First Solar's record quarter rests on $2.1 billion of tax credits Washington keeps poking at

---

There is a particular kind of corporate earnings report that arrives dressed for a victory lap and leaves you counting the exits. First Solar's first quarter of 2026, announced on April 30, was exactly that kind of report. The headline numbers were the sort that earn a green tape and an analyst upgrade: net sales of $1.04 billion, up roughly 24% from the $844.6 million of a year earlier; net income of $347 million, or $3.22 per diluted share, against $210 million and $1.95 the prior year; adjusted EBITDA of $520 million versus $379 million. Management reaffirmed full-year 2026 guidance — 17.0 to 18.2 gigawatts of volume sold, $4.9 billion to $5.2 billion of net sales, $2.6 billion to $2.8 billion of adjusted EBITDA — without flinching. Gross margin, that most-watched line, climbed to 46.6% from 40.8%. The company shipped a record 3.8 GW of its cadmium-telluride thin-film modules and ran its U.S. factories at a 96% utilization rate.

It is a beautiful set of numbers. It is also, on inspection, a set of numbers that does not belong primarily to First Solar's customers. It belongs to American taxpayers, routed through the Internal Revenue Code. The forensic question this article asks is simple and uncomfortable: how much of First Solar's profitability is an industrial achievement, and how much is a transfer payment that exists only because a particular statute is still on the books? The answer, once you do the arithmetic the press release would rather you not do, is that the subsidy is not a tailwind to the earnings. The subsidy is, to a first approximation, the earnings.

The credit is the same size as the profit

Start with the number that matters most and that the company discloses in plain sight. First Solar told investors it expects to generate between $2.10 billion and $2.19 billion of Section 45X advanced-manufacturing production credits during 2026. Section 45X, created by the 2022 Inflation Reduction Act, pays a per-watt bounty to companies that manufacture solar components inside the United States. For a thin-film module maker running its fabs flat out on domestic soil, it is close to free money: produce the watt, claim the credit, sell or monetize it.

Now set that subsidy beside the profit it is supposed to be merely "supporting." Full-year 2026 adjusted EBITDA guidance is $2.6 billion to $2.8 billion. The Section 45X credits alone, at the midpoint of around $2.15 billion, are equivalent to roughly 75% to 80% of the entire adjusted EBITDA the company expects to earn this year. Put more starkly: if Section 45X disappeared tomorrow and First Solar sold exactly the same modules at exactly the same prices, the great majority of its operating profitability would vanish with it. The modules would still ship. The factories would still hum. The income statement would collapse.

This is not a hidden footnote First Solar is concealing — it is, to the company's credit, disclosed and quantified. But the framing matters enormously. The market values First Solar as a high-margin technology manufacturer with a structural cost advantage in CdTe thin-film. The financial statements describe something closer to a toll operator collecting a government-set tariff, where the toll is set in Washington and can be changed in Washington. A 46.6% gross margin is an extraordinary number for any manufacturer of a commoditizing energy product. It is far less extraordinary when a substantial slice of it is a per-unit federal payment rather than pricing power wrung from customers.

A subsidy you have already seen converted into cash

Skeptics of the "it's all a paper credit" critique will object that these are accounting artifacts, not real money. First Solar has helpfully refuted its own defenders on that point. The credits are real and the company has been turning them into hard cash. In 2025, First Solar monetized roughly $1.6 billion of Section 45X credits. In October 2025 it signed two separate agreements to sell $699.7 million of 2025-vintage 45X credits for aggregate cash proceeds of $668.2 million — selling the credits at a few cents on the dollar discount to a counterparty hungry for tax offsets. Cumulative 45X transfer deals have since passed the $2 billion mark.

That cash is genuine. That is precisely the problem. A company whose cash generation depends on finding buyers for federal tax credits is exposed to two distinct risks stacked on top of each other. First, the statute that creates the credit can be altered. Second, the market for buying the credit — the universe of profitable corporations willing to purchase transferable credits to lower their own tax bills — can shrink if corporate tax rates fall or if the transfer mechanism itself is curtailed. First Solar is not merely betting that solar demand persists. It is betting that a specific, politically contested feature of the tax code persists, and that a liquid secondary market to monetize it persists alongside. Two policy bets dressed as one industrial story.

The statute is a moving target, not a fixture

Here is where the foreboding earns its keep. Section 45X was written into law in 2022 with a generous runway, but the political ground beneath it has been shifting ever since. The credit was originally designed to phase down over the back half of the decade and into the early 2030s; subsequent legislation and rulemaking have layered on conditions. Chief among them are the "Foreign Entity of Concern" (FEOC) provisions — domestic-content and supply-chain-sourcing requirements that ratchet upward over time, restricting which inputs and ownership structures qualify for the credit. A manufacturer that fails to keep its supply chain on the right side of an evolving FEOC line risks watching its per-watt subsidy shrink or evaporate even while the underlying statute technically survives.

