Solar module prices are lower right now than at any point in the industry’s history. That fact alone tells you almost nothing about whether a project will get built. Tier-1 mono TOPCon panels are trading FOB China at $0.085 to $0.095 per watt, according to BloombergNEF, down 38% from the 2022 peak. Manufacturers can barely sell below that. Polysilicon inventory topped 570,000 metric tons in early 2026, per InfoLink Consulting, enough latent supply for 300 gigawatts of modules nobody has ordered yet.
Cheap hardware and bankable projects are two different problems. A lender financing a 500 MW plant for 18 years doesn’t care much about the sticker price on a pallet of modules. It cares whether the plant will still be producing rated output in year 15, whether an insurer will pay out if it doesn’t, and whether the offtaker on the power purchase agreement will still be solvent to honor it. Getting from low-cost modules to a financeable project means solving for all three at once.
Why price alone doesn’t get you financed
A 2024 procurement spec might have said “M10 monofacial, 540 watts.” A 2026 spec reads more like a bifacial, high-efficiency TOPCon panel rated above 700 watts, from a BNEF-listed tier-1 manufacturer, with warranties an insurer will actually underwrite. That shift matters because balance-of-system costs scale with module count, not wattage. A higher power class lowers the installed cost per kilowatt even when the price per watt looks similar on paper.
Origin matters just as much as spec. The same physical module trades near $0.09/W out of China. Rebuild the supply chain in Southeast Asia with non-Chinese polysilicon and it climbs toward $0.20/W. Source it fully domestic in the US and it can reach $0.47/W, based on BloombergNEF’s 2026 estimates. US buyers are already living this: median module pricing sat at $0.28/W in Q1 2026 as Foreign Entity of Concern rules and stacked anti-dumping duties reset the market, according to procurement platform Anza. None of that shows up if you’re only comparing cents-per-watt across a spreadsheet of quotes.
What lenders are actually underwriting
Project finance debt is non-recourse or limited-recourse. Repayment depends on the cash flow the plant generates, not the sponsor’s balance sheet. That changes what “bankable” means at every stage.
On offtake, lenders want a long-dated PPA, often 20 to 35 years, with a creditworthy counterparty and tariff risk clearly allocated. On hardware, they want modules from manufacturers with a multi-year audited track record, backed by warranties an insurer will price. On delivery, they want an EPC contractor with a completion guarantee and scale experience, so schedule slippage isn’t the base case. Then they model it all against a minimum debt service coverage ratio, typically 1.2 to 1.4, before construction starts.
Two projects built five years apart show how these pieces come together.
Case study: Bhadla Solar Park, Rajasthan
Bhadla sits in the Thar Desert, where summer temperatures pass 48°C and dust storms are routine. Rajasthan’s state renewable energy corporation had to solve the biggest barrier to financing desert-scale solar before developers would even bid: infrastructure risk. Instead of leaving each developer to build its own substations and transmission access, the state built shared infrastructure and leased land at subsidized rates across four bidding phases between 2015 and 2020.
That de-risking pulled in concessional finance alongside commercial debt. The Asian Development Bank put in $175 million for transmission, plus $50 million from the Clean Technology Fund. The World Bank added $100 million for shared park infrastructure, and individual project vehicles drew loans as small as $29 million apiece. Developers including Adani, Acme, SB Energy, and Hero Future Energies then competed for capacity through SECI and NTPC auctions, with winning tariffs landing between ₹2.44 and ₹2.62 per kWh. The park now runs at 2,245 MW across roughly ten million panels, with around 2,000 robotic cleaners handling the dust problem the site itself created.
Case study: Al Dhafra PV2, Abu Dhabi
Al Dhafra shows what bankability looks like in a market that’s matured past needing concessional support. The 2 GW plant, owned by TAQA, Masdar, EDF Renewables, and Jinko Power, went from tender launch to financial close in under two years, backed entirely by commercial lenders. Bank of China, BNP Paribas, Crédit Agricole, HSBC, MUFG, Sumitomo Mitsui, and Standard Chartered arranged roughly $1 billion in limited-recourse debt.
The winning bid came in at 1.35 US cents per kWh on a levelised cost basis, a world record at the time, and financial close nudged it lower still, to 1.32 cents, through hedging and financing optimisation. That tariff held because the offtaker signed a multi-decade PPA and because the four million bifacial modules on single-axis trackers carried warranties strong enough for underwriters to actually price the risk. In January 2025, the project refinanced through an $870.75 million green bond, certified under the Climate Bonds Standard, a sign the original financing package held up in practice.
AgriPV and the conference circuit
The next bankability test is already showing up on panels at the Solar Power conference circuit and the Renewable Energy Summit series: AgriPV, where panels sit above active farmland or grazing land instead of replacing it. Lenders are starting to ask for crop-yield and land-lease documentation alongside the usual PPA and warranty package, since dual land-use terms affect both revenue assumptions and permitting timelines. It’s not yet a standard underwriting line item the way module bankability tiers are, but developers pitching AgriPV projects at these events are already treating it as one.
Frequently Asked Questions
1. What makes a module bankable rather than just cheap?
A bankable module comes from a manufacturer with several years of audited financial history, carries product and performance warranties a third-party insurer will actually underwrite, and has independent test data confirming it holds rated output over time. A cheap module from a thinly capitalised or newly listed manufacturer can match the same spec sheet and still get rejected because there’s no confidence the warranty means anything by year 12.
2. Why doesn’t module price per watt tell the full story anymore?
Because origin, power class, and compliance status now move the effective cost as much as the sticker price does. A Southeast Asian module built with non-Chinese polysilicon can cost more than double its China FOB equivalent. A US-delivered module can cost three to five times as much once tariffs and domestic content rules are factored in. Comparing quotes without normalizing for these variables gives you a misleading number.
3. How do lenders decide whether to finance a utility-scale solar project?
They model the plant’s cash flow against its debt obligations and generally require a minimum debt service coverage ratio before releasing funds. That model depends on a creditworthy long-term offtaker, an EPC contractor with a completion guarantee, and hardware with warranties an insurer will stand behind. Weakness in any one of those areas raises the cost of debt, or stalls financial close entirely, regardless of how competitive the module price looks on its own.
