Stock behind Donald Trump data-center crashes

Fermi America (FRMI) just learned a hard lesson about the gap between ambition and execution. On paper, its gigantic data center campus Project Matador looks like a slam dunk investment. That was until its initial tenant terminated a $150 million Advance in Aid of Construction agreement, triggering FRMI stock to crater of nearly 50% in premarket trading.

The reaction was swift—and brutal—especially given that demand for large-scale data center capacity remains strong across the industry.

Project Matador is being pitched as something unprecedented.

An 11-gigawatt mixed-use data center campus in Texas, formally called the President Donald J. Trump Advanced Energy and Intelligence campus. Fermi describes it as the largest project of its kind currently under development.

But scale cuts both ways.

Despite the vision, Fermi has yet to finalize an anchor agreement with a major data center developer. The loss of the construction advance made that painfully visible. Investors weren’t just reacting to a single tenant walking away—they were repricing the risk that the project’s timeline and economics may be far more fragile than advertised.

Power permits are secured. Customers are not.

To its credit, Fermi has done much of the hard, unglamorous work. The company says it has permits in place for 6 gigawatts of gas-fired generation and is coordinating with federal regulators on plans to build four AP1000 nuclear reactors at the site.

That’s not trivial. Power availability is one of the biggest bottleneck facing data center development today.

Fermi says it remains on track to bring hundreds of megawatts online early next year. But investors are clearly signaling that power readiness alone isn’t enough. Without signed tenants, even the most energy-rich campus remains a concept rather than a business.

Hyperscalers move slower than developers expect.

Fermi has positioned Project Matador squarely at hyperscale customers—major technology firms with insatiable compute needs. The company says Palantir has held discussions and is expected to visit the site, and that talks are ongoing with multiple candidates, including the tenant that withdrew its advance.

Still, management expressed surprise that more companies haven’t rushed to secure space.

That surprise may itself be the problem.

Hyperscalers don’t sprint into first-of-kind megaprojects. They wait. They diligence. And they demand certainty—on timelines, costs, regulatory stability, and execution risk—before committing tens of billions of dollars.

Infrastructure ambition is colliding with political and local constraints.

Data center developers everywhere are running into the same resistance: local opposition over energy use, water consumption, and costs that resemble those of small cities.

Fermi says it has mitigated these concerns by securing new power generation and access to millions of gallons of water per day from nearby aquifers. That puts it ahead of many competitors.

But mitigation doesn’t eliminate friction. It just delays when friction shows up.

Project Matador sits at the intersection of two powerful trends: exploding demand for AI infrastructure and the growing difficulty of actually building it at scale.

Fermi has aimed for the extreme end of that spectrum. But as Oracle found out yesterday, investors are no longer willing to reward ambition alone.

Fermi (FRMI) has a Disruption Score of 0.

Click here to learn how we calculate the Disruption Score.