Opinion — While Tony Blair is trashing UK govt energy policy, NESTA analysis finds "early signs [it] may be wo

The UK Power Desk's Oldest Assumption Just Broke, and Almost Nobody Repriced For a decade, trading British power has rested on one reliable shortcut: when in doubt, it is a gas derivative. Model the gas curve, add a spark spread, and you have a serviceable view on electricity. That shortcut is now wrong, and the evidence is a single ratio that should be on every UK power desk's front page. In 2022, gas roughly doubled and electricity rose about 83%, near one-for-one. In 2026, gas rose about 28% and electricity just 6%. The pass-through has collapsed from roughly 1:1 to roughly 5:1, and the political class is busy arguing about whether the policy that produced it is good or bad while the market implication goes unhedged. Set the politics aside, because the politics is the least interesting part. Tony Blair attacking government energy policy and NESTA defending it are both arguing about whether bills are acceptable. A trader does not care whether the policy is popular. A trader cares that the correlation between gas and power, the single relationship most UK energy books are built on, has weakened by roughly four-fifths in four years. That is not a talking point. That is a model risk sitting live in portfolios that still treat power as gas plus a margin. Start with the spark spread, because it is the cleanest casualty. The spark spread is only a clean gas-margin trade if gas is reliably the marginal price-setter. When gas sets the power price in almost every hour, as in 2022, the spread behaves and the gas plant is the swing unit. When gas sets the price in a shrinking share of hours, as the 2026 ratio implies, the spark spread stops measuring what traders think it measures. It becomes a spread between an input that is on the margin sometimes and an output that is increasingly set by something else. A desk running spark-spread positions sized on the old pass-through is carrying a basis risk it has not named. The same logic runs through contracts-for-difference. A CfD is, in effect, a bet on the gap between a strike and the reference power price, and the value of that gap depends on how power actually moves. If electricity decouples from gas, the distribution of power prices changes shape, and the value of every CfD struck under the old correlation regime drifts. Renewables generators sitting on CfDs and the offtakers on the other side are both holding instruments whose risk profile has shifted underneath them, whether or not they have remarked it. Here is the part that should bother anyone running a book. The decoupling is not a gradual drift you can adjust to at leisure. The mechanism, renewables pushing gas down the merit order and electrified demand increasingly coinciding with renewable supply, is non-linear. Once renewable output and flexible demand start lining up in the same hours, the frequency with which gas is on the margin falls faster than installed capacity alone would suggest. So the correlation does not decay smoothly; it can step down. A model calibrated on last year's relationship can be materially wrong this year, and the 1:1-to-5:1 move is exactly the kind of step-change that breaks backward-looking calibration. Now the discipline, because I am not arguing the old relationship is dead. I am arguing it is conditional, and the condition is system stress. A 5:1 pass-through measured in a year when gas rose a moderate 28% may not hold in a cold, low-wind winter, the dunkelflaute weeks when renewables vanish and gas has to set the price hour after hour. In that environment the pass-through could snap back toward 1:1, and a desk that has fully repriced to the new regime would be caught the other way. The correct read is not "power has decoupled from gas." It is "the gas-power correlation has become state-dependent, high in stress, low in calm," and that is a harder, more valuable thing to trade than a single ratio. That state-dependence is itself the opportunity. If the pass-through is low in benign conditions and spikes in stress, then the correlation between gas and power has become a tradeable variable in its own right, not a constant to assume. Optionality on that relationship, structures that pay when the gas-power link reasserts in a tight winter, is worth more than it was when the correlation was a stable 1:1. The desks that recognise the regime change first will price that optionality before the ones still treating power as a gas clone. The market failure here is one of attention. The decoupling arrived dressed as a political story, a fight between a former prime minister and a think tank over whether bills are tolerable, and so it got filed under policy rather than under risk. But a four-fifths collapse in the central correlation of a market is a risk-management event, not a policy footnote. The number to act on is not the level of the bill that Blair is attacking. It is the pass-through ratio, which has told you that the instrument you are trading no longer behaves the way your model says. In 2022 almost all of a gas move reached power. In 2026 a fifth did. If your book still assumes the former, the market has already moved on without you.