Story
The Full Story
Between Q3 2022 and Q1 2026, First Solar's story arc bent in three distinct ways: from IRA-fueled hyper-growth (2022–2023, "sold out through 2026" at record ASPs), to selective discipline under Chinese pressure (2024, AD/CVD petition, CuRe launch, but the Series 7 warranty bombshell), to policy whiplash and backlog erosion (2025–2026, tariffs, FEOC, the BP affiliate default that vaporized 6.6 GW of bookings, and a 2026 revenue guide below 2025). Management's core message — differentiated CdTe, domestic supply chain, balanced approach — has been remarkably consistent. What has changed is whether the evidence still supports it: backlog has fallen from 80.1 GW to 50.1 GW, the original $17–$20 FY25 EPS guide collapsed to $14.21 actual, CuRe has slipped over a year past its original Q4 2024 launch date, and 2026 is the first year since the IRA where management has stopped guiding to EPS at all.
1. The Narrative Arc
The arc is V-shaped. Late 2022 through 2023 was the most bullish period in First Solar's modern history: the IRA gave management permission to pre-sell capacity through 2027 at $0.30+ ASPs, and they did. From early 2024 onward, the story has been a slow concession that the IRA tailwind was insufficient against three forces management did not fully anticipate: (1) Chinese oversupply that crushed module pricing globally, (2) a domestic policy regime that turned hostile to non-domestic content even from a U.S. champion, and (3) a Series 7 manufacturing issue plus a major customer default that exposed the fragility of the "sold out" narrative.
2. What Management Emphasized — and Then Stopped Emphasizing
Three patterns stand out:
What disappeared. "Tandem with crystalline silicon" was a featured roadmap pillar in 2023; by Q4 2024 management had flipped to "no tandem without thin film." Europe was a stated growth corridor through 2023; the Q4 2024 call disclosed that the EU sales office had been shut down. "Sold out through 2026" was repeated nearly every quarter in 2022–2023 and is now absent.
What replaced it. Trade enforcement (AD/CVD), IP litigation (TOPCon), and FEOC restrictions have moved from minor footnotes to dominant talking points. By Q4 2025, IP enforcement consumed multiple paragraphs of prepared remarks. The narrative has shifted from offense (capture demand) to defense (deny demand to Chinese-tied competitors).
What persisted. The differentiated CdTe pitch, the IRA / 45X anchor, and the "balanced growth, liquidity, profitability" framework have been repeated verbatim across 14 quarters. These are the load-bearing beams of the story.
3. Risk Evolution
The 2025 10-K added entire new sub-sections on IEEPA tariffs, Section 232 derivative-product investigations, port fees on Chinese vessels, and OBBBA tax-credit acceleration — none of which existed as discrete risks in the 2021 filing. Customer-termination risk, which got one boilerplate paragraph in 2021, became a multi-paragraph disclosure in 2025 with a named litigation against BP affiliates. Conversely, COVID and shipping demurrage — the dominant 2021–2022 risk — has effectively disappeared. The shift in risk vocabulary is the cleanest tell that this is no longer the same business: the 2021 10-K read as a manufacturing operator worried about cost curves; the 2025 10-K reads as a U.S. policy beneficiary worried about which way the policy and the courts swing next.
4. How They Handled Bad News
Series 7 manufacturing issue (Q3 2024). Disclosed cleanly when found, with a $50M Q3 charge plus a $56–100M warranty range. The first-quarter response from analysts (Phil Shen at ROTH) flagged that some customers were privately reporting 5–7% underperformance — implying project-level economics could be at risk beyond the warranty cap. By Q4 2025 the range had narrowed to $35–75M with $50M reserved, and BP affiliates filed counterclaims alleging "overall PV plant underperformance." Management's response — pointing to "third-party prediction modeling, weather variability, terrain variability, EPC quality" — sounds defensible but reads as classic deflection. Verdict: disclosure was timely; the through-line on whether the issue is fully behind them is still uncertain.
CuRe rollout slippage. Original plan (Q1 2024 call): launch on Ohio lead line in Q4 2024, full fleet replication by 2026. Q4 2024 call: pushed to Q1 2026 "to address key manufacturing learnings from the initial production run." Q4 2025 call: now expected to convert in Q1 2026 with first Series 7 line in India in early 2027. That's two-plus years of cumulative slippage on the most-touted technology promise. Each disclosure was incremental, never framed as a setback — always couched as "disciplined, phase-gate introduction." The slippage is real; the framing is generous.
