Fermi Inc stock (NASDAQ: FRMI; LSE: FRMI) is ending Christmas Eve with a familiar theme: extreme volatility, headline risk, and a market trying to price a huge AI-infrastructure ambition against very real execution and financing uncertainty.
On Wednesday, December 24, 2025, FRMI stock set a fresh 52-week low—trading down to $7.83 and last noted around $8.04 on roughly 1.18 million shares (at the time of that update). Market data cited the move as occurring far below the stock’s 50-day moving average near $18.77, underscoring just how violent the post-IPO unwind has been. [1]
What’s driving it is not subtle: investors are still digesting the collapse of a key construction-funding agreement tied to Fermi’s flagship development—Project Matador—plus new legal “investigation” headlines that often follow sharp drawdowns in newly public stocks. [2]
What happened: the $150 million AICA termination that reset the stock
The central datapoint is in Fermi’s Form 8-K filed December 12, 2025.
According to the filing, Fermi signed a non-binding letter of intent (LOI) on September 19, 2025 with an “investment grade-rated” prospective tenant (labeled the “First Tenant”) for a portion of the Project Matador site. In November 2025, Fermi and that tenant entered an Advance in Aid of Construction Agreement (AICA) under which the tenant could advance up to $150 million to fund certain construction costs—but Fermi had not drawn any funds under the AICA. The LOI’s exclusivity period expired at midnight on December 9, 2025. [3]
Then came the break: Fermi disclosed that on December 11, 2025, the First Tenant notified the company it was terminating the AICA, while the parties “continue to negotiate” a lease under the LOI framework. Fermi also said it had started discussions with other potential tenants for power delivery in 2026 and reiterated confidence in its expected power-delivery schedule, citing robust demand for behind-the-meter AI power. [4]
That one change—losing a potential $150 million early funding source tied to the first major tenant—was enough to puncture the story’s “inevitable momentum” narrative.
How markets reacted: a post-IPO shockwave that’s still reverberating
When the AICA termination became public, Fermi’s stock suffered its worst day on record. Reuters reported shares fell roughly 34% on December 12 after the company said the prospective tenant terminated the construction funding deal. [5]
Other coverage framed it as a major hit to a newly listed, high-profile AI-power infrastructure play. MarketWatch emphasized that Fermi’s REIT-like model depends on tenants and that the withdrawn funding commitment put pressure on credibility and timelines, with the stock falling sharply and trading well below its $21 IPO price. [6]
Barron’s similarly described the move as the stock’s worst day since listing, tied directly to the termination of the $150 million AICA (while noting Fermi said lease talks were ongoing). [7]
Financial Times added an extra layer of color: analysts interviewed there suggested the breakdown may have been driven by pricing negotiations, and the company’s leadership signaled it didn’t want to compromise pricing in a way that could weaken future deals. FT also noted the scale of the vision—an AI-focused energy and data campus potentially costing tens of billions—which makes early tenant commitments disproportionately important. [8]
The Amazon question: Business Insider says yes, Reuters says Fermi says “no”
In this kind of story, the identity of the “first tenant” becomes the whole plot.
Business Insider reported that Amazon was the prospective tenant tied to the AICA withdrawal and said discussions remained “constructive,” portraying the termination as connected to an exclusivity period ending rather than a total collapse. [9]
But Reuters subsequently reported that Fermi denied Business Insider’s claims—including the attribution of specific comments to CEO Toby Neugebauer—and said Amazon did not respond to requests for comment. [10]
For FRMI stock, this matters because the bull case is heavily dependent on landing one or more “hyperscaler” tenants (think the largest cloud/AI players) at massive scale. The bear case is that tenant wins are harder, slower, and more expensive to secure than the market initially priced in—and that “almost a deal” is not a deal.
