Fermi Inc. (FRMI) Earnings Call Transcript & Summary

Thanks, operator, and good morning to all of you. It's a pleasure to welcome everyone to Fermi America's earnings call for the fourth quarter and the full year 2025. With me on the line today are Toby Neugebauer, who serves as our Co-Founder and CEO, and Miles Everson, our CFO. Ahead of getting started, please note that forward-looking statements will be part of this call, as defined under the Private Securities Litigation Reform Act of 1995. Management's present expectations form the basis for those statements, and they carry risks and uncertainties that might lead actual results to differ materially. For a thorough look at those risks, please see our annual report on Form 10-K covering the fiscal year that ended December 31, 2025, which we will file with the SEC later on today. The structure for today's call involves 2 parts. To kick things off, Toby and Miles will provide an operational update regarding Project Matador. After that, Miles will take you through our financial results. Following that, we'll open the line for questions. So with that, I'll turn things over to Toby.

Hello, everyone, and thanks for making time to be with us today. On the most recent Saturday, we held a Board meeting, and during it we went through everything the team has achieved across the last 100 days; across 30 years in business, this was the first occasion where I saw applause happen more than once. From where I stand, the recognition those team achievements drew was fully earned. Naturally, Fermi received the market's message clearly and unmistakably: bring in a tenant. The hard truth, though, is that securing a tenant is the simpler piece. What you as shareholders need Fermi to accomplish is winning the confidence of investment-grade counterparties, along with the investors whose financing backs construction projects costing multiple billions of dollars per gigawatt; delivering that demands excellence stretching from engineering all the way to accounting. Here at Fermi, what we're building is a private community, run on a private grid, so that we can lead in supplying power to artificial intelligence that will define tomorrow. Yet everything begins and ends with our tenants placing their trust in us — trust with their business and, every bit as crucially, trust with their balance sheet — to carry out this enormous endeavor in conditions where plenty of others are falling short. At Fermi, our view is that the finest route to earning that trust is action, is execution. I ask — I nearly beg — each of you to visit the site. Whenever tenants arrive at our site, the scale and the pace of what we're doing leaves them astonished: 450 million cubic feet of gas pipeline brought in, 10 million gallons of water pipeline brought in, grid connection brought in, a substation built for an 800-megawatt capacity, 60% to 70% finished, and foundations for our gensets either done or nearly done. What truly converted our shoppers into buyers, however, was the air permit. The moment we secured that air permit was the moment our customers' C-suites became extremely, extremely serious about purchasing. As you're aware by now, we hold the 6 gigawatt air permit. On Friday, we submitted a filing for an additional 5, which we are eligible for and which I hold a strong expectation we will execute on. Before anything else, though, I want every one of you to grasp what Fermi expects: tenants. To get the most out of our power gensets, we require multiple tenants. My thinking is that multiple tenants are necessary, since achieving proper efficiency from what we're building with this private grid depends on having diversity of demand. Next, the Board feels considerable concern about any disclosure that could affect negotiations on transactions involving multiple parties and sums in the tens of billions of dollars. So even as we sign new LOIs, we're operating in a mode of coffee as for closers; no coffee gets served until a close is fully complete; it's evident that problems have existed with the stock. And since so many of my friends, my family, and all of you placed your capital in my hands, making me a steward of it, I cannot say strongly enough how gravely I regard that responsibility. At the same time, I can't get caught up in daily ups and downs; what we're creating is a significant business addressing an essential requirement for our clients — safeguarding consumers while serving our nation. At this point I'll pass things over to our Chief Financial Officer, Miles Everson.