The blunt reality is that 45X is a creature of a tax code the current Congress and administration have spent two years actively rewriting. Project-level incentives — the investment tax credit and the production tax credit that subsidize the buyers of solar, not the makers — have already seen their timelines compressed relative to the IRA's original promise. Manufacturing credits like 45X have so far fared better, with phase-outs pushed toward the early 2030s rather than killed outright. But "so far" is the operative phrase. A subsidy whose schedule has been renegotiated repeatedly is not a fixed asset on which to underwrite an equity. It is a variable in someone else's spreadsheet, and the people holding that spreadsheet do not work for First Solar's shareholders.

When the bull case for a stock requires you to forecast not just module demand and manufacturing cost but the durability of a specific tax provision through multiple election cycles, the analyst has quietly become a political prognosticator. Equity markets are bad at pricing legislative risk because legislative risk is binary and lumpy: nothing happens, nothing happens, nothing happens, and then a markup in a reconciliation bill rewrites the economics overnight. First Solar's stock has already lurched on exactly these fears, falling sharply on days when policy and trade scrutiny dominated the tape rather than fundamentals.

The bookings cliff hiding behind the backlog

If the subsidy were the only crack, one might wave it away as a known, disclosed, slowly-evolving risk. It is not the only crack. The second forensic problem lives in the gap between First Solar's enormous backlog and its suddenly anemic order flow.

The company entered 2026 with a contracted backlog of 47.9 GW, carrying an aggregate transaction price of $14.4 billion. That number is the centerpiece of the bull case: years of visibility, demand locked in, factories spoken for. But a backlog is a stock, and what keeps a backlog alive is the flow of new bookings replacing what gets delivered. In the first quarter of 2026, First Solar booked just 1.7 GW of gross new orders. Against a backlog of nearly 48 GW that the company is steadily shipping down, 1.7 GW per quarter is not replenishment — it is depletion in slow motion. At that pace, refilling the pipeline would take something on the order of 28 quarters. Seven years. That is not a growth company adding to its order book; that is a company drawing down a reservoir it filled in a friendlier policy era.

There is a reason new bookings have gone quiet, and it loops directly back to the first problem. Developers signing multi-year module contracts today must price in the same policy uncertainty that hangs over First Solar's credits — uncertainty about the investment tax credit, about tariffs, about interconnection, about the cost of capital for utility-scale solar in a higher-rate world. When the rules are in flux, buyers wait. A 47.9 GW backlog booked under the old regime can mask, for several quarters, the fact that the new regime is producing a trickle of orders. The income statement looks like a growth story precisely because it is delivering yesterday's bookings. The bookings line — the leading indicator — is flashing the opposite.

The denominator illusion in the margin

It is worth dwelling on that 46.6% gross margin, because it is the number that does the most persuasive work in the bull case and the number most distorted by the subsidy. A gross margin is revenue minus cost of goods sold, over revenue. But Section 45X credits effectively act as a reduction in the per-unit cost of manufacturing — they flow through in a way that flatters the very margin investors cite as proof of First Solar's technological moat.

Ask the right counterfactual question: what is First Solar's gross margin on a module if you remove the manufacturing credit attached to that module? The disclosed figures do not let an outsider compute that to the decimal, but the direction is unambiguous and the magnitude is large. A company harvesting more than $2 billion of credits against roughly $5 billion of net sales is earning a subsidy worth on the order of 40 cents of credit for every dollar of revenue. Layer that onto reported margins and the "structural cost advantage" of CdTe thin-film starts to look at least partly like a structural subsidy advantage — an advantage that exists because First Solar manufactures in the United States and claims the bounty, not solely because its physics are superior. The moat may be real, but a meaningful portion of it is a moat the government dug.

This is the quality-of-earnings heart of the matter. Two companies can report identical margins, and one earns it from customers while the other earns it from the Treasury. The market pays a premium multiple for the first kind of margin because it is durable and within management's control. It should pay a discount for the second kind, because it is contingent on a vote. First Solar is being valued, much of the time, as the first kind of company while reporting the economics of the second.

Customer concentration and the BP cautionary tale

The backlog is not only thinning at the top; it has already proven it can crack at the contract level. First Solar is in active litigation after a counterparty walked away from a large module commitment. The company terminated a contract "for customer default" with affiliates of BP involving the supply of 6.6 GW of modules, and filed a breach-of-contract suit against BP Solar Holding and Lightsource entities seeking to recover not less than $405 million, contending that the defendants failed to make required down payments and owe termination payments exceeding $400 million.

Litigation outcomes are not yet adjudicated, and the allegations are First Solar's; BP's affiliates will tell their own story in court. But the episode is instructive regardless of who ultimately prevails. A 47.9 GW backlog with a $14.4 billion transaction price is only as solid as the creditworthiness and resolve of the developers behind it. When 6.6 GW — more than a tenth of the headline backlog — can become a courtroom dispute, the comfort investors take from "years of contracted visibility" deserves a discount for the optionality those contracts implicitly hand to buyers. In a softening, policy-uncertain demand environment, a contracted backlog booked at yesterday's prices is exactly what a developer wants to renegotiate or escape. The BP fight is a live demonstration that the backlog is a probability distribution, not a promise.