What’s new today, Dec. 24: 52-week low prints and legal-investigation headlines
Two “today” items are shaping the December 24 news cycle around Fermi Inc stock:
- A new 52-week low
MarketBeat reported FRMI hit $7.83 intraday and traded around $8.04, calling out the technical damage (far below the 50-day average) and highlighting that the drawdown remains the dominant factor in sentiment. [11] - More law-firm investigation announcements
Robbins Geller and the Schall Law Firm both published December 24 notices saying they are “investigating” potential securities-law claims related to the sharp decline and disclosures. [12]
A reality check that belongs in any responsible market coverage: these press releases are common after big drops—especially after IPOs—and are not, by themselves, proof of wrongdoing. They can, however, add to uncertainty and keep negative headlines circulating.
The underlying bet: what Fermi is trying to build at Project Matador
Fermi Inc (doing business as Fermi America) is pitching itself as a next-generation AI infrastructure developer—combining power generation and data center campus development at enormous scale.
In its Q3 2025 shareholder letter, the company described Project Matador as a behind-the-meter campus planned to deliver up to 11 gigawatts of power, with a target of 1.1 gigawatts for first power delivery in 2026 and a site spanning over 5,200 acres under a 99-year lease with the Texas Tech University System. [13]
Fermi’s IPO also drew attention because of who was involved and the “AI power scarcity” thesis. Reuters reported the company raised $682.5 million in its IPO by selling 32.5 million shares at $21, valuing the firm at $12.46 billion at the offering price, and listing on Nasdaq and the LSE under FRMI. [14]
In its shareholder letter, Fermi said the IPO generated about $731.4 million in net proceeds after underwriting discounts and offering expenses (and described a larger gross figure including the over-allotment). [15]
The “good news” pile investors haven’t forgotten: power agreements and infrastructure steps
Even after the tenant funding shock, Fermi has pointed to tangible infrastructure progress—particularly around near-term power access.
Xcel/SPS electric service agreement (up to 200 MW; 86 MW starting Jan. 2026)
On December 5, Fermi America announced a signed Electric Service Agreement with Southwestern Public Service Company (an Xcel Energy subsidiary) to provide up to 200 MW for Project Matador, with 86 MW expected to begin delivery in January 2026 and ramp over time. [16]
This is meaningful because markets are increasingly distinguishing between AI-power projects that are “PowerPoint plausible” and projects with real interconnects, delivery schedules, and counterparties.
“$4+ billion project financing” discussions for the first tenant (Nov. 25)
In a November 25 press release, Fermi said it was in discussions for $4+ billion in project financing for its first tenant, describing the financing as potentially non-recourse through a special purpose vehicle. [17]
Read next to the December 12 AICA termination filing, that earlier statement is part of why investors are now laser-focused on whether the tenant relationship ultimately converts into a definitive, financeable lease—or whether Fermi must pivot to other tenants and funding structures.
Water-saving cooling agreement (Dec. 1)
On December 1, Fermi announced a non-binding MOU with MVM EGI to develop hybrid dry–wet cooling systems aimed at reducing water use for the campus, with the first cooling tower’s construction scheduled to begin in January 2026 and a longer-term buildout timeline extending into the 2030s. [18]
This kind of announcement won’t move the stock like an anchor-tenant lease, but it’s part of the broader “execution breadcrumbs” investors watch in megaproject stories.
Financial snapshot: big losses, but also a large cash war chest post-IPO
Fermi is early-stage, and its reported results reflect that.
In its Q3 2025 shareholder letter, Fermi reported (from inception through September 30, 2025) a net loss of $353.2 million (about -$0.95 per share) and cash and cash equivalents of $183.0 million at quarter-end (with both restricted and unrestricted components). The company also broke out large non-cash items contributing to the loss, including fair value adjustments and a sizeable charitable contribution of equity units, among others. [19]
For the quarter ended September 30, 2025 specifically, the shareholder letter showed a net loss per share of -$0.84. [20]
Translation: the income statement isn’t the story yet. The story is (1) cash runway and financing capacity, and (2) whether Fermi can turn its infrastructure build into long-duration contracted tenant revenue before markets lose patience.
