The Funding Path Improved and the Stock Fell 21%: Upgrading Centrus to Outperform

Results vs. Consensus

Centrus reported after the close on February 10 and held its call the following morning, February 11, at 8:30 a.m. ET. This is our second note on the company. In November we initiated at Hold and named three disclosures, any one of which would move the rating: a costed Piketon plan with a target unit cost, a definitive strategic investment agreement with a named amount, or a DOE task order of material size. The third arrived on January 5. The stock, meanwhile, fell 20.7% on the reaction session.

The quarter itself is the least interesting part of the release. Revenue landed within 0.6% of both consensus estimates we could find. EPS missed badly, and the reasons are disclosed and mechanical. What changed in the three months since our initiation is the funding architecture, and it changed in the direction we said would justify an upgrade.

Fourth-Quarter Year-over-Year Comparison

Full-Year Comparison

Segment Performance

LEU segment

Full-year revenue of $346.2M was down 1.1%, and the mix within it moved decisively. SWU revenue rose 21% or $51.9M on a 23% increase in volume sold, partially offset by a 1% decrease in average price. Uranium revenue fell 54% or $55.6M. Cost of sales fell 8% or $21.3M, with SWU costs up on volume but average unit cost down 13%. The result is segment gross profit up 18.7% to $111.5M, attributed to higher SWU volume and better SWU margin from contract and pricing mix.

Assessment: This is a genuine improvement and it corrects the picture our initiation drew from one quarter. In November the third-quarter release showed the average SWU price down 69% and a segment gross loss; the full year shows SWU volume up 23%, average price down only 1%, unit cost down 13%, and segment gross profit up nearly a fifth. Management's insistence that the annual frame is the informative one was right, and we said we would hold it to that. On the annual frame the resale business earned $111.5M of gross profit on $346.2M of revenue, a 32.2% segment gross margin, which is a better business than the third quarter suggested. What has not changed is the price exposure: a 1% decline in realized SWU price in a year when the published curve compounded at 24% annually since 2019 confirms that spot and realized remain disconnected.

Technical Solutions segment

Revenue rose 11.3% to $102.5M on a $10.5M increase from the HALEU Operation Contract, while cost of sales rose 30% or $22.0M, driven by a $22.8M increase in costs under the same contract. Segment gross profit fell 65.9% to $6.0M. The disclosed reason is unchanged from the third quarter and now dated: because Phase 2 was delayed, DOE extended it through January 31, 2026, and costs incurred after November 2024 have not yet been subject to a fee because that portion remains undefinitized.

The operational result was better than the financial one. Management said Phase 2 was completed by contractually delivering 900 kilograms of HALEU UF6 to the DOE, with well over 1 metric ton produced by the end of 2025.

Assessment: This segment took $22.8M of additional contract cost and earned $11.6M less gross profit, which is what a cost-plus contract running with an unpriced fee on a year of scope looks like. The number to hold on to is the receivable-in-kind: over a year of accumulated cost carrying no fee, subject to negotiation with a counterparty that has just selected the same subsidiary for a $900M award. That is a negotiation the company enters from a stronger position than it held a quarter ago. Judge the segment on the metric ton delivered, and treat the fee resolution as an unquantified asset rather than a loss.

Backlog

Assessment: Total backlog fell $0.1B over the quarter, which is what happens when a business delivers $146.2M of revenue and does not replace all of it. The composition is unchanged: $2.3B of the LEU figure remains contingent and $0.2B of that still awaits a definitive agreement. Management said it added roughly $300M to the contingent book during 2025 and is making progress on removing contingencies, and provided no detail on either. The one structural change is on the other side of the ledger. Contingency removal depends on securing substantial public and private investment, and $900M of public investment was just selected. That does not convert the contingent backlog on its own, and it is the first event that could.