Tariffs: a tailwind that is also a hostage

To be fair to the other side of the ledger — and we will be more fair still in a moment — trade policy is one area where First Solar's exposure cuts in its favor. Because First Solar makes cadmium-telluride thin-film modules rather than crystalline-silicon panels, it sidesteps much of the tariff machinery aimed at imported silicon cells and polysilicon. A potential Section 232 tariff on imported polysilicon components, widely discussed as imminent, would raise rivals' costs while leaving First Solar's CdTe supply chain comparatively insulated. Analysts have called Section 232 a likely positive catalyst for First Solar for exactly this reason.

But notice what that argument concedes. The bull case on tariffs is, at bottom, another bet on the durability and direction of federal policy. First Solar wins if Washington taxes imported silicon, wins if Washington keeps 45X intact, and wins if Washington preserves the project-level credits that make its customers' projects financeable. That is three favorable policy outcomes the equity is implicitly underwriting. A company whose competitive position depends this heavily on the simultaneous persistence of multiple government interventions is not insulated from politics. It is leveraged to it. Tariffs are a tailwind today and a hostage tomorrow; the same authority that can impose a Section 232 duty can lift it, and the same Congress that can extend 45X can let it lapse.

Priced for the policy to hold

Pull the threads together and the asymmetry becomes the thesis. First Solar trades as a premium clean-energy manufacturer with secular tailwinds, a fortress backlog, and best-in-class margins. Every one of those attributes, examined forensically, is contingent on policy holding its current shape: the margins lean on 45X; the backlog was booked under incentives now in flux and is being depleted faster than it is refilled; the tariff "moat" is a federal decision that could go the other way; and even the cash conversion depends on a functioning market for transferable credits that itself depends on the tax code.

In a priced-for-perfection setup, the buyer is paid little for the upside and exposed fully to the downside. If 45X survives intact through 2032, if FEOC rules prove navigable, if Section 232 lands favorably, if developers resume booking at scale, and if the BP-style defaults stay rare, First Solar earns roughly what the guidance promises and the stock is reasonable. But that is five contingencies that must all break the company's way, several of them outside management's control entirely. Lose any one — a 45X markdown in a reconciliation bill, a tightening FEOC interpretation, a wave of contract renegotiations as policy uncertainty drags on — and the earnings base that justifies the multiple is suddenly a fraction of what the screen says. That is the definition of negative asymmetry: capped upside if everything works, severe downside if any single political variable turns.

What the bulls genuinely get right

The bear case above is real, but intellectual honesty requires conceding where the bull thesis is genuinely strong — and it is strong in several places that matter.

First, the disclosure is exemplary. First Solar does not hide its 45X dependence; it quantifies the credit guidance to the tens of millions and lets investors do exactly the arithmetic this article performed. That transparency is the opposite of the obfuscation that usually accompanies a forensic short, and it deserves credit.

Second, the manufacturing achievement is authentic. A 96% utilization rate, a record 3.8 GW shipped in a quarter, and a genuine technological differentiation in CdTe thin-film are not subsidy artifacts. First Solar built real factories that make real modules at real scale on American soil, and its cost structure in thin-film is a legitimate engineering accomplishment that predates the IRA. The company was profitable and competitive before 45X existed; the credit supercharged an already-viable business rather than conjuring one from nothing.

Third, the backlog, for all its policy-era vintage, remains enormous and contracted at attractive prices. New U.S. utility-scale bookings in the quarter came in around $0.35 per watt inclusive of adjusters — a healthy price in a commoditizing market — and 47.9 GW of visibility is a luxury most manufacturers would trade a great deal for. Even a depleting reservoir of that size buys years of runway.

Fourth, on the policy bets themselves, the bull is not obviously wrong. Manufacturing credits like 45X have so far survived the legislative churn better than project credits, precisely because domestic-manufacturing incentives enjoy bipartisan support in a way that consumer-side green subsidies do not. Reshoring and energy-security politics cut in First Solar's favor. And the Section 232 tariff dynamic genuinely could lift First Solar's relative position. The bull case is not a fantasy; it is a coherent bet that the policy scaffolding holds. The disagreement is about how much an investor should pay for a business whose profitability is that scaffolding-dependent — not about whether the business is real.

The kicker

Markets adore a clean industrial story: a company with a better mousetrap, fat margins, and years of orders. First Solar offers a version of that story, and the modules are real, the factories are real, and the engineering is real. But the margins that make the story sing are, to the tune of roughly $2 billion a year, a line item in the U.S. tax code — and the bookings that are supposed to refill the backlog have slowed to a 1.7 GW trickle while the income statement keeps delivering the fat orders of a friendlier era. Strip out the subsidy and you have a good manufacturer of a commoditizing product trading like a secular growth compounder. Keep the subsidy and you have a wonderful business whose wonderfulness is renewed, or revoked, by people who do not answer to its shareholders and who change the rules every session. That is not a moat. It is a lease.

The most dangerous number on First Solar's income statement is the one Washington can erase with a single line in a reconciliation bill, and right now it is roughly the size of the entire profit.