Back to News
Earnings AnalysisFRMI

FRMI FY2025: $935M Deployed, Zero Binding Tenants, Stock Down 76% From IPO

Matt RuncheySHORELINE, WA — March 30, 2026 · 12:30 PM PDT6 min

Fermi Inc. filed its first 10-K annual report, confirming what the committee's 6-lens analysis anticipated: massive physical infrastructure progress colliding with a complete absence of binding commercial commitments. The company has deployed $935M in construction-in-progress, secured a 6 GW air permit, confirmed the AP1000 Westinghouse reactor design, and closed $500M in MUFG financing. But the only financial commitment from a tenant — a $150M AIAC agreement — was terminated in December 2025. Revenue remains $0. The stock fell 24% on earnings day to ~$5.12, extending the decline to 76% from its $21 IPO price.

HIGHER_SCRUTINY
Committee posture unchanged
$5.12
-76% from $21 IPO
-$486M
Net Loss (FY2025)
91% non-cash
$935M
Construction in Progress
Total assets: $1.4B
$409M
Cash Position
~4 years at $34M/yr burn
$0
Revenue
First revenue expected 2027

The Core Tension: Infrastructure Without Tenants

Fermi's FY2025 earnings present a genuine paradox. The infrastructure progress is among the most tangible of any pre-revenue energy company: 42 Siemens SGT-800 gas turbines secured (with 5 installed), a 6 GW air permit finalized in February 2026 with an additional 5 GW filed, and the first NRC large reactor application in 15+ years with 4 AP1000 Westinghouse reactors confirmed. The MUFG $500M facility closed post-year-end, fully refinancing the Macquarie term loan and eliminating the 2026 maturity wall that was the single largest near-term risk in our prior analysis.

But the commercial validation gap has widened, not narrowed. The $150M AIAC (tenant advance for infrastructure construction) was terminated in December 2025 — the only binding financial commitment from any tenant. CEO Neugebauer characterized prospects as being in a “contracting phase” and said the air permit was what “really turned our shoppers into buyers.” But there are no buyers yet. Zero binding leases. Zero signed commitments. The board is explicitly withholding tenant timeline guidance.

CFO Risk Disclosure
CFO David Jones stated directly: “These financings are not certain to occur... we could be forced to delay investments, amend purchase commitments, or potentially surrender collateral.” This candor is atypical for pre-revenue companies and should be weighted accordingly. The cash position of $409M provides approximately 4 years of runway at the $34M true operating burn rate — but capital commitments for Phase 0 and Phase 1 development far exceed available cash without additional financing or tenant deposits.

Signal Change: 1 Reclassified, Multiple Confidence Upgrades

Our 6-lens update produced one signal reclassification out of seven monitored signals. The remaining six confirmed at prior levels with confidence upgrades from MEDIUM to HIGH based on the substantially richer evidence base from the first 10-K:

EXPECTATIONS_PRICED — EXUBERANT → STRETCHEDImproved
ACCOUNTING_INTEGRITY — QUESTIONABLE (confirmed, confidence HIGH)Unchanged
GOVERNANCE_ALIGNMENT — MIXED (confirmed, confidence HIGH)Unchanged
FUNDING_FRAGILITY — STRAINED (confirmed, MUFG de-risks near-term)Unchanged
NARRATIVE_REALITY_GAP — DISCONNECTED (gap shifted to commercial)Unchanged
REGULATORY_EXPOSURE — EXISTENTIAL (NRC timeline multi-year)Unchanged
REVENUE_DURABILITY — FRAGILE (zero revenue, zero leases)Unchanged

The EXPECTATIONS_PRICED reclassification from EXUBERANT to STRETCHED reflects the 76% stock decline compressing market cap from ~$13B to ~$3.2B. At $5.12, Fermi trades at 2.3x book value and 1.8x total capital assembled — the market no longer implies near-certainty of the full 17 GW vision. However, $3.2B for a pre-revenue company with zero binding tenants and a terminated AIAC still requires meaningful execution milestones to justify. Notably, 8 of 8 sell-side analysts maintain Buy ratings at an average $29 target (466% upside) — a disconnect that itself functions as a Myth Meter signal.

Evidence Quality Upgraded Across the Board
The first 10-K filing is a substantially richer evidence base than the prior 10-Q. Full-year financial statements with detailed notes enabled confidence upgrades on ACCOUNTING_INTEGRITY, GOVERNANCE_ALIGNMENT, FUNDING_FRAGILITY, CAPITAL_DEPLOYMENT, and COMPETITIVE_POSITION. Evidence level on COMPETITIVE_POSITION upgraded from E1 to E2 as all competitive claims are now grounded in verifiable primary sources (AP1000 design confirmed in NRC COLA filing).

Key Developments: De-Risking vs. Commercial Void

DevelopmentImpactAssessment
MUFG $500M closedRefinanced Macquarie, eliminated 2026 maturity wallPositive
6 GW air permit finalizedGas-side regulatory bottleneck removedPositive
AP1000 reactor design confirmedNRC COLA filed for 4 reactors, pilot program selectedPositive
NRC environmental review initiatedApplicant-prepared EIS may reduce timelinePositive
$150M AIAC terminatedOnly binding tenant commitment eliminatedNegative
Zero binding tenant leasesCommercial validation gap widenedNegative
Lock-up expiry (March 30, 2026)Pre-IPO investors free to sellRisk factor
REIT 5/50: Neugebauer at 38%Must sell down, hired advisor for block saleStructural

CEO Neugebauer described a strategic shift toward reverse due diligence — Fermi is now evaluating tenant MEP (mechanical, electrical, plumbing) readiness rather than simply seeking tenants. The company hired an ex-Meta executive specifically for this purpose. Management targets 500+ MW minimum deals with 2–3 tenants for the first 2 GW. Modular MEP via factory-built solutions (Schlumberger visit noted) may compress deployment timelines. There is also “significant engagement by chip makers” — a new tenant category beyond hyperscalers.