Key Topics & Management Commentary

Overall Management Tone: Markedly more concrete than three months ago and still withholding the cost figures. This call carried a dated first cascade, a named EPC contractor, a costed public award, a capacity threshold for the cost curve, quantified hiring, the first revenue and capital guidance in our coverage, and a commitment to report trailing-twelve-month financials from next quarter. Every one of those is new. What remains withheld is the same thing as last quarter: the capital cost of the plant and the unit cost of its output, which management again declined explicitly.

1. The $900M DOE HALEU award, and why the form matters

On January 5, DOE selected American Centrifuge Operating, a Centrus subsidiary, for a $900.0M task order to expand Piketon for commercial-scale HALEU production, subject to negotiation. Management sized the potential upside and, importantly, described the instrument.

"The $900 million HALEU enrichment award has the potential to exceed $1 billion, and still needs to be finalized through negotiations. This potential funding, which comes through a procurement and involves neither debt nor equity, would serve 2 important points. First, it would be another pool of low-cost capital... And second, it is supporting the 12 metric tons of HALEU capacity ahead of a commercially-viable advanced reactor market." — Amir Vexler, President and Chief Executive Officer

The CFO added the mechanics: $170M of additional options on top of the $900M, and payment on milestones, with the final contract still being negotiated.

Assessment: This is the single most important development since our initiation and it is exactly the class of event we said would move the rating. The critical feature is the one management named: a procurement is not dilution and not leverage. Public money that arrives as payment for delivered milestones changes the funding arithmetic in a way that neither the $805M of converts nor the $533.6M of at-the-market equity did, because it does not compete with existing shareholders for the residual. Two cautions belong next to it. The award is subject to negotiation and is not signed, and milestone payments mean the company funds the work before it is paid for it, which is precisely why the $2.0B cash balance and the 2026 capital deployment guide sit side by side.

2. nth-of-a-kind cost is reachable below full capacity

The most valuable analytical disclosure was a headline item in the release and the CEO's prepared remarks, and the Q&A sharpened it. The base-case build-out, sized for the $2.3B contingent LEU backlog plus 12 metric tons of HALEU, is expected to be sufficient to reach nth-of-a-kind cost.

"The nth-of-a-kind move will be achieved way before the 3 million SWUs, as you mentioned... As you know, one of the greatest obstacles for any new entrant into the market is the nth-of-a-kind cost. And the fact that we are commencing, we're building towards a certain volume and capacity, and the fact that we are able to get through the first-of-a-kind in this short period of time, I think, is remarkably important." — Amir Vexler, President and Chief Executive Officer

Assessment: Read carefully, this lowers the capital hurdle for a competitive cost position. The questioner's own working assumption, stated in the question, was that the full 3.5 million SWU of capacity was required to get down the learning curve. If the base case clears it, the company needs less capital before its output is competitively priced, which shortens the funding path and reduces the amount of equity that must be raised at whatever price prevails. It is also the closest thing to a cost disclosure the company has given, and it is still a relative statement rather than a number. Management again declined to disclose full cost in the same answer.

3. An EPC contractor, and a dated first cascade

Two execution facts arrived that did not exist in November.

"It is with great pleasure that I'm able to announce that we just recently entered into an agreement with a best-in-class partner, Fluor, who will serve as our primary EPC in Piketon." — Amir Vexler, President and Chief Executive Officer

And on schedule: "Our first new cascade of centrifuges is expected to come online in 2029, with subsequent cascades to come thereafter."

Assessment: A named tier-one engineering, procurement and construction contractor is a real de-risking event and a real cost commitment, and it is the kind of counterparty that does not sign for a project it expects to stall. The 2029 date is the first schedule the company has published and it should be treated as a first estimate rather than a commitment: management spent much of the call discussing efforts to compress the 42-month lead time it has previously described, which is an admission that the schedule is the binding constraint and an indication that it is being worked on rather than accepted.

4. The first guidance in our coverage, and what it implies

The company guided for the first time in our coverage: 2026 revenue of $425M to $475M and total capital deployment of $350M to $500M.