Thank you, Toby. We're filing our initial Form 10-K as a listed company, and it covers the stretch starting with our founding on January 10, 2025, and running through December 31, 2025 — roughly 11.5 months. During that window, we went from being newly formed to completing an IPO, and we got the first stage of Project Matador essentially finished. My preference is to describe the numbers the same way our internal management reporting treats them. At this point in our development, a standard income statement can't really show the full economics of what we're building; we haven't begun generating revenue and we're deep in construction. Yes, our GAAP net loss is large, but the vast majority of it doesn't involve cash. What tells the real story is the balance sheet along with the cash flow statement — in particular, the fact that close to $570 million in capital from investors has gone into the physical infrastructure at Matador. Now let's go through the balance sheet. At December 31, 2025, we had total assets of about $1.4 billion. Property, plant and equipment came to $935 million, and practically all of that sits in construction in progress, since nothing has been put into service yet. Cash and cash equivalents stood at $409 million when the year closed; on the other side, accounts payable and accrued liabilities reached $177 million, which mirrors how quickly we're building and how active our vendors have been. Stockholders' equity in total was $1.1 billion. Coming into March 2026, we had roughly 630 million common shares outstanding. Turning to the income statement and what happened operationally, the net loss across the full year came to $486 million. The key point is that about $445 million of it was noncash. G&A expenses added up to $178 million, and $133 million of that figure was noncash share-based compensation connected to equity incentive arrangements put in place when we were formed and around the time of the IPO. The cash portion of G&A came to roughly $45 million, which breaks down as $12 million of personnel costs for a small team of about 35 people, $22 million for professional services, and $11 million of other corporate items like recruiting, travel and marketing. Other expenses, net, were $312 million and were almost entirely noncash. The biggest pieces: $174 million tied to a charitable contribution of Class B units made before the IPO, $61 million of fair value losses on Series B convertible notes, $46 million of losses on embedded derivatives linked to preferred unit financing, and lastly, $24 million connected to preferred unit issuances. Looking at it from a cash standpoint, operating cash use over the year was $34 million. That's our actual operating cash burn as we formed the company, finished the IPO, locked in a 99-year ground lease, built out the organization, and pushed Phase 0 construction forward — we see that as a solid showing of capital discipline. As for investing activities, which sit at the heart of our financial narrative and represent how we put our investors' money to work, net cash used there was $570 million, and essentially all of it went straight into property, plant and equipment at Project Matador, booked as construction in progress. Natural gas power generation took the majority of that capital — over 50% — covering turbine purchases for Siemens F-Class and SGT-800 units plus GE6B fleets, and also mobile generation and balance of plant equipment. Data center infrastructure, substation and electrical interconnection, general construction, land and water development and early-stage nuclear predevelopment made up what was left. Turning now to financing activities: roughly $1 billion came in as cash provided by financing activities. Within that, $746 million was net proceeds from our IPO, $108 million came from preferred units, $100 million was a Macquarie term loan, $76 million came from Series A convertible notes and $26 million was from seed convertible notes. After year-end, we closed 3 equipment financing facilities: a $500 million MUFG nonrecourse turbine warehouse backing Siemens F-Class procurement, which additionally refinanced in full a Macquarie term loan, and a $120 million facility with Keystone National Group that can expand to $220 million for high-voltage equipment such as transformers and switchgear. Then the third, signed this week, for $165 million with Yellowstone, financing more Siemens SGT-800s. Together with the cash we already hold, those facilities provide liquidity sufficient to meet our financial obligations for no less than 12 months, though we are taking a deliberate approach to what comes next. At Project Matador, the following phase of capital deployment will be tied to 2 milestones: first, executing a definitive tenant agreement; second, closing project financing. Those are the gates. We will not commit significant capital toward the next construction phase until both exist. That is how shareholder value gets delivered. Both work streams are moving forward in parallel; on the tenant side, potential tenants are competing for initial power. Active negotiations are underway with multiple counterparties. As of today, though, no definitive lease agreement has been executed. Tenant revenues should begin in 2027, yet even then they will fall short of covering our full operating capital requirements until Matador is fully built out and running at scale; on the financing side, we are actively talking with multiple lenders and moving technical diligence forward now, so we can act quickly once definitive lease agreements are signed. Underwriting for the project level financing rests on the future cash flows that a tenant commitment unlocks; for Project Matador, we expect Phase 0 and Phase 1 combined to surpass $3 billion in total aggregate capital deployment. Our ability to execute tenant leases, raise project level debt and bring in strategic equity where necessary determines the path forward; we believe this can be achieved. These financings, however, are not certain to occur. Should capital be unavailable in the amounts, timing or terms we require, we might be compelled to delay investments, amend purchase commitments or potentially surrender collateral to preserve liquidity; we are stating that plainly because it reflects the reality of building a multibillion-dollar infrastructure platform from a standing start. Also, we think investors ought to actually hear this from us, rather than simply come across it buried in a risk factor. Here's the key point: our liquidity is sufficient to cover what we owe. When and how we put capital to work is a deliberate, strategic decision. At Project Matador, the upcoming stage of capital deployment will be timed to align with signed definitive tenant agreements, along with the project financing tied to them that comes afterward. Additional work remains ahead, and we tackle it daily. Let me also address the REIT election. As previously stated, our intention is to elect REIT status for U.S. federal income tax purposes, starting with our short taxable year that ends December 31, 2025. In our view, this structure fits nicely with the long-duration, infrastructure-focused real estate assets currently under development by us. Because of the expected level of noncash depreciation, we don't foresee producing material REIT taxable income in the near term. So we don't expect to pay dividends until taxable income at some point requires it. Last, for those pre-IPO investors who had a lockup agreement, I'd like to point out that it expires today. Thank you.

