- The problem: when lots of renewable power hits the grid at once, wholesale prices fall, sometimes below zero.
- The blind spot: models using the average price of electricity can overstate an asset’s revenue. What it really earns is the captured price.
- What Germany shows: rules that stop negative prices can turn a price problem into a volume problem.
- The solution: model the capture rate and the volume you can actually sell, lock in part of your revenue with long-term contracts, and add battery storage or a second technology.
It’s a windy Sunday afternoon in Scotland. Your wind farm is generating at full output, exactly what you built it for. A great day? Not quite. You don’t have a monopoly on the wind. Every competing farm is generating too, and Sunday demand is low. Cheap power floods the market, nobody wants to switch off, and prices fall. Then you check the screen: the price is negative. This wasn’t in the model. How did we get here?
01Why prices fall, and sometimes go negative
Electricity is sold through a merit order: plants are ranked from cheapest to most expensive by their short-run marginal cost (the cost of producing one extra unit), and the most expensive plant needed to meet demand sets the price for everyone. Wind and solar have free fuel and pay no carbon cost, so their marginal cost is near zero. They go first and push out the gas and coal plants that normally set the price.
At low penetration (a small share of supply), this barely moves prices. At high penetration, many farms generate at once, so prices drop sharply in exactly the hours they sell. They can even go negative, because some generators earn subsidies per unit produced and would rather pay to keep running than lose that payment. No one planned this: each owner acted rationally, and together they pushed prices down for everyone. This is price cannibalisation.
02The number your model missed
Most project models use one number: the average (baseload) price, the mean of wholesale prices across every hour. But a wind or solar farm only sells when the wind blows or the sun shines. The price it actually earns is the captured price, the average price weighted by when it generates. The ratio of the two is the capture rate.
Capture rate = captured price ÷ average price
Say your farm is rated at 100 MW with a 35% capacity factor, generating about 306,600 MWh a year. Assume it is built to last 25 years. At an average price of £60/MWh, a static model estimates £18.4M a year. Apply an assumed 80% capture rate and the farm earns £48/MWh, or £14.7M. (These figures are illustrative only.) For a project financed with debt, that £3.7M gap can be the difference between repaying comfortably and struggling.
Is this just a story, or does it happen in real markets?
Negative prices are now common across Europe, and Germany offers a clear view of how it plays out and of how a market can respond. It logged a record 573 negative-price hours in 2025. On Easter Monday 2026, a surge in renewable output met unusually weak demand, and intraday prices fell to about −€324/MWh in the afternoon. Renewables supplied roughly 80% of the country’s load, and Germany shed over 500 GWh of excess generation over the Easter period, roughly 90% of it wind.
03How German lawmakers responded
German lawmakers tried to fix this by changing the subsidy rules. Under the Renewable Energy Sources Act (EEG), many renewable plants receive a market premium: a top-up paid on top of the market price. §51 EEG switches that top-up off when prices are negative, and the rule has tightened over time:
The Solar Peak Act, in force since 25 February 2025, removed the waiting period for new plants. They now lose the premium from the very first negative quarter-hour (a 15-minute trading period), but are still paid in full when the price is exactly zero. So new plants stopped offering power at negative prices and began offering it at €0 instead.
The result? In the first half of 2026, hours with negative prices fell 23%. But commercial curtailment, meaning generation switched off because selling it isn’t worthwhile, rose 20%. Fewer negative prices, but more renewable power going unsold. The problem didn’t disappear. It moved from price to volume.
04The volume problem: pro rata allocation
Much of Europe’s electricity is traded in an auction held the day before it is used. In this day-ahead market, producers tell a power exchange (the marketplace that runs the auction) how much electricity they will sell in each hour of the next day, and the lowest price they will accept. Picture one of those hours. Demand is 80 GWh. Three farms each offer their output at €0, meaning they will sell at any price from zero upwards. Together they offer 100 GWh. (Illustrative numbers.)
