Electricity Markets: How the Price Is Actually Made
Electricity is unusual among commodities: it cannot be stored at scale, supply must match demand within seconds, and the physical network makes it impossible to direct a particular electron to a particular buyer. The market design that emerged from those constraints is elegant, widely misunderstood, and currently under more political pressure than at any time since liberalisation.
The Merit Order
Every hour, generators tell the market operator what they would charge to produce. The operator stacks those offers from cheapest to most expensive and works up the stack until the total matches expected demand. The plant whose offer is needed last sets the clearing price, and every plant that was dispatched receives it - including the ones that offered far less.
This looks unfair at first glance and is deliberate. Paying each generator its own offer would make bidding a guessing game about what others will do, and generators would inflate offers to capture margin. Paying everyone the marginal price makes the honest strategy the optimal one: offer your actual marginal cost, and you will be dispatched whenever it is efficient for you to run. The design produces the cheapest possible dispatch of the existing fleet, which is what it was built to do.
What a plant offers is its marginal cost - the cost of producing one more megawatt-hour, not its total cost. For a wind farm or a solar array that is essentially zero, because the fuel is free and the capital was spent years ago. For a gas plant it is the fuel bill plus the carbon cost. For a nuclear plant it is low but the plant prefers not to cycle, so it often bids very low to stay running.
The consequence is an ordering that barely changes: renewables and nuclear at the bottom, coal and gas above, and the most expensive peaking plant at the top. Demand determines how far up the stack the market climbs on a given hour, and therefore what everyone is paid.
Why Cheap Wind Does Not Mean Cheap Bills
This is the single most common misunderstanding in energy policy, and it follows directly from the mechanism. Adding wind and solar pushes expensive plants out of the stack, so the marginal plant on a windy afternoon may be cheaper than it would have been. That is the merit order effect, and it genuinely reduces average wholesale prices.
But on a still winter evening the wind contributes little, demand is high, and the marginal plant is gas. Gas then sets the price for everyone, including all the nuclear and renewables that were dispatched below it. Building more wind does not change what happens in that hour unless the wind is actually blowing.
In 2022 this became visible to everybody. Gas prices rose by a factor of ten, gas set the price in most hours across Europe, and electricity prices followed gas even in countries where most electricity came from other sources. Nothing was broken; the market did exactly what it was designed to do, which is pass the marginal cost through. Whether that is the right design when the marginal fuel is a geopolitical instrument is a separate question, and several countries have been asking it since.
The reform proposals divide into two families. One decouples retail prices from wholesale marginal pricing through long-term contracts, so that a consumer pays a blend of what the fleet actually cost rather than what the last plant charged. The other keeps marginal pricing and taxes away the inframarginal windfall. Both have been tried in part since 2022, and neither changes the physics of dispatch.
Negative Prices and What They Reveal
Prices below zero mean a generator is paying to keep producing. They sound absurd and they are informative: they happen when there is more generation that cannot easily stop than there is demand to absorb it.
Several things cannot easily stop. A nuclear plant loses money and stresses equipment if it cycles. A combined heat and power plant supplying a district heating network must keep producing heat, and the electricity comes with it. A wind farm under certain subsidy schemes earns per megawatt-hour generated and will pay up to that amount rather than curtail. Add a sunny, windy Sunday in spring with low industrial demand, and the stack overflows.
Negative prices have gone from a curiosity to a regular feature. Germany, the Netherlands, Spain and California all now record hundreds of negative-price hours per year, and the count rises with solar penetration. They are a signal that the system has more energy than it can use at that moment, which is precisely the signal storage, flexible demand and transmission exist to answer.
They also point at the value deflation problem. Because solar output peaks simultaneously everywhere in a region, each additional solar farm arrives at a moment when the price is already depressed by all the others. The marginal value of solar falls as its share rises, which is invisible in a levelized cost comparison and central to what a grid will actually pay for.
Paying for Capacity Instead of Energy
An energy-only market pays for megawatt-hours. That works while most plants run most of the time. It breaks when a plant is needed for a hundred hours a year to prevent a blackout, because the revenue from a hundred hours does not cover the cost of existing for eight thousand seven hundred and sixty.
The market answer is scarcity pricing: when supply is tight, prices spike to thousands of euros per megawatt-hour and a peaking plant earns its year in a few days. This works in theory and is politically intolerable in practice, so most markets cap prices - and the cap removes exactly the revenue the peaking plant needed. That is the missing money problem, and it has been debated since liberalisation began.
Capacity markets are the usual response. Generators are paid a retainer for being available at a defined time, whether or not they run, and the system operator buys enough capacity to meet a reliability standard. Interruptible demand, storage and demand response compete in the same auctions, which is how a shopping centre agreeing to shed load becomes a substitute for a gas turbine.
The argument against is that capacity payments subsidise plants that would otherwise exit, including fossil ones, and that a properly functioning scarcity market would not need them. The argument for is that no country has been willing to let a scarcity market run to its conclusion during an actual shortage. That disagreement is unresolved and is one of the live questions of electricity market design.
Frequently asked questions
What is the merit order?
The ordering of generators from cheapest to most expensive marginal cost. The market operator dispatches upward through that stack until supply meets demand, and the offer of the last plant needed sets the price paid to every plant dispatched. Renewables sit at the bottom because their fuel is free.
Why do all generators get the same price?
Because paying each its own offer would turn bidding into a guessing game and encourage inflated offers. A single clearing price makes offering your true marginal cost the optimal strategy, which produces the cheapest possible dispatch of the existing fleet. The design is intentional, not an oversight.
If wind is nearly free, why are bills high?
Because the price is set by the last plant needed, not the average one. On a still winter evening that plant is usually gas, and gas then sets the price for everyone including the renewables dispatched below it. More wind lowers average prices but does not change the hours when the wind is not blowing.
How can electricity prices be negative?
When more generation that cannot easily stop is running than there is demand. Nuclear plants, district heating plants and some subsidised wind farms keep producing rather than shut down, so on a sunny, windy Sunday supply exceeds demand and prices go below zero. They are now a regular feature in high-renewable systems.
What is a capacity market for?
To pay plants for being available rather than for generating. A plant needed a hundred hours a year cannot cover its costs from a hundred hours of revenue, and price caps remove the scarcity pricing that would otherwise pay for it. Capacity auctions buy availability, and storage and demand response compete in them alongside generators.