Operator, we're ready for questions.

Congratulations on everything you've accomplished with the site. There are 2 things I'd like to ask about in the near term. First, when you're trying to reach a definitive agreement, what are the main topics being discussed with the prospective tenants you're negotiating with? Second, and more specifically, regarding the near-term energization milestones — with the SGT-800 frames that have arrived in Houston, and the equipment now in your possession, how is the schedule shaping up for getting those assets deployed, and how might that tie into the talks you're having with prospects?

Thank you, Paul. What our tenants raise as the #1 issue, the #2 issue and the #3 issue is that they want every bit of our power, and they want it permanently; however, our model only functions if we have several tenants, which lets you handle the variations in loads; that's really the crux of it. And so from — once they arrive at the site, they're astonished by it and they understand this is somewhere you can produce a substantial amount of power, and they want all of it; as for the SGT-800s, we've been fairly fortunate. So when — and I'll return to the tenant topic. The units arrived, as you know, in Houston. We held them in a free trade zone, hoping perhaps we could catch a break while we waited on our environmental permit. And literally, the moment the Supreme Court issued a ruling on the tariffs, we brought them into the country. We likely saved ourselves $27 million to $30 million; and we make — I'll get the pictures posted. You can see the SGT-800 foundations are set to be poured; but on the tenant side, since I know that's where everyone's attention is. What we're able to share is that we're now in the contracting phase with several new potential tenants. But it's important for everyone to grasp that these transactions are complicated. They involve multiple parties. And Macquarie, as you know better than anyone, these involve billions of dollars.

Yeah, hi Paul — Miles here. I'd just like to tack something on — and this has been our position the whole time — another point, and it isn't some kind of friction, but one item we won't budge on is our insistence on investment-grade wraps. That's the reason Toby mentioned a bit earlier that what this really comes down to is firms asking themselves whether they'll commit their balance sheet. And by firms, I mean the investment-grade firms doing the wrapping on these projects. So the next point I'd tack on is that it isn't merely about our own actions and the things within our control — the end users of the output, or the development partners, also have to examine this and ask whether they can hold up their end of things. Put differently, all of their MEP plus the racks, and so forth. So we're now doing, I'd say, more reverse due diligence than I'd anticipated needing to do, to confirm they can get their equipment ready in time to consume our power, since our power arrives sooner than what most folks can manage for the remaining pieces. That's ranked among the larger surprises — if you'd told me back in the fall, the concern was whether, well, persuading people that we'd be able to deliver the power. Now we're the ones who want reassurance that they've got the MEP, since the power is indeed there and we've got to put it to use.

