T1 Energy

A domestic solar manufacturer betting on the 45X credit

T1 Energy is a US-listed solar company running module production at G1_Dallas (3.1-4.2 GW guided for 2026) and building a solar cell facility at G2_Austin (first cell production targeted Q4 2026). On paper the thesis fits a real theme: the One Big Beautiful Bill's Section 45X advanced-manufacturing tax credit pays domestic producers per watt, the US wants a non-Chinese solar supply chain, and AI-data-center power demand is pulling forward US solar deployment. If the company is what it claims, 45X turns a thin-margin commodity business into a subsidized, policy-protected one with a domestic moat.

Q1 2026 looked like validation on that thesis. Revenue of $177.65M dwarfed the $53.45M from a year earlier and beat consensus by ~60%. Gross margin improved to 17%, adjusted EBITDA hit $9.1M, and net income from continuing operations was a positive $3.9M (the first profitable continuing-ops print). G2_Austin construction is reportedly on schedule with concrete work in April and first steel before end of May. Roth Capital, the sell-side defender of the name, frames TE as FEOC-compliant, fully transparent, and a model for the Trump administration's vision of US tech transfer.

The institutional positioning matters. 170 institutional investors opened positions in Q1 2026. Situational Awareness LP, Aschenbrenner's fund focused on the AI build-out and run by a credible thinker on AI capital expenditure, disclosed a 10-million-share, ~$44M stake. That is not retail momentum; that is named, sophisticated, public capital underwriting the bull case.

The build itself is real and progressing. The question is not whether T1 is building factories; it is whether the reported economics that justify the equity are accurate, and whether the build gets funded without heavy dilution.

The entire bull case depends on one contested number

This is the part the discipline forces into view, and it cannot be hand-waved. Fuzzy Panda's short report, published in May, alleges that T1 aggressively booked $41.4M of unearned Q1 2026 45X tax credits, and that without those credits, estimated future operating margins fall from positive 6% to negative 31%. The report also alleges Chinese links via Trina Solar that may compromise FEOC compliance and 45X eligibility, that G2_Austin is behind schedule, and that the IP transfer to Singapore-based Evervolt does not resolve the China-content question.

At ~$6.85, the market is implicitly pricing roughly a 65-70% probability of the bull resolution. Whether that is the right probability is the trade. Until the DOJ/SEC inquiry resolves, no single fair value is honest.

These allegations are not market chatter. The US Department of Justice has issued subpoenas. The SEC is demanding documents tied to share sales by a manager and a board member. A patent suit with First Solar is open. Roth Capital has publicly defended T1, calling the short report misleading; T1 itself has emphasized FEOC compliance and continued G2_Austin progress. There is, as yet, no specific point-by-point public rebuttal of the 45X accounting allegations. When the single number that makes the thesis work is the exact number under federal investigation, the thesis is unsizeable until it resolves either way.

And there is a separate, structural pressure on the equity: a $225M financing gap on G2_Austin Phase 1, with management proposing to double authorized common shares from 500M to 1B at the upcoming annual meeting. Whatever the accounting truth, the build itself needs to be funded, and the funding will likely include equity dilution on top of the $160M convertible already raised.

A real revenue print, a contested margin

Revenue is real and growing fast. The contested question is what flows through to the bottom line. With the booked 45X credits, Q1 net income from continuing operations was $3.9M (the headline beat) and forward margins land in low-single digits positive. Without them, on Fuzzy Panda's math, those margins invert by nearly forty points. A reverse-DCF on a name with this much accounting uncertainty is not meaningful; there are effectively two different companies depending on which set of figures is correct, and the market is currently paying for the optimistic one.

Q1'26 figures from the 12 May 2026 release and 10-Q. Forward columns illustrative two-scenario, contingent on credit resolution. The bear column reflects Fuzzy Panda's claimed margin inversion if 45X credits are restated.

The estimated future operating margin swings ~37 points depending on whether the contested credits are earned. This single fork is why a single fair value is dishonest for TE and why the position is gated until resolution.

A genuinely two-sided book, on the page and on the tape

Unlike Nokia (Street uniformly bullish after the rate-rate) or USAR (Street firmly long with no sellers), TE has the most divided coverage of the three names. The disagreement runs across sell-side, hedge funds, and short sellers in roughly equal weight:

Two things to read from this table together. First, the institutional flow is meaningfully one-sided long; that is not a stock the market is collectively shunning. Second, the bears are not retail noise; Fuzzy Panda has a credible publication record, and federal regulators do not subpoena on hunches. Trading at ~$6.85 (down from a $9.92 peak), TE is below Needham's recently-cut $8 target but well above the bear-case fair value of $1-3. Annualized 30-day volatility above 130% says the market is pricing genuine uncertainty in both directions, not consensus.

Prototype Skyway Coverage brief · facts current to 27 May 2026 (Q1'26 results, post-Q1 institutional filings, Fuzzy Panda report, DOJ/SEC inquiry status, Roth defense all sourced). Forward estimates and fair-value derivation are explicitly bifurcated pending regulatory resolution. Not investment advice. The brain advises, the trader decides.