"Total company revenue of $425 million to $475 million, with the midpoint of $450 million representing flat year-over-year growth. And total capital deployment of $350 million to $500 million. 2026 capital deployment will also include prepaid expenses that do not appear as CapEx, but impact free cash flow as we invest in our partners as they scale up ahead of our production." — Todd Tinelli, Chief Financial Officer

Assessment: Guidance is a genuine governance improvement and its content is the most likely reason the stock fell. Flat revenue tells a shareholder that 2026 delivers nothing new from the existing business, while capital deployment at the $425M midpoint is roughly 94% of guided revenue and roughly 22x 2025 capital expenditure. That is a company converting from a cash-generative reseller into a capital-consuming developer inside one year. The disclosure that some of the spend is prepayments to suppliers rather than capital expenditure is important and easy to miss: it will not appear in the capex line, it will appear in free cash flow, and it means reported capex will understate the cash going out of the door.

5. The equity was raised, and at what price

The at-the-market program announced in November was used immediately and heavily, and management disclosed the average price.

"In 2025, we raised gross proceeds of $533.6 million through 2 ATM programs. We raised gross proceeds of $390.4 million through our recently announced November 2025 ATM program at an average of $269.21 per share." — Todd Tinelli, Chief Financial Officer

Assessment: This is the dilution we flagged at initiation, executed. The disclosure of the average realized price is exactly the transparency the November announcement lacked, and the price is the striking part: $269.21 against a $210.16 reaction close. Whatever the market thinks of the shelf, the company sold stock 28% above where the shares now trade, which is what opportunistic issuance is supposed to look like and is a point in management's favour rather than against it. The offsetting fact is permanent: full-year diluted average shares rose 21.7% and the fourth quarter's rose 34.3%, and per-share earnings fell while net income rose.

6. A shipping delay, not a permitting problem

The CFO volunteered the specific cause of the margin shortfall and was careful to distinguish it from the risk shareholders actually worry about.

"Importantly, we had a schedule and permitted fourth quarter shipment from Russia that did not leave as expected due to a shipping issue. That shipment, which was pushed out to the first quarter of 2026, would have driven down the average cost per SWU and positively impacted gross margin and net income." — Todd Tinelli, Chief Financial Officer

Asked about the flat guide, he repeated the distinction: "I would say this was not due to any permitting or any waivers that had to go along with any Russian material."

Assessment: Volunteering the cause of your own margin miss, quantifying its direction, and pre-empting the obvious wrong inference is good disclosure. The distinction is material: a logistics delay recovers in the next quarter, a permitting or waiver failure would have been a structural break in the supply that funds most of current revenue. It also sets up an easy comparison to grade in three months, because the deferred shipment lands in the first quarter and should show up as a lower average cost per SWU.

7. Trailing-twelve-month reporting, starting next quarter

Management announced a change in presentation that follows directly from its own argument about variability.

"To provide a better picture of our earnings and account for potential variances and future shipments as well as our quarterly earnings fluctuations, next quarter we will begin to present financials in both a quarterly as well as a trailing 12-month format." — Todd Tinelli, Chief Financial Officer

Assessment: Straightforwardly good, and overdue. A company whose quarterly results are set by which multi-year contracts happen to deliver should report on a rolling annual basis, and doing so removes the excuse-making that otherwise accompanies every quarterly miss. It also cuts both ways for management, since a trailing-twelve-month series makes a deteriorating trend as visible as a lumpy one. We take it as a signal of confidence rather than of cover.

8. The demand number got bigger and more specific

The CEO restated the demand case with a larger figure than the roughly 2.5 million SWU he cited in November, and added a price-curve statistic.

"Near-term domestic LEU demand alone is set to increase by approximately 6.5 million SWUs, stemming from Russia's exiting the market and the additional demand from restarts, uprates and new pledged reactors. The LEU pricing curve, which has experienced a 24% compound annual growth rate from 2019 to 2025, is an indicator of this increasingly constrained market and pent-up latent demand." — Amir Vexler, President and Chief Executive Officer

Assessment: The 6.5 million SWU figure is broader than November's 2.5 million because it now includes the volume vacated by Russia rather than only new-build and uprate demand, so the two are not comparable and neither is a contracted order. Against 3.5 million SWU of planned base-case capacity, the gap is the argument for expansion beyond the base case. The 24% compound annual growth in the price curve is the more useful statistic because it is observable and it is six years long, which makes the tightness structural rather than a spike.