Miles, if I may squeeze in just one additional question, given that the conversation about demand appears quite strong. On the subject of pricing — could you offer any insight into where pricing is headed compared with what you anticipated at the point you initially began weighing, in a sense, the financial strategy for that initial gigawatt of power?

We remain where we were; and although we'll aim to submit bids at a higher level, we are certainly prepared to state that we remain where we were.

Congratulations on wrapping up your entire fiscal year. What I'd like to do, if possible, is go a bit deeper into those conversations with tenants. I'm curious whether the hurdles differ when Fermi is the landlord compared with your prospective tenants, and how those hurdles might end up pushing back when deals get signed. I believe you mentioned a 12-month figure in your remarks — across the coming 12 months. So I'd like to understand more clearly what the sticking points look like from each side; and thinking about timing, ought we to plan on 12 months? Or might it happen faster?

Yes, I realize the way it lands < >. As soon as they arrive at the site, the reality is they're after the power, and they're attempting to secure every bit of our power at a price set today, since I believe our tenants genuinely understand how scarce the gensets are and that their energy costs are going to rise; so quite reasonably, their focus is on securing the greatest amount of power they can at today's pricing; that's truly the whole story. And it's — as I mentioned, this process has us in the contracting stage right now. On the question of giving guidance about timing, the Board, the coaching, I have an excellent — Fermi's Board is truly excellent < > examine this. Their view is that whenever we give guidance about timing or expectations, that alters how the negotiations unfold, to Fermi's detriment. So on Saturday the Board was quite resolute with me that timing is not something we'll talk about, because doing so shifts the dynamics of — these aren't transactions worth $1 billion. These are transactions worth multiple billions of dollars, and my aim is simply to keep those dynamics as level as I can.

All right, and there are only a couple of items I'd like cleared up. First, I believe the letter to shareholders brought up term sheets along with a number of agreements. I simply want to confirm, first, whether an actual LOI exists with any among those 5, 6, 7 tenants you're in negotiations with, or whether that is more or less the following stage. And second, could you clarify any of the short-term financings, such as the one from MUFG? Does any of these financings carry a condition requiring that a lease be signed by XYZ by a certain date? I believe what we've said — the thing we're at ease with — is that executing new LOIs will be part of our ordinary business operations, and we will not be offering comments about them. After that, I think one lesson we've picked up so far is that we don't wish to alter the negotiating dynamics. Each of these financings stands on its own. And I'm certain we've disclosed it. Miles, I just would rather not speak about the...

Yes. We have no tenant signing covenants, so to speak, Vikram, attached to our financing arrangements; and the other point to keep in mind is that these financings are nonrecourse to parent, which I think matters a great deal when you consider the overall public company. And then what has not shifted is that we do have an agreement with Texas Tech under which there will be a tenant by the end of 2026. That stays the same, and we're working collaboratively with them to move that forward. So that is probably the most important thing at the moment — to understand that we're still fully engaged, and we feel really good about where we stand, frankly.

Right, it's just a 200-megawatt tenant. I'm not trying to — I don't mean to talk it down, but that ought to be < >. That falls short of what we expect. Here's how we see it. There isn't a tenant asking us for 200 megawatts. That isn't the issue on our end.

They're asking for additional capacity.

Right. Suppose we were forced into a situation where a single tenant could only be 200 megawatts — that would create an issue on our end. The real challenge is that what they're after is gigawatts.

Miles, you brought this up before, but at this stage, what further work at the site do you intend to do before we get a first tenant lease signed and you lock in project financing? Can you tell us anything about more specific CapEx spending or cash burn you expect during this year?

Yeah; so when it comes to development, we'll be very disciplined about aligning it with both the signing of a project financing and our tenant leases. The — but as far as how far we'll go — and look, these things shift; this is a long-term project, not one that lasts 30 days. But here's what we'll do: get the site ready so it can receive our power generation equipment. It's mostly there today. A bit of incremental work remains to be done, which would let us put those generation assets into service the moment we see tenant agreements that line up with when that installation happens.

