Infrastructure Fund Underperformance: How Falling Costs Masked Rising Risk

Infrastructure funds are underperforming because falling renewable energy costs masked a simultaneous rise in revenue risk, climate-driven yield erosion, and model assumptions that no longer hold. The result: assets that looked cheaper to build became harder to profit from.

The Cost Paradox

Solar got cheap. Costs fell 90% in a decade. Battery costs dropped 40% in 2024 alone. The slide decks looked better than ever.

But infrastructure fund returns moved in the opposite direction. Pre-2020 renewable infrastructure projects routinely delivered 10 to 15% IRR. New projects are clearing at 5 to 8%. Bottom-quartile infrastructure funds from 2022 to 2024 vintages are returning negative IRRs. The gap between top-quartile and bottom-quartile outcomes has never been wider.

As one former CIO of a major global renewables platform stated: “These assets are not performing as initially underwritten. Solar plants are flowing maybe 5% below what the expectation was. Panels malfunctioning or degrading faster than expected, inverters failing more often and sooner than expected.”

This is not a story about costs. Costs did exactly what everyone expected. This is a story about what happened on the other side of the equation: what investors actually got paid for the energy those assets produced, and how a changing climate eroded the generation itself.

The Three Compressions

Three forces converged on infrastructure fund returns simultaneously. Understanding them individually matters, but what made 2022 to 2025 so punishing is that they moved together.

1. Capture Rate Collapse

LCOE measures the cost per megawatt-hour produced. It says nothing about what you get paid for that megawatt-hour. The industry learned this the hard way.

When everyone builds solar, everyone produces at the same midday window. Wholesale electricity prices during peak solar hours collapse. Germany’s PV capture rate, the share of the average wholesale price that solar generators actually receive, hit 52% in 2025. Spain reached 54%. Portugal, 50%. Solar generators across Central Europe are now earning roughly half the average wholesale electricity price.

Germany recorded 724 negative-price hours in 2025. Nearly a quarter of all German PV generation occurred during hours when the wholesale price was below zero. That means producers were paying to generate electricity. The trend is structural: annual RES capacity additions continue to outpace battery storage deployment through 2030, locking in further oversupply.

Capture rate collapse does not show up in an LCOE calculation. It does not appear in most P50 yield forecasts. But it directly determines whether an asset clears its return hurdle.

2. PPA Duration Compression

As revenue uncertainty grew, offtakers shortened their commitments. Twenty-year power purchase agreements became 10-year contracts. Then 7-year contracts. Each compression pushed more revenue into the merchant tail of the asset’s life, exactly the period with the highest price uncertainty.

In Spain, the PPA market effectively froze after 2024’s wave of negative pricing events. Buyers grew unwilling to lock in long-term pricing when the capture rate trend was still falling. For investors modelling returns based on contracted revenue, the shrinking PPA window meant modelling more years of unhedged exposure, more merchant risk, and wider confidence intervals around base-case returns.

3. Hurdle Rate Escalation

Meanwhile, the cost of capital moved against investors. Return hurdles shifted from approximately 6% to 8 to 10% as interest rates rose. Funds raised in 2022 and later must generate materially more operational growth just to hit their targets.

The mathematics are unforgiving. When your cost of capital increases by 200 to 400 basis points while your revenue per megawatt-hour is falling and your contracted revenue window is shrinking, the squeeze hits from every direction. There is no single variable to optimise. The entire risk-return profile of the asset class has shifted.

The Climate Variable Nobody Modelled

Behind the market-level compressions sits a more fundamental problem. The assets themselves are producing less energy than expected. And the gap between projections and reality is not random. It is directional.

The Degradation Gap

Financial models for solar assets typically assume panel degradation of 0.5% per year. NREL’s compendium confirms a median around 0.5% to 0.75% under laboratory conditions, but field data from operating portfolios in hot climates shows effective degradation closer to 1.0% per year when you include climate-driven efficiency losses. Over a 25-year asset life, the difference between 0.5% and 1.0% annual degradation compounds into a 10+ percentage point output shortfall.

But degradation is only part of the story. On top of the panel-level degradation, operators are seeing climate-driven yield drag: inverters shutting down under extreme heat, trackers malfunctioning in high winds, soiling losses increasing as dry spells lengthen. One climate risk specialist recounted that operators are examining “how to re-energise. The last year for a lot of wind operators in Europe, it was a particularly bad year. And the question across the boardroom for the next few months was: is this a bad year or is this a structural shift?”

The causes are often not the generation resource itself but the equipment responding to changing conditions. Inverters fail under heat stress. Trackers malfunction. Panels degrade faster than laboratory tests predicted. When you add climate-driven yield drag on top of higher-than-modelled degradation, the cumulative shortfall over a 20-year hold period can exceed what most financial models allow for.

Shifting Wind Patterns

Wind portfolio underperformance has been even more pronounced. Europe’s largest listed wind fund, Greencoat UK Wind, has missed its generation budget every year since 2016. In 2024, the energy yield shortfall reduced the fund’s NAV by GBP 146.8 million, equivalent to 6.5 pence per share. When the fund conducted a comprehensive energy yield review with an independent third party, 91% of assets received downward P50 revisions.

