Solar market value: why solar power is worth less and less on the exchange

On the power exchange, solar electricity is a commodity with a flaw: it always arrives when every other solar plant is delivering too. The more photovoltaics is connected to the grid, the deeper the price falls in exactly the hours in which a solar park produces. The figure that measures this is the solar market value (Marktwert Solar). The transmission system operators calculate it every month, and it is also the second variable in the market premium through which a solar park in a direct investment earns its money.

This article lays out the official series since 2015, uses hourly data from the Federal Network Agency to trace where the value is lost, explains a rule from Annex 1 of the EEG that most explainers leave out, and shows how much of the cannibalisation actually reaches a direct investment with an EEG award. It describes the market, not a specific project.

What is the solar market value, and who calculates it?

The solar market value is the average revenue of a kilowatt-hour of solar power on the spot market, weighted by the quantity actually generated in each quarter-hour. The legal basis is Annex 1 to § 23a EEG. Under number 3.3.4, for every quarter-hour of a calendar month the spot market price is multiplied by the solar power that the transmission system operators report in their online extrapolation for that quarter-hour; the sum is divided by the monthly quantity. The result is the monthly solar market value, MW Solar, in cents per kilowatt-hour. Number 4.3.4 describes the same calculation over a whole calendar year; the result is the annual solar market value, JW Solar.

The four transmission system operators must calculate and publish both values on their joint site netztransparenz.de, the monthly values at the start of the following month, the annual values under number 5.3 by the tenth working day of the following year, to three decimal places. Alongside them stand the market values for onshore and offshore wind and the unweighted monthly average of the spot market price. That comparison is what makes the figure readable: the ratio of solar market value to spot average is called the market value factor, or capture rate. At 1, solar power is worth as much as an average kilowatt-hour; at 0.5 it brings only half.

For a solar park with EEG support the market value is not a statistic but an input: the sliding market premium is the difference between the plant's award value and the solar market value (Annex 1 No. 3.1.2 and 4.1.2 EEG). If the market value falls, the premium rises, and vice versa. How this formula works day to day with the direct marketer and where the award value comes from is explained in Direct marketing and the market premium: how a solar park earns its money; this article is about the variable itself.

How has the solar market value developed since 2015?

From a premium to a discount. Up to 2018 the monthly solar market value stood above the spot average in six of twelve months; solar power arrived at midday in the expensive hours and was therefore worth rather more than the average. In 2019 it was still five months, in 2020 four, in 2021 three, and since 2022 the solar market value only manages this in one or two winter months. The annual market value, which the transmission system operators have reported since 2019, shows the same path in a single number.

Two things stand out. First, the 2025 annual solar market value of 4.508 cents is barely below the 2024 value of 4.624 cents, although the price level rose by 14 percent; the factor therefore fell from 0.59 to 0.50. Cannibalisation does not necessarily show up as a falling cent amount, but as a widening gap to the rest of the market. Second, 2022 was an outlier: in the gas crisis prices were high in all hours, the factor rose to 0.95, and solar parks without a price cap earned more than ever. Anyone deriving the market value of a 20-year investment from 2022 is working with a year that need not repeat itself.

Within a year the range is even wider than between years. In 2025 the monthly solar market value ran from 11.511 cents in January to 1.843 cents in June; in 2026 April followed with 1.317 cents, the lowest monthly value the series has seen since 2015, although the spot average in the same month was 7.852 cents. Summer is therefore not only the time of highest generation but also of the lowest prices, and the two together make the annual value.

What is cannibalisation, and in which hours does it happen?

Cannibalisation means that solar plants spoil each other's price because they all deliver at the same time. The exchange price of an hour is formed from the order of power plants by their marginal cost; renewable plants with marginal costs close to zero stand at the front. If a lot of photovoltaics feeds in at midday, expensive plants drop out of the market, the price falls, and if supply exceeds demand it falls below zero. For the individual plant this is not a fault, for the fleet it is a system effect: every additional gigawatt lowers the price in exactly the hours in which all the other gigawatts are producing too. At the end of 2025, according to the Federal Network Agency, 117 gigawatts of photovoltaics were installed, 16.4 gigawatts of them new in 2025, and solar generation rose to 74.1 terawatt-hours.