FY2025 guidance reset. February 2025 initial guide: $17–$20 EPS. April 2025 (post-tariffs): cut to $12.50–$17.50. November 2025: cut again to $14–$15. February 2026 actual: $14.21. That is roughly a $4 per-share reset on the original midpoint, justified by tariffs (genuinely exogenous) plus the BP default plus an Alabama glass-supply hiccup. The sequencing matters: management was guiding to 50% YoY EPS growth nine months before delivering 18%. Customers like the BP affiliates would presumably have shown stress in late 2024; the original FY25 guide either underweighted that signal or was outright optimistic.
The "sold out through 2026" framing. In Q3 2022 and through 2023, management repeated nearly every quarter that they were "sold out through 2026." That framing was technically correct but quietly stopped working in 2024 as customers began invoking shift rights, terminations, and defaults. By Q4 2024, ~2.4 GW of contracts had been terminated; in 2025, another 8.3 GW of debookings (6.6 GW from BP affiliates alone). The contracted backlog peaked at 80.1 GW (Q4 2023) and has fallen 37% to 50.1 GW. To management's credit, they preempted the issue by emphasizing termination clauses and payment security; to their debit, the headline number was always more attractive than its underlying quality, and the gap is now visible.
5. Guidance Track Record
Three of the last four full years missed the initial guide midpoint, and the FY2025 miss is the largest in absolute dollar terms (~$4.30 per share). FY2023 — the only clean beat — was the year when 45X credit recognition first kicked in at scale, which was largely an accounting tailwind. The pattern: management is reliable on cost structure and production volume, less reliable when projecting the demand side, and systematically too optimistic about how customer behavior and policy interact in the year ahead. The 2026 switch from EPS guide to EBITDA guide is itself a tell — Pillar Two tax volatility is the stated reason, but it also conveniently masks year-over-year EPS comparability at exactly the moment the 45X benefit peaks.
Credibility Score (1–10)
— Trending down 3-yr trend
Credibility: 5.5 / 10. Management is unusually transparent about what is happening — disclosures of warranty issues, customer defaults, and CuRe slippage have been timely and detailed. Where credibility erodes is on forward-looking framing: the original FY25 guide, the multi-year CuRe timeline, and the "sold out through 2026" story all aged worse than the steady tone of the calls would have suggested. They are honest about the past and persistently rosy about the next twelve months.
6. What the Story Is Now
The current First Solar story is narrower and more dependent than the 2023 version. The bull case has compressed to four pillars: (1) a U.S. manufacturing footprint built at the right time, (2) eligibility for the Section 45X credit through the OBBBA window, (3) FEOC enforcement that effectively disqualifies most Chinese-controlled competitors, and (4) the TOPCon patent portfolio creating a litigation overhang on every crystalline-silicon competitor. What has been de-risked: U.S. capacity is largely built (Alabama, Louisiana online, South Carolina finishing line in flight), the IRA's 45X credit is enshrined in tax law and being monetized at scale ($1.4B sold in 2025), and the CdTe technology moat is intact. What still looks stretched: the 50.1 GW backlog still includes meaningful international product whose viability depends on tariff outcomes; the 2026 net-sales guide is below 2025, which is unusual for a company claiming the AI-driven power demand story; CuRe revenue uplift is now mostly a 2027–2028 story; and Southeast Asian factories are running at ~20% utilization with no clear path to recovery.
What the reader should believe: the company is a real beneficiary of the U.S. policy regime and has the scale, technology, and balance sheet to monetize it. What the reader should discount: the "tailwinds" framing in 2026 prepared remarks (FEOC, Section 232, AD/CVD all sound great until they don't materialize on the predicted timeline), and any guidance midpoint, which has run high three of four years. The biggest single upside catalyst is FEOC clarity favorable to non-Chinese suppliers; the biggest single downside catalyst is another large customer (post-BP) walking from a multi-gigawatt framework. The 2026 EBITDA-only guide is the clearest signal that management itself expects a noisier year than the headline numbers will suggest.