9. The competitive framing sharpened

The CEO described the market structure Centrus is entering in terms shareholders should note.

"Absent Centrus, the market must rely on the duopoly of state-backed competitors with a shared signal point of failure centrifuge manufacturing risk. Providing Centrus with orders derisks customers' businesses and provides us with the clarity needed to stand up our operations." — Amir Vexler, President and Chief Executive Officer

Assessment: This is the clearest statement of the commercial pitch and it reframes the sale. Centrus is not asking utilities to buy cheaper enrichment; it is asking them to pay for supply diversification against a concentrated, state-backed alternative with a common manufacturing dependency. That argument supports price rather than requiring a discount, which matters for the eventual contract economics. It also explains why the contingent backlog exists at all: $2.1B of definitive agreements for output from a plant that does not exist is a diversification purchase, not a price purchase.

10. Hyperscaler prepayments are now named as a funding source

The prospective-investor conversations described in November acquired a specific instrument.

"We're already in conversation with hyperscalers for future prepayment or offtake-like agreements, another potential source of low-cost capital financing." — Amir Vexler, President and Chief Executive Officer

The CFO listed the same category alongside national security funding and foreign direct investment as parallel avenues being pursued.

Assessment: Prepayment is the most shareholder-friendly capital available to this company: it is neither dilution nor debt, it converts a customer conversation directly into construction funding, and it removes backlog contingency at the same time. That is why it is worth watching more closely than the equity market. It also remains, one quarter after it was first raised, a conversation with no named counterparty and no amount.

11. Hiring and site work moved from plan to execution

The operational tally for 2025 was specific: more than 300 new jobs announced at Piketon, more than 50 of those hired in the fourth quarter, over 140 employees added across Piketon and Oak Ridge in the year, design work initiated on a 150,000 square foot training, operations and maintenance facility at Piketon, and domestic commercial centrifuge manufacturing begun in December.

For 2026 the company guided to at least 150 net new employees, at least 100 in Oak Ridge against an announced 400, and at least 50 in Piketon.

Assessment: These are the least glamorous and most falsifiable commitments in the release, which makes them the most useful. Headcount and a Certified for Construction package are things a shareholder can check in twelve months without any cost disclosure from the company. Note the implied pace: at least 100 Oak Ridge hires against an announced 400 is roughly a quarter of the target in the first year, which is consistent with a 2029 first cascade and inconsistent with the schedule being pulled forward much.

Guidance & Outlook

The guidance's stated assumptions are worth reading as risk disclosure. The release conditions 2026 guidance on no significant change in restrictions on receiving and selling Russian LEU or other uranium products, no significant economic disruption, successful implementation of the planned expansion, and the finalization and funding of the DOE $900M task order. The second half of that list is the honest statement of what has to hold, and the last item is the one that matters most: the award that justifies the upgrade is also a guidance assumption.

Implied 2026 shape: flat revenue at the $450M midpoint against $448.7M delivered, with roughly $425M of capital deployment at the midpoint, means 2026 is a cash-consumption year by design. Against $1,957.2M of cash, the guided deployment is roughly 22% of the balance, so the company can fund this year and next from the balance sheet before any milestone payment, prepayment or additional raise. The first cascade is 2029, so there are three more years like this before revenue from new capacity arrives.

Street at: Q4 revenue consensus was $145M to $147M against $146.2M delivered, essentially in line. EPS consensus was $1.42 to $1.63 against $0.79, a miss of 44% to 51% depending on which estimate is used. The dispersion in EPS estimates is itself informative: with the convertible notes moving in and out of the diluted calculation and an at-the-market program running, the denominator was hard to know from outside, which may explain some of the disagreement. It does not explain the miss: even at the prior-year diluted count of 16,778k, $17.8M of net income is $1.06 a share, below both estimates, so most of the shortfall is in net income.