Where the value is lost is shown by an analysis of the SMARD hourly series that we sorted by hour of the day for three years: how high was the price on average in that hour, what share of the year's solar generation fell into it, and what did the solar kilowatt-hour actually earn there?

Three findings. First, the generation profile does not change: in all three years, 60 percent of solar generation falls between 10 a.m. and 3 p.m. What changes is the price in that window. In 2019 the midday average was 37 euros and thus only just below the daily average; in 2025 it was 55 euros, while the morning brought 104 and the evening 101 euros. Second, the midday kilowatt-hour is hit even harder than the midday average shows, because generation within the window is largest in the cheapest quarter-hours: in 2025 it earned an average of 23.5 euros per megawatt-hour between 1 and 2 p.m. at an hourly average of 46 euros, and in 99 of those hours the price was negative. Third, the shoulder hours become more valuable but are of little use to a solar park: between 7 and 8 p.m. the price in 2025 was 136 euros, but only one percent of annual generation arises there.

The share of solar generation that falls into hours with a negative price is the second axis of cannibalisation. By our analysis it was around 3 percent in 2019, 8 percent in 2023, 18 percent in 2024 and 24 percent in 2025; in the current year up to August again 23 percent. In those hours the kilowatt-hour does not earn little, it costs money, and since § 51 EEG as amended by the Solar Peaks Act there is no market premium there for new plants either. What that means in detail and why a direct marketer then curtails the plant is explained in Negative electricity prices: what they mean for solar and storage investors.

Monthly or annual market value: which applies to a solar park from 2023?

The annual market value. Annex 1 No. 2 EEG 2023 divides plants into two groups: those commissioned before 1 January 2023 or awarded before that date continue to have their market premium calculated monthly, as award value minus monthly market value (number 3.1.2). For all other plants number 4 applies: the market premium is calculated for the calendar year, as award value minus annual market value (number 4.1.2), retrospectively after the year end. During the year the grid operator pays monthly instalments under § 26 (1) EEG, which it may base on the previous year's annual market value; the difference is settled with the final statement in the following year.

The difference sounds technical but is money. Under the monthly calculation the floor of zero applies to each month separately: in a January with a market value of 11.5 cents the premium becomes zero and the plant keeps the high winter price; in a June with 1.8 cents the premium tops up to the award value. The good months stay with the operator, the bad ones are compensated. Under the annual calculation this advantage disappears: winter and summer are netted in the annual market value before the floor applies. For a new park the spread between January and June is therefore irrelevant for the premium; all that matters is where the annual average lands.

In 2023 the rule had no effect, because the market value exceeded the award value in every month and the premium was zero anyway. In 2024 and 2025 the annual calculation cost the new plant 0.3 to 0.5 cents per kilowatt-hour compared with an older plant with the same profile and the same award value, because the high winter months raise the annual market value and thus lower the premium, without the summer months being topped up individually. Anyone using the monthly formula in a revenue forecast for a park awarded from 2023 overstates revenue by exactly that amount. Both values nevertheless exceed the award value, because even after curtailment the fleet profile earns more in positive hours than the annual market value; the negative hours depress the annual market value, but the curtailed plant does not take part in them.

How much cannibalisation can a direct investment withstand?

More than a look at the series suggests, but not an unlimited amount. As long as a solar park receives an EEG market premium and the annual market value is below its award value, the premium offsets the falling market value: if the annual market value falls by one cent, the premium rises by one cent. Three channels remain open, and through them cannibalisation reaches the investment.

1. Hours with negative prices. In them there is no award value under § 51 EEG, so neither revenue nor premium; the direct marketer curtails. Every percentage point of generation that falls into such hours is yield that the yield assessment shows and the settlement does not. In 2025, by our analysis, it was 24 percent of fleet generation.
2. The plant's own profile against the fleet average. The premium is calculated on the annual market value of all solar plants, not on what the park itself earned. A park that delivers in the same hours as everyone else earns the fleet value; one that delivers more into the shoulder hours keeps the lead; one that lies below it, for instance because of a south-facing profile with a high midday peak, carries the gap alone. How large this difference is per layout is shown in South, east-west or vertical: which solar park layout pays off for investors.
3. Revenue without EEG backing. A power purchase agreement, continued operation after the support period or a park without an award hangs on the market value without a floor. What providers promise in such contracts and what they do not is sorted out in Understanding PPAs: how power purchase agreements make solar park revenues predictable; the years after the award are covered by A solar park after 20 years: continued operation, repowering or decommissioning?.