Forecasts and analyst outlook: still “Buy”-leaning, but targets are being tested
As of December 24, MarketBeat summarized Street sentiment as still broadly positive—an average “Buy” rating and a consensus price target around $31.56, with some targets historically as high as the upper $30s. MarketBeat also listed earlier coverage initiations (e.g., UBS with a $30 target, Mizuho with $27, Redburn with $31, and Evercore with an Outperform and a target that was previously higher). [21]
But the December drawdown is forcing revisions. Investing.com reported that Evercore ISI cut its Fermi price target to $20 from $37 while maintaining an Outperform rating—an example of analysts trying to balance long-term thesis support with near-term damage control. [22]
The most defensible way to read “FRMI stock forecasts” right now is not as a precise map to fair value, but as a sentiment indicator: analysts largely still believe in the macro demand story (AI needs power), yet the market is demanding proof in the form of signed tenants and financeable structures.
The macro backdrop: AI power demand is real, but the market is getting pickier
Fermi is operating inside a larger 2025 trade: “AI requires electricity, so anything adjacent to electricity wins.”
The Wall Street Journal ran a Christmas Eve cautionary note on the broader energy trade, arguing that 2025’s widespread gains across energy-related segments can’t be sustained indefinitely without concrete proof—contracts, backlogs, and execution—especially for speculative, low- or no-revenue names. [23]
That framing fits FRMI stock almost too well: the upside case is enormous if Fermi becomes a real “AI power landlord.” The downside case is equally large if the project slips, costs rise, or tenant conversions disappoint.
Key catalysts to watch next for Fermi Inc stock
Going into early 2026, the next major potential catalysts for FRMI stock are fairly clear:
- Definitive tenant lease announcements (either with the original “First Tenant” or a replacement) and clarity on whether the AICA termination was a negotiating tactic, a real rupture, or simply a restructure. [24]
- Financing updates, especially anything that substantiates the prior “$4+ billion project financing” discussions with specifics (terms, counterparties, structure). [25]
- Power delivery milestones—including whether the 86 MW starting January 2026 proceeds on schedule under the SPS/Xcel agreement. [26]
- Conference disclosures and investor events: Market calendars show an Evercore conference appearance slated for January 7, 2026, which could become a checkpoint for new detail. [27]
- Earnings timing: the company hasn’t posted a confirmed upcoming earnings date on some calendars, and Nasdaq’s page reflects that earnings-date data may be unavailable or estimated—so treat “next earnings date” references as tentative until Fermi confirms. [28]
Bottom line
Fermi Inc stock is behaving like what it is: a high-conviction, high-volatility bet on a single giant idea—build an AI-focused, behind-the-meter power-and-data campus big enough to matter to hyperscalers—colliding with the market’s insistence on near-term proof.
As of December 24, 2025, the debate is no longer about whether AI needs electricity. It’s about whether Fermi can turn ambitious engineering, early infrastructure agreements, and LOIs into durable, financeable tenant contracts quickly enough to justify its post-IPO valuation swings. [29]
References
1. www.marketbeat.com, 2. www.sec.gov, 3. www.sec.gov, 4. www.sec.gov, 5. www.reuters.com, 6. www.marketwatch.com, 7. www.barrons.com, 8. www.ft.com, 9. www.businessinsider.com, 10. www.reuters.com, 11. www.marketbeat.com, 12. www.prnewswire.com, 13. www.sec.gov, 14. www.reuters.com, 15. www.sec.gov, 16. www.prnewswire.com, 17. www.prnewswire.com, 18. www.prnewswire.com, 19. www.sec.gov, 20. www.sec.gov, 21. www.marketbeat.com, 22. www.investing.com, 23. www.wsj.com, 24. www.sec.gov, 25. www.prnewswire.com, 26. www.prnewswire.com, 27. www.marketscreener.com, 28. www.nasdaq.com, 29. www.sec.gov