Right. I'd say what we were getting at is that Fermi isn't doubtful, exactly, but we really need to confirm that when we develop, it lines up with the MEP our tenants are able to secure. And to your question about whether our stance as a business has shifted as we've grown — I'd put it this way: a year ago, our emphasis was on giving people room to carry out strong due diligence on us. Since then we've changed direction and brought people on board; we brought in an excellent individual from Meta who assists us in evaluating how quickly our prospective tenants actually execute.

Sorry in advance if you hear anything in the background — I'm at an airport right now. So thinking about what your tenants require in terms of power, and the mix of tenants, plus the sequencing of when their data centers get built out and when they actually need that power — what does that look like? When you're in conversations with customers, how far ahead are they planning around locking in power, measured against the point where the data center is at full operation?

First off, that answer changes from one tenant to the next. And it's the #1 — well, not the #1, but it lands in the top 3 items of diligence for us, or rather the way we rank tenants, is clearly, can you be financed, meaning do you carry investment grade, that's number one; but the #2 item is that we, in contrast to most companies out there, hold a tremendous amount of power generation that we're eager to deploy. And so — that is — there isn't a single answer here. But if you consider how we operate the business and how we rank the folks we'd sign contracts with, that's likely our #2 item. How quickly we're not — we're able to have the power given all the site work we've completed up to now. We're able to have the power, which we now understand probably sooner than some of them are able to have the MEP. So that turns into how we rank who we sign with.

And for the follow-up, it's sort of — just returning to that first possible tenant and those talks that got called off, or at least perhaps called off but pushed back, when exactly did they require power? Because I'm just kind of wondering, if power is lacking, how are they getting power if they're not getting it from you?

Looking back, and again, I don't think — I hope that first tenant turns out to be a tenant. So that's just what I want to get across. When we consider the timing of their power needs, it really made us realize that we must truly ensure these tenants have the MEP. In my view, that's a larger bottleneck than we had initially expected.

Miles, what I mean is, you had a hand in that...

Right. So, at first their plan was to have electricity available for themselves to sell and generate income from in 2027, all right? But your broader question has 2 pieces, I think — the first being the timing of when off-takers require power. If you examine the portfolios they've mapped out, the majority fall in '27 and '28, which is when consumption of that power becomes necessary. That said — and this matters — you can find it reported in the papers. Certain other sites simply can't supply the power they promised to these parties. As a result, we anticipate, and have already observed, some possible reallocations regarding where they'll obtain power they believed was already secured but in fact is not.

Our position is unusual, since we hold considerable power, and that fact is acknowledged globally.

I wanted to ask about one thing. Given that there are several new LOIs moving forward along with the first one, I was hoping to get a sense of whether the possible range of those tenants has grown — as in, are different kinds of tenants involved? Or does it remain roughly those hyperscalers? Clearly investment grade, but I'm just trying to determine whether those horizons have widened somewhat.

The single item I'd call notable, I believe, is meaningful involvement from chip makers — not some of them directly, not directly — I think they're becoming quite worried that those chips hold value only because of the power supporting them; so regarding scope, that's the sole shift I would point to. Miles?

I'd put it the same way. The people making the chips are asking, in a much more hands-on way, where will the electricity actually be supplied from? And honestly, what they really care about is where all the usage of their chips is going to be supplied from.

My take is that everyone's overlooking those markets—the top chip manufacturers are starting to see that they've fallen behind quite a few businesses. That shifts how things work, so yes, it does widen things out. On our end, the crucial point—and you're as aware of this as anyone—is that we truly require a varied mix of load to get the most efficiency out of our gensets. So what I believe you're watching us do involves somewhat more math—not somewhat, considerably more math. Having a broader, more varied load base serves us better for getting the most efficiency from this private grid we're in the process of constructing.