How strongly the first two channels act can be played through with two variables: the fleet's revenue in hours with a positive price and the share of generation in negative-price hours. From both follows the annual market value, and from that the premium. We hold the award value at 4.79 cents and let the plant curtail in negative hours.

The table reads in two directions. Downwards, with falling prices in the positive hours, revenue barely changes once the premium kicks in: going from 65 to 45 euros in positive hours costs 0.4 cents at a 20 percent negative share, from 55 to 45 euros only 0.16 cents. That is the protection the award provides. To the right, with a growing negative share, revenue falls in every row: at 55 euros the steps from 10 to 40 percent together cost 0.43 cents, at 45 euros 0.56 cents, which is 9 to 12 percent of the award value. The market premium secures the price, not the quantity; a model that carries the negative hours forward as a constant underestimates precisely the channel that is growing.

Why does cannibalisation hit solar harder than wind?

Because wind power spreads across the day and the year and solar power does not. In 2025 the annual market value for onshore wind was 7.441 cents, for offshore wind 8.059 cents and for solar 4.508 cents, at a spot average of 8.932 cents. The market value factor for onshore wind was therefore 0.83, for solar 0.50. Both technologies cannibalise themselves, but wind has more hours in which little other wind is blowing, and it produces in winter when prices are high. Solar produces at midday in summer, when every solar plant is producing.

The gap has reversed. Up to 2018 the monthly solar market value stood above that of onshore wind in every single month; in 2025 only in four of twelve months, in 2026 up to August in one. For an investor that is not an argument for wind but a pointer to how quickly a technology's value on the market can shift when build-out overtakes the profile.

Does storage build-out slow cannibalisation down?

It shifts it, it does not end it. Battery storage buys at midday when the price is lowest and sells in the evening when it is high. In doing so it lifts the midday price and lowers the evening price, and both narrow the spread on which cannibalisation feeds. In the 2026 figures this is not yet visible: the solar market value factor from January to August was 0.52, as in the previous year, and the negative-hour share of solar generation 23 percent. Storage build-out is still small against 16 gigawatts of new photovoltaics per year.

The honest counter-argument belongs here: the more storage supports the midday hours, the more every solar plant benefits without having a storage system itself. But the effect accrues to everyone at once and is therefore not an advantage of an individual project, and every new gigawatt of photovoltaics partly eats it up again. For a revenue forecast over 20 years, storage build-out is therefore an assumption you disclose, not a safety net you build on.

What does the EEG 2027 change about the market value?

Nothing about the market value itself, quite a lot about what a new park keeps of it. The government draft of the EEG 2027 continues to calculate the market premium for plants awarded from 2027 on the annual solar market value, but makes it two-sided: if the annual market value exceeds the award value, the operator pays the difference back as a refinancing contribution. A year like 2022 with a factor of 0.95 and an annual market value of 22 cents would be a year of repayment for such plants, not of windfall revenue. Which thresholds the draft provides for and what the 2019 to 2025 retrospective yields under this rule is calculated in EEG 2027: what the refinancing contribution means for a solar park direct investment; the draft is still in parliamentary deliberation.

Which questions should you ask the provider?

• Which market value does the revenue forecast use: the fleet's annual market value or a separate capture price for the park's own profile? And where does the figure come from?
• Is the market premium calculated monthly or annually, and does that match the plant's award date (Annex 1 No. 2 EEG)?
• What share of generation is assumed as curtailed in negative-price hours in the forecast, and does it rise over the term or stay constant?
• How is the market value set after the support period, if the model contains years without a premium?
• Does the model show revenue in positive hours separately from the premium, so that both assumptions can be checked?
• Is storage build-out included in the model as an assumption for rising midday prices, and if so, how?

How we treat the market value in project review

For every solar park we present to investors, we place the provider model's market value assumption next to the official series and next to our own hourly calculation for the park's specific profile. We check whether the premium is calculated under the right rule, how the negative-hour share is set over the term, and what the revenue still carries at a market value factor of 0.4 instead of 0.5. If that gives a different picture from the prospectus, it goes into our report. If you would like a specific offer reviewed or want to know how a direct investment fits into your situation, book an initial consultation.