Analyst Q&A Highlights

Compressing the build schedule

The first question went to the timeline, and to whether the previously discussed 42-month lead time can be shortened.

Q: "Maybe we could talk about the time line a little bit more. It sounds like you're looking to improve on the 42-month forecast that you've talked about in the past. Could you talk about some of the potential initiatives that might help pull that forward, whether it's more funding, other federal programs or anything else we should be aware of?"
— Ryan Pfingst, B. Riley Securities

A: "Continuous improvement and the ability to reduce unit cost and go faster is paramount as we build out our capacity. We have to do it in parallel. And I will tell you that we have quite a few folks dedicated to this, and we're dedicating resources to this. And it is evident to me that we already have a good list of opportunities that we're executing on."
— Amir Vexler, President and Chief Executive Officer

Assessment: The answer names no initiative and no target, but the framing is a change from November, when no dated schedule had been given. A management team that describes cost reduction and cycle time as day-one activities with dedicated resources is behaving like an operator rather than a promoter. The absence of a quantified compression target means the 2029 date is the one to model until it moves.

When contingent backlog becomes firm

The question a shareholder most needs answered was asked directly and declined.

Q: "On the commercialization side of the -- LEU commercialization, is that capacity ramp sort of similar time line with the HALEU ramp? And when does that kind of reach a point where you have that backlog turn from contingent to sort of final?"
— Robert Brown, Lake Street Capital Markets

A: "This is not something that at this point we will get into a lot of contractual details. Needless to say that we, as I stated before, we have the intention and the plan and we're progressing towards fulfilling our commitments... And again, I will not comment specifically on the details of the contractual arrangements, but this is exactly what we're executing towards."
— Amir Vexler, President and Chief Executive Officer

Assessment: A non-answer on the most consequential mechanical question in the story. $2.3B of contingent backlog becoming firm is the event that converts this company from a developer into a contracted producer, and there is no stated trigger, threshold or date. Management said in prepared remarks it is making strong progress on removing contingencies and gave no measure of that progress. Until a conversion is announced, treat the contingent portion as demand evidence rather than revenue.

The shape of the capital spend

An analyst asked whether the 2026 capital number is representative of the years to 2029.

Q: "If you could talk about, if possible, maybe the linearity of that CapEx expected throughout the year. And then also, as we think about the number that you gave, is that kind of a representative number on an annual basis as you work towards that initial 2029 operational date?"
— Eric Stine, Craig-Hallum Capital Group

A: "Looking past 2026, obviously, our CapEx spend manufacturing process gets more linear. Currently, right now, this -- the number that we have provided for guidance is capital spend. So that does include some long-lead procurement, potential prepayments on supplier agreements and also engineering work at our Piketon facility. So the first year is not indicative of the linear spend that I anticipate seeing in '27 through '29 period due to those factors."
— Todd Tinelli, Chief Financial Officer

Assessment: The most useful modelling answer of the call, and it settles less than it seems. 2026 carries long-lead procurement and supplier prepayments, and 2027 through 2029 is described as more linear, but the CFO did not say whether that spend runs above or below 2026. A shareholder should not assume the annual capital requirement falls from the 2026 midpoint, against a cash balance of $1,957.2M and a revenue line guided flat.

How much capacity the cost curve requires

The most valuable exchange on the call tested whether full capacity is needed to reach a competitive cost.