Narrative Expanded While Evidence Gap Widened
The original analysis flagged the narrative scope at 11 GW. The 10-K expanded it to 17 GW (11 GW gas + additional nuclear capacity). Physical progress has validated the “is the project real?” question — but the core question has shifted to “will anyone sign?” The AIAC termination and zero binding leases mean the commercial validation gap has actually widened since our initial analysis, even as the infrastructure gap has narrowed.

Reading the Numbers: $486M Loss in Context

The headline $486M net loss is primarily non-cash. A $104M Q4 SBC surge is explained by IPO-triggered vesting acceleration, not ongoing compensation. The charitable donation (previously flagged in our analysis) remains opaque. The true operating cash burn is ~$34M/year, which against $409M in cash provides roughly 4 years of runway at current rates — though capital commitments for Phase 0 and Phase 1 development require substantially more than available cash.

The balance sheet tells the asset story more clearly than the income statement: $1.4B total assets anchored by $935M in CIP with $1.1B in equity. The ~628M shares outstanding at ~$5.12 produces the ~$3.2B market cap. Management saved $27–30M by clearing SGT-800 turbines through customs before the Supreme Court tariff ruling — a genuine operational execution win.

Prediction Markets: 3 Resolved

Three prediction markets resolved with this earnings report at an average Brier score of 0.1432 — solid calibration with one excellent call, one strong call, and one moderate call:

CIP impairment charge in FY2025 10-KNO (88%)Brier 0.0144
MUFG facility closing by 10-K filingYES (57%)Brier 0.1849
Reactor design disclosed in FY2025 10-KYES (52%)Brier 0.2304

The CIP impairment call was the ensemble's best (Brier 0.0144) — the models correctly assessed that management would not impair assets during an active construction program with increasing capital commitments. The MUFG closing prediction (Brier 0.1849) was solid: the ensemble assigned moderate probability and the deal closed post-year-end. The reactor design disclosure (Brier 0.2304) was directionally correct at 52% but underweighted the regulatory filing dynamics that made disclosure near-certain once the NRC COLA was submitted.

Today: Lock-Up Expiry and Structural Pressures

The pre-IPO investor lock-up expired today, March 30, 2026. This coincides with two structural selling pressures: the Neugebauer family holds 38% of shares and must sell down for REIT 5/50 compliance (no single holder may own >50%, and top 5 holders combined must exceed 50%). They have hired an advisor for an “accretive” block sale, though the mechanism for making a mandatory sell-down accretive is unclear.

Additionally, the Texas Tech ground lease imposes a 200 MW tenant + Notice to Proceed deadline by end of 2026 — only 9 months away. Failure to meet this condition could jeopardize the land rights underlying the entire Project Matador development.

New Earnings Call Dynamics
Two management disclosures stand out from the earnings call. First, CEO Neugebauer on the air permit impact: “What really turned our shoppers into buyers was the air permit.” Second, the emergence of chipmaker engagement as a new tenant category beyond hyperscalers. Both suggest commercial traction may be closer than the zero-binding-leases headline implies — but “shoppers turned buyers” without signed leases remains aspiration, not evidence.

What to Watch Next

1.First binding tenant lease: The single most important near-term catalyst. Converts narrative into contractual reality. Management targeting 500+ MW minimum deals with 2–3 tenants for the first 2 GW.
2.Texas Tech deadline (end of 2026): 200 MW tenant + Notice to Proceed required. Failure could jeopardize land rights. Only 9 months remain.
3.Post lock-up Form 4 filings: Monitor insider selling patterns over the next 30–90 days. The Neugebauer block sale mechanism and pricing may signal management confidence.
4.NRC environmental review milestones: Notice of Intent published March 20, 2026. Scoping period open. The applicant-prepared EIS pilot program may reduce the typical multi-year timeline.
5.REIT 5/50 compliance progress: Neugebauer family must sell down from 38%. Timeline and mechanism unclear. Failed compliance would threaten the REIT classification that provides Fermi's tax structure.
Sources and methodology

Primary sources: FRMI 10-K Annual Report (FY2025), filed with SEC on March 25, 2026; Fermi Inc. FY2025 Earnings Call Transcript (March 26, 2026); NRC COLA Application and Notice of Intent (March 20, 2026).

Analysis methodology: Multi-lens committee analysis using the Runchey Research AI Ensemble. Six lenses (Fugazi Filter, Stress Scanner, Moat Mapper, Regulatory Reader, Myth Meter, Black Swan Beacon) updated with earnings data. Each lens employs structured discourse between multiple AI models with persona-based orchestration.

Prediction markets: 9-model ensemble generating binary probability forecasts scored via Brier metric (0.0 = perfect, 1.0 = worst).

Full 6-lens analysis with signal table, cross-lens reinforcements, monitoring triggers, and committee discourse

FRMI Full Analysis