That is 20 GWh more than the market needs, and the exchange can’t solve it in the usual ways. The price has nowhere lower to go, because new plants won’t offer below €0 (they would lose their market premium). And it can’t take the cheapest offers first, because every offer is at the same price. So it applies the pro rata rule: the accepted volume is shared out proportionally among all the sellers offering at that price. Everyone is cut by the same percentage, which is worked out like this:
Pro rata factor = volume the market can take ÷ total volume offered
Here, that is 80 GWh ÷ 100 GWh = 80%. Each farm then sells 80% of what it offered:
| Offered | Calculation | Sells | Unsold | |
|---|---|---|---|---|
| Farm A | 50 GWh | 50 × 80% | 40 GWh | 10 GWh |
| Farm B | 30 GWh | 30 × 80% | 24 GWh | 6 GWh |
| Farm C | 20 GWh | 20 × 80% | 16 GWh | 4 GWh |
| Total | 100 GWh | 80 GWh | 20 GWh |
The total sold, 80 GWh, matches demand exactly. And because the market premium is paid only on volume actually sold, the unsold 20% earns nothing.
Now a fourth farm joins with 25 GWh. Offers rise to 125 GWh against the same 80 GWh of demand, so the factor falls to 80 ÷ 125 = 64%, and Farm A sells 32 GWh instead of 40 (50 × 64%), though nothing about Farm A has changed.
It hurts most for subsidised plants, which would otherwise earn their premium in those hours; unsubsidised plants earn nothing at €0 anyway. Older plants without the new rule can still bid below zero, which is why negative prices haven’t disappeared.
Get the next edition in your inbox.
05How investors are adapting
Cannibalisation isn’t a reason to stop building renewables. It’s a reason to build, model and finance them differently. None of the responses below removes the risk, but each makes it easier to see and manage before the money is committed.
- They model the capture rate and the volume they can sell. Instead of assuming today’s average price (the £60/MWh in our example) holds for the farm’s whole operating life, they ask what it will earn in its early, middle and later years as the grid fills with renewables. They also account for the share of output that goes unsold, as in the pro rata example. They check how their subsidy, if they have one, behaves when prices turn negative. Finally, they test several scenarios so lenders can see the downside before committing.
- They lock in part of their revenue with long-term contracts. A power purchase agreement (PPA) is a contract in which a buyer agrees to pay a set price for the power, or at least a minimum price (a price floor, which lets the seller keep any gains above it). Revenue agreed in advance no longer depends on what the market pays in the hours the farm sells. The contract terms decide who carries the risk of unsold volume.
- They add batteries and diversify the mix. Batteries store power when prices are low and sell it when they recover, which also reduces the volume that goes unsold. Pairing wind with solar smooths output, since the two often peak at different times.
These are the kinds of trade-offs teams work through in our simulation.
Make the Mistake Here, Not in the Market
Cannibalisation is easy to understand and hard to see coming. Our Energy Investment Simulation lets teams experience it in a safe environment. Teams build the financial model for a realistic energy project. Midway through the session, they have to react to a market shock: they re-run the numbers and quickly come up with a new plan. They then pitch their case to a committee of seasoned industry investors, who decide whether the project goes ahead at final investment decision (FID). One of our cases, a 100 MW/200 MWh merchant battery in Germany, asks exactly this: revenue stacking versus price cannibalisation.
Delivered at INSEAD · ESSEC · Imperial College London
If you run a school programme, a club or a company team and want your people to learn about energy investment in a hands-on way, email us and let’s discuss how to make that happen.
Email us about the simulationNew editions on energy investment, delivered when they publish. By subscribing you agree to our privacy policy. Unsubscribe at any time.
- Montel: From negative prices to zero hours
- Euronews (27 July 2026): Why Germany keeps switching off its solar and wind farms
- Bloomberg (7 April 2026): Germany Power Prices Turn Deeply Negative on Renewables Surge (may require a subscription)
- IENE: Comment No. 66 (PDF)
- Kpler: What can power markets learn from the Easter break
- Copernica Partners: Negative electricity prices 2026
- Next Kraftwerke: 6-hour and 4-hour rules (in German)
- Insurance Journal (5 January 2026)