.

Sources: National Grid plc disposal announcements, FY2026 results and RIIO-T2 Regulatory Financial Performance Reports; Centrica/ECP acquisition release (August 2025); Ofgem RIIO-T3 and RIIO-ED2 Final Determinations and RIIO-2 performance data; Ofwat PR24 Final Determinations; National Audit Office, "Regulating energy networks"; Scottish Hydro Electric Transmission and SP Transmission RFPRs; DESNZ/IEA international electricity price data; UK Debt Management Office gilt yields. Figures real-terms unless stated; the water allowed return and the gilt real yield to be confirmed at publication.