As one senior infrastructure executive reflected: “What I have noticed over the past 20 years is that wind patterns, particularly in Europe, might be changing. I don’t know if that is due to climate change or not. But my broad experience is the underwriting was over-optimistic.”

The pattern extends across the sector. SSE’s UK wind portfolio showed -60.4% cumulative underperformance against projections over the 2020 to 2025 period. Scottishpower, -43.4%. These are not small operators. If integrated energy companies with billion-pound portfolios are seeing this scale of underperformance, the issue is not operational. It is structural.

When P50 Stopped Working

The renewable energy industry built its financial models on P50 forecasts: the median expected production level, typically derived from 20 to 30 years of historical weather data. For decades, this was considered robust. The assumption was that wind and solar resources would mean-revert over time, that a bad year would be balanced by a good one.

That assumption is breaking down.

The problem is that P50 models are backward-looking in a forward-moving climate. When the historical dataset includes periods of unusually high wind speeds (as the 1990s were for Northern Europe), and those years fall out of the 20-year rolling average, the “long-term mean” itself shifts downward. The benchmark is moving, and models calibrated to an older baseline are systematically overestimating future generation.

DNV’s 2025 half-year wind update confirmed a “wind drought” across the North Sea, with windiness falling 4% to 8% below normal. DNV attributed this to a “poleward displacement” of the jet stream, redirecting weather systems away from the UK. Whether this is cyclical or structural remains debated. But for investors with 5 to 7-year hold periods, the distinction is academic. Underperformance in the hold period is underperformance in the returns.

A veteran renewables investor who has underwritten assets across two decades summed it up: “The underwriting was over optimistic. People that have been in the industry for long enough noticed that 10 to 5 years ago. People that have not been in the street for that long are only realising that more recently.”

The market is already responding. As one climate risk researcher recounted from a recent conversation with a major European utility: “They started implementing parametric insurance to ensure their yields. There is a new product for when they have bad yield years. They’re protecting against P50 to P90 downside risk.”

When utilities are buying insurance against their own yield forecasts, the P50 is no longer a planning tool. It is a liability.

The practical consequence: when you base your DCF on a P50 that no longer reflects actual conditions, you are pricing the asset at a valuation the weather will not support. The gap between modelled and actual returns is not noise. It is signal.

The Dispersion Problem

If infrastructure fund underperformance were uniform, it would be a market-timing problem. But it is not uniform. The dispersion between top-quartile and bottom-quartile infrastructure funds has widened dramatically.

Top-quartile infrastructure funds are delivering 12%+ IRR. Bottom-quartile funds from the same vintages are in negative territory. Private debt tells a similar story: bottom-quartile IRRs sit in the mid-single digits while top-quartile funds reach the mid-teens. The spread between winners and losers is 6 to 7 percentage points.

What distinguishes the top quartile? Three factors emerge from the data.

Geography. Portfolios concentrated in Northern and Western European wind (France, Germany, Finland) have experienced the worst underperformance. Operators with UK exposure are now showing recovery signals. Those with Southern European or Iberian solar-plus-storage positions are capturing value from price volatility rather than suffering from it. Geographic selection is not a nice-to-have. It is the primary driver of return dispersion.

Technology mix. Pure-wind operators have been hit hardest. Operators with solar exposure experienced less severe shortfalls (though still negative). Those with battery storage co-location are capturing the price spreads that are punishing standalone assets. Daily price spreads in Spain hit a record EUR 94/MWh in 2025. A standalone solar asset can only participate in the low-price hours. A solar-plus-storage hybrid captures the spread.

Risk quantification. Part of the problem is structural: nobody in the value chain owns the risk. As one infrastructure risk specialist described it, after mapping conversations across the entire energy value chain: “When we talk to the operators, they say investors should demand better modelling. When we talk to the originators, they say the lenders are controlling the assumptions. The lenders say the insurers and the regulators set the risk framework. And the insurers say they have to model out the prices from the certification bodies. Blame keeps getting pointed around.”

The funds that are outperforming are the ones that have moved beyond generic climate scores to asset-level, forward-looking analysis. The difference between modelling one-third of gross yield using historical data and modelling full gross yield including forward-looking atmospheric projections and operational impacts is the difference between quartiles.

What the European Market Is Telling Us Right Now

The European renewable energy market in 2026 provides a real-time stress test of every assumption infrastructure investors relied on.

Germany: 724 negative-price hours in 2025. PV capture rate at 52%. Almost a quarter of all solar generation occurring during negative prices. New legislation now removes compensation during negative prices for new PV installations. The standalone solar business case in Germany is structurally challenged.

Spain: Solar valuations crashed 42% since Q1 2024, from EUR 916k/MWp to EUR 648k/MWp. Capture rates fell from 83% in 2023 to 56% in 2025. Over 500 hours of negative pricing. Grid saturation at 83.4% of connection nodes. Curtailment hit 11% in July 2025.

France: Record 1.7 TWh of renewable energy curtailed in 2024. Nuclear plants now flex 6GW daily to compete with solar at midday. Ten percent of France’s green energy was wasted in the first half of 2025.

