Solar projects have traditionally relied on long-term, fixed-price contracts that give lenders predictable revenues. That is starting to change. More developers are now selling power directly into wholesale markets through corporate PPAs, green certificates and spot trading. This “merchant” model offers flexibility, but it also exposes projects to swings in both price and output that can leave them short of cash when debt payments fall due.
A recent case study from Oxford’s Saïd Business School looks at what this means in practice. It models a 150MW solar plant in Bavaria selling into the German wholesale market and asks whether it can service its debt, and what it would cost to hedge the risk.
Why Germany is a useful test case
Germany has one of Europe’s most liquid power markets. Renewables supplied about 63% of net generation in 2024, with solar alone at roughly 14%. That liquidity makes trading easier, but it also creates a structural problem for solar owners.
When the sun is shining, everyone’s panels are producing at once. Between April and September, midday prices regularly fall below €50/MWh and sometimes go negative. In winter, and during “Dunkelflaute” periods when it is dark and windless, prices can spike to 10 to 20 times normal levels, exactly when solar output is at its lowest. As more solar capacity is added, this pattern intensifies. About 14GW of new solar in 2023 helped push average peak prices down from €106/MWh to €88/MWh in 2024.
The case study project
The modelled plant costs about €90 million (roughly €600,000 per MW) and generates around 188GWh a year. Because merchant revenues are uncertain, it is financed more conservatively than a fixed-tariff project would be: 50% debt and 50% equity instead of the typical 75% debt. At a 4% cost of debt and a 7.45% cost of equity, its weighted average cost of capital (WACC) is about 5.07%.
Based on 2024 prices, the project earns about €10.4 million in revenue and remains profitable overall. The monthly picture is less comfortable. January was a loss-making month. At 75% gearing, every low-generation winter month would have run negative cash flows. Seasonality, not just annual totals, determines whether debt can be serviced.
Five years, five very different outcomes
Running the same plant against actual prices from 2020 to 2024 shows how much results depend on market conditions:
2020 (COVID-19): average price about €30/MWh. Net loss of almost €1 million.
2021: about €97/MWh. Net profit of about €5.8 million.
2022 (energy crisis): about €235/MWh. Net profit above €24 million.
2023: about €95/MWh. Net profit of about €5.5 million.
2024: about €80/MWh. Net profit of about €2.8 million.
Over a 30-year life, three forward scenarios were tested:
2023 prices falling 6% a year: 9.8% IRR, 8-year discounted payback.
2024 prices falling 3% a year: 6.9% IRR, 12-year payback.
2020-level prices rising only with inflation: 3.6% IRR, 21-year payback.
The third scenario is the warning sign. It represents a future where cheap renewables and storage dominate, and in it the project fails to clear its cost of capital.
Hedging: protection at a price
The study proposes a two-sided hedging strategy covering roughly the first half of the debt term:
Sell power forward for the high-output months (March to September). This locks in prices and guards against negative-price periods.
Buy power forward for the low-output months (October to February). This protects against having to cover shortfalls at spike prices.
Hedges are layered across the year-ahead, day-ahead and intraday markets, with most volume in the more liquid year-ahead contracts.
Together, the hedges cover around 60% of annual revenue. Total hedging costs come to about €0.8 million over roughly eight and a half years. Insurance products, such as volume-deficit cover for forecasting errors, can complement the hedges by covering imbalance risk.
The author’s caution is worth repeating: don’t over-hedge. These instruments are expensive and can erode the returns they are meant to protect.
Key takeaways for investors and lenders
Gear down for merchant exposure. Debt at 30–50% of capex, plus a debt service reserve account, gives a buffer through low-price periods.
Model monthly, not annually. Winter cash shortfalls can breach covenants even in a profitable year.
Stress-test a low-price future. The “cannibalisation” of solar prices is already visible, and a 2020-style price world makes projects marginal.
Hedge selectively. Target the debt-repayment period and the months that matter most, and weigh the cost against the impact on IRR.
Merchant solar can be bankable, but only with financial structures designed around the market’s seasonal and structural volatility.