When Owning Stops Paying: Leverage, Overhead and the Developer-to-IPP Shift

After years of developing projects to sell, many solar and storage developers set out to keep more of them. Owning operating assets promises stable revenue, a share of the long-term value that buyers used to capture, and freedom from selling into a weak market. For some developers it has worked. For others, it has turned a development business into a heavily indebted one, and then into a forced seller.

This final part of our series looks at how much leverage an owned portfolio can carry at each level of the capital structure, how much overhead the platform can afford, and the point at which keeping projects starts to destroy value.

Why developers keep projects, and how it can go wrong

The case for ownership is straightforward. Operating assets earn contracted revenue in hard currencies, smooth out lumpy development gains, and capture value that would otherwise go to the buyer. Part 2 showed why it became more tempting: buyers now pay little for development optionality, and many sellers are selling at or below book value.

The risk is in how ownership is funded. One large listed developer announced a target in early 2024 to own several gigawatts of operating solar and storage in the U.S. and Europe by 2026. By mid-2025 its owned solar was a fraction of that target. Debt in its development business rose by more than half in 18 months, interest expense reached about $40 million a quarter, and the business reported an operating loss in every period. In late 2025 it told investors it would step up project sales in 2026 to manage its debt. The constraint that forced the reversal was funding.

How much leverage each layer can carry

Leverage in a renewable portfolio sits at three levels, and each has its own limit. At the project level, lenders size debt to the asset’s cash flow. Norton Rose Fulbright’s 2026 survey of bank lenders puts the minimum P50 debt service coverage ratio for contracted solar at 1.25–1.30x and for contracted storage at 1.15–1.20x, rising to 1.75x and 2.0x for merchant projects. Lenders describe merchant exposure above about 25–30% as very difficult to finance.

In Europe, battery leverage depends on contracts. A Capcora and Suena study of German storage finds that a five-year full toll supports about 70% gearing, needing about €150/kW of equity, against about 60% and €250/kW for a merchant project with debt, and €550/kW for one funded entirely with equity. Lenders typically want around 80% of capacity tolled for five to seven years.

Above the projects, the cost of capital rises steeply. The same Norton Rose Fulbright survey prices clean construction loans at around SOFR plus 150 basis points, but holding-company loans at 350–450 basis points and development-stage facilities at up to 800.

Where leverage breaks: the parent

Project debt is non-recourse: if an asset underperforms, its lenders take the asset, not the developer. Debt at the parent is different. It sits behind project lenders and tax-equity investors, whose cash traps and sweeps can cut off the distributions the parent needs to pay it. That is where recent failures have started.

The investment-grade template keeps the parent light. Brookfield Renewable, rated BBB+, carries about $3.7 billion of corporate borrowings against $31.2 billion of non-recourse debt: about 89% of its debt sits at the assets. Clearway Energy targets corporate leverage of 4.0–4.5x and a BB rating, while its consolidated debt is about 7x adjusted EBITDA; the difference is the project debt the parent does not guarantee. Rating agencies watch parent metrics closely: S&P would downgrade EDP if funds from operations fall below 19% of debt, and names a missed asset-rotation target as a downgrade trigger.

Overhead: what the platform can afford

A development platform has to be paid for. LevelTen data on North American project sales in 2024 put late-stage projects at 5–12 cents per watt, or about $50–120 million per gigawatt. That margin has to cover development staff, interconnection deposits and every project that fails, before there is anything to reinvest. A developer that stops selling gives up that cash, and has to fund overhead and parent-level interest from operating assets instead.

There is no reliable industry benchmark for developer overhead per megawatt, so the test is internal: does cash from operating assets cover the platform’s overhead and parent-level interest without relying on asset sales? At the developer described above, revenue from asset sales in 2025 was smaller than the segment’s operating expenses. Lean owners look very different: Clearway’s reported G&A was about 3% of revenue in 2025, although it buys many services from its sponsor, so its full platform cost is higher.

The retain-or-sell test

The decision to keep an asset reduces to a comparison: is the asset worth more to the owner than to the best buyer? Two numbers drive it. One is the return the asset earns on the owner’s equity. The other is what that equity costs. Lotus Infrastructure puts the cost of equity for renewables assets at 10–12%, up from 7–9% a few years ago. Listed vehicles face a higher cost still: the two largest diversified UK listed renewables funds, Greencoat UK Wind and TRIG, trade at roughly 26–31% below net asset value with yields around 11%. Smaller solar and battery storage funds trade at wider discounts, as Part 2 sets out.

The gap in that chart is the argument for selling. Where buyers will underwrite a contracted solar asset at a 7–8% levered return and the owner’s equity costs 10–12%, the asset is worth more to the buyer. The most disciplined owners sell into that gap: Brookfield Renewable recycled a record $4.5 billion of assets in 2025 at about 2.4 times invested capital, and EDP Renováveis sold a contracted California solar and storage portfolio in July 2026 for a projected gain of about 40% on invested capital. Ownership makes sense for assets whose returns clear the owner’s cost of equity, which in practice means owners with low-cost capital, or assets with returns buyers will not pay for.

What it means for developers and investors

Data as of October 15, 2026. Return and leverage figures are drawn from lender surveys, rating agency reports and company filings and are indicative; their bases differ. The developer case is drawn from public filings and presented without naming the company. This piece is for informational purposes only and does not constitute investment advice.