UK Wind: Every major listed wind fund missed its generation budget in 2024. Greencoat UK Wind: -13% below budget. Octopus Renewables: onshore wind -18% below budget. TRIG: -7% below budget in Q3 2025. DNV confirmed a North Sea wind drought with windiness 4 to 8% below normal.

Netherlands: 90% of businesses affected by grid congestion. Over 10,000 users and 7,500 projects waiting for grid connections. Even well-performing assets cannot sell what they cannot connect.

None of these data points are hidden. They are in annual reports, grid operator filings, and market analytics. But most infrastructure fund models have not been updated to reflect them. The industry’s standard assumptions, P50 yield, contracted revenue, historical degradation rates, were built for a market that no longer exists.

What Separates Winners from Losers

The gap between top-performing and bottom-performing infrastructure funds is not closing. It is widening. And the pattern that emerges from the data suggests the divergence is not temporary.

The funds that are navigating this environment share several characteristics.

They model forward, not backward. Traditional climate risk assessment uses historical weather data to project future performance. The funds that are outperforming have moved to forward-looking yield projections that account for shifting atmospheric conditions, not just last decade’s averages. The difference between backward-looking P50 and forward-looking, asset-specific yield modelling is measurable. Portfolios using forward-looking analysis identified the 2020 to 2025 wind drought before it showed up in earnings.

They price the full yield stack. Most yield assessments model only a fraction of gross yield, typically just irradiation for solar or wind availability for wind. They miss the interaction between atmospheric variables (irradiation, cooling effects, heating effects for solar; wind speed, air density, hub-height dynamics for wind) and they stop before netting down to operational reality. The operational layer, including soiling losses, inverter shutdowns, safety tracking downtime, blade icing, and offshore access constraints, can account for an additional 3 to 5% of yield variance that never appears in standard models.

They use volatility, not avoid it. The same market dynamics causing underperformance, oversupply, negative pricing, capture rate collapse, are creating opportunities for investors positioned to capture price spreads rather than suffer from them. Battery storage co-located with solar in markets like Spain (EUR 94/MWh daily spreads) and Germany (EUR 80 to 90/MWh spreads) is delivering 15 to 19% IRR on day-ahead trading alone in parts of Central and Eastern Europe.

They stress-test climate assumptions. LP expectations are evolving. 46% of institutional investors now say GPs should do more to weigh climate risks, up from 38% the prior year. 43% consider themselves underallocated to infrastructure. The capital is available. But it is increasingly conditional on demonstrating that risk has been quantified, not just acknowledged.

The cost-risk paradox that defined 2020 to 2025 is not reversing. Cheaper assets attract more deployment, which floods the grid, which crashes the revenue you receive, which compounds the impact of any yield shortfall. Understanding this dynamic, and pricing it accurately, is no longer a competitive advantage. It is a prerequisite for fund survival.

Frequently Asked Questions

Why are infrastructure funds underperforming despite falling renewable energy costs?

Falling costs increased deployment, which flooded electricity markets and crashed the prices renewable generators actually receive. German solar capture rates hit 52% in 2025, meaning generators earned roughly half the average wholesale price. Simultaneously, climate-driven yield erosion meant assets produced less energy than modelled, and rising interest rates increased the return hurdle funds needed to clear. The compression came from three directions at once: lower revenue per MWh, lower generation volumes, and higher cost of capital.

How much are renewable energy assets underperforming their forecasts?

Solar assets are underperforming expectations by approximately 5% on average, while wind assets are 5 to 10% below. Field data shows panel degradation running at roughly 1.0% per year versus the 0.5% assumed in most financial models. For wind, Greencoat UK Wind, Europe’s largest listed wind fund, has missed its generation budget every year since 2016, with 91% of its portfolio requiring downward P50 revisions.

What is causing the gap between top-performing and bottom-performing infrastructure funds?

Geographic selection, technology mix, and risk quantification are the three primary differentiators. Portfolios concentrated in challenged markets (France, Germany wind) have significantly underperformed. Those with storage co-location, exposure to recovery markets, and forward-looking yield analysis are outperforming. The IRR spread between top-quartile and bottom-quartile infrastructure funds now exceeds 12 percentage points.

Are P50 forecasts still reliable for renewable energy investment?

P50 forecasts are increasingly questioned. They are backward-looking models calibrated on historical weather data that may no longer represent future conditions. Climate change is shifting wind patterns, increasing temperature extremes, and altering atmospheric conditions in ways that systematically bias P50 models toward overestimation. When Greencoat UK Wind rebased its P50s, 91% of assets were revised downward, with 13% of assets requiring reductions exceeding 10%.

How does climate change directly affect infrastructure fund returns?

Climate change affects returns through two main channels. First, chronic yield erosion: heat stress degrades inverter efficiency, shifting wind patterns reduce generation volumes, and extreme weather increases equipment downtime. Second, market-level impacts: the rapid buildout of renewables driven by the energy transition floods grids, collapses capture rates, and creates negative pricing events that undermine revenue assumptions. Both channels are amplified by climate change, and most financial models account for neither.