Q: "Achieving the nth-of-a-kind, I'm just wondering if I'm reading into this correctly. I had always assumed that maybe it took the 3.5 million SWU to achieve that. And it sounds like that's not the case, that you can achieve that target with less capacity. And so I'm just curious if you might be able to give a little bit more framework on the relative capacity to achieving the nth-of-a-kind target."
— Jonathan Dorsheimer, William Blair

A: "The nth-of-a-kind move will be achieved way before the 3 million SWUs, as you mentioned... As you know, one of the greatest obstacles for any new entrant into the market is the nth-of-a-kind cost... As always, I'm very careful. We're not disclosing full cost. And so I will -- I'll stop there."
— Amir Vexler, President and Chief Executive Officer

Assessment: Confirmation, in response to a well-constructed question, that a competitive cost position needs less than full capacity, a point the CEO had already made in prepared remarks. This is one of the disclosures that most change the equity's risk profile since our initiation, because it shortens the path between here and an economically self-sustaining plant. It is also the second time in two calls that the answer stops precisely where a number would begin.

Washington and the 2028 cliff, asked again

The same analyst who raised the import prohibition in November raised it again, and the answer was the same with one addition.

Q: "Has the government started to come around on this Jan 1, 2028 cutoff for Russian imports? Has there been any willingness to discuss that given domestically you guys should be first to market, that there isn't really going to be supply in 2028, 2029 domestically?"
— Joseph Reagor, ROTH Capital

A: "On the government side, I really can't report anything because there's nothing really to report. We're not -- there's nothing out there in the public domain that expresses any concerns. But as you know, the real concerns and the real push will come from the customers, the industry, the utilities themselves if and when the problem presents itself."
— Amir Vexler, President and Chief Executive Officer

Assessment: Unchanged from November, which means the risk is unchanged and one quarter closer. The added observation is that pressure to revisit the deadline would have to originate with utilities rather than with Centrus, which is credible and also means the company has no lever on the single largest identifiable discontinuity in its own revenue line. Note the tension with the same answer's demand argument: a supply crisis in 2028 and 2029 is bullish for a producer and hostile to a reseller, and Centrus will still be the latter in 2028.

Why the revenue guide is flat when SWU prices are rising

An analyst put the central valuation question to the CFO.

Q: "The '26 guide is flat year-over-year for revenue. Just wondering if you could kind of remind us of some of the contract dynamics for your long-term supply arrangements, and then also if and when we should expect to see any kind of step-up in line with some of the changes we've seen in long-term SWU prices in line with some of the indices."
— David Choe, UBS

A: "Our guidance is flat. We picked the midpoint... But we see upside to our potential guidance. The market continues to improve itself. We've seen some record-high spot prices in SWU. We hope those dynamics continue. And a lot of our supply is contracted, both from 2 foreign sources. We feel comfortable with our -- on our supply side. Our goal is to maximize the margin on the revenue side and sales to our customer."
— Todd Tinelli, Chief Financial Officer

Assessment: The question asked when rising indices reach the revenue line and did not get a date, which is now the second consecutive call in our coverage in which that has happened. What the answer does supply is a claim of upside to the guide and confirmation of contracted supply from two foreign sources. The full-year figures make the underlying point better than the answer does: realized SWU price fell 1% in 2025 while the published curve compounded at 24% a year since 2019. The reprice is real and it arrives on contract renewal, not on the index.

Milestones that would de-risk the schedule

An analyst asked what observable checkpoints exist for outsiders.

Q: "My question is regarding the time line that you guys have put out to initial enrichment capacity. I'm curious, are there more important milestones or roadblocks that would more tangibly derisk the time line that you guys have put out that you could communicate along the way?"
— Jeffrey Grampp, Northland Capital Markets

A: "If you think about lead times to production, a lot of it has to do with our own cycle times, our supplier cycle times, the lead times that we can get our suppliers to commit to, how we configure our supply chain. All of these are being worked on... I foresee that when those mature and there's something to report, we certainly will be very transparent in how we announce them."
— Amir Vexler, President and Chief Executive Officer

Assessment: The CFO's follow-on was the more substantive half, committing to provide updates on how milestones can be tracked and saying the company would hold itself accountable. That is a promise a shareholder can hold, and combined with the 2026 operational guidance, the first Certified for Construction package and the hiring targets, it means the next four quarters finally contain checkable events. That is a meaningful change from November, when there were none.

What 12 metric tons of HALEU represents

An analyst tested the scale of the HALEU target against reactor requirements.

Q: "Could you talk about the HALEU production target that you have? On a metric ton basis, it looks like it's roughly enough for an initial load for 1 of the advanced reactors. So could you just confirm whether that quantity includes both an assumption of market demand as well as the demand that you see from the DOE task orders?"
— Ted Giletti, on for Vikram Bagri, Citigroup

A: "We got to remember the intent of the DOE program was to stimulate the market. So here, we had a market where there was not necessarily demand that anybody could commit to, and the Department of Energy had wisely said, 'Well, what we're going to do is we're going to invest, we're going to create the capacity, and then we're going to allow the OEMs to contract for that capacity.' And exactly as they planned, we're starting to see that."
— Amir Vexler, President and Chief Executive Officer

Assessment: The observation embedded in the question is the important one: the analyst suggested that 12 metric tons is roughly one advanced reactor's initial load, and management said only that it depends on the reactor design and that there is no specific answer. If the analyst's sizing is right, the HALEU business is a market-seeding position rather than a volume business, which is exactly why the DOE is paying for it, and the award should be valued as capacity creation and cost-curve progress rather than as a revenue stream.

How SWU prices need to evolve to underwrite the capital

An analyst asked the economics question in its most answerable form and got the fullest reasoning of the call.

Q: "When we think about the CapEx you're spending to build out capacity over the next decade, how do you think about the evolution of SWU prices to sort of support the economics behind it? And I know you're not going to give me an exact number, but like how do we think about SWU for HALEU and LEU?"
— Stephen Gengaro, Stifel

A: "The SWU prices, in my view, are going to adjust down only if 1 of 2 things happen... Either demand goes down, in other words, we're going to start shutting down reactors, or basically the supply side is going to outpace the demand... And for the latter, most of the announced expansions, either it's us or our competitors, is based on contracted SWUs. So there is really nobody out there that is building what I'd call just-in-case or ready-to-serve SWUs that they will be eager to get off and sell on the market."
— Amir Vexler, President and Chief Executive Officer

Assessment: The strongest argument management has made in two calls, and it is a structural one rather than a forecast. If most announced expansions in the industry are underpinned by contracted volume and, as the CEO said, nobody is building just-in-case capacity, then no speculative capacity exists to break the price, which means the downside case for SWU pricing requires reactor shutdowns. That is a genuinely asymmetric setup and it supports the upgrade. It still does not tell a shareholder what price the plant needs.

Whether the DOE money is front-loaded

An analyst probed the cash timing of the award.

Q: "The question is about the $900 million from the task order -- DOE task order. Given that advanced reactors are likely to come online towards the end of the decade, should we expect this 10-year contract to be really be 5-year contract and like to get the money up front-loaded?"
— Sameer Joshi, H.C. Wainwright

A: "Obviously, longer-term commitments are more favorable to us. That would be reflected in the pricing as well... Your question was about front-loading some of these contracts from a cash perspective. I mean, obviously, these are all the things that we constantly try to optimize and we will continue to try to optimize in the best interest of our shareholders, particularly that most of the SWUs that we're selling is what I call expansion SWUs. These are SWUs that we're investing capital in."
— Amir Vexler, President and Chief Executive Officer

Assessment: No cash schedule was given, which combined with the CFO's disclosure that payments are milestone-based means the company funds construction ahead of receipts. That is the practical reason the $2.0B cash balance is not excessive and the reason the 2026 capital guide can exceed the DOE award's annual pace. The phrase worth keeping is "expansion SWUs": management is explicitly telling shareholders that future sales are the return on capital being deployed now, which is the correct framing and also the admission that the return is years out.

What is inside management's control

An analyst asked how much of the schedule depends on third parties.

Q: "I was wondering if you can double-click on what opportunities you have to pull in the 42-month time frame you spoke of, as well as, I'd say, to build out more broadly? I guess what I'm getting at is, what's under your control versus suppliers or other third parties? How much is contingent on additional financing, negotiations with suppliers and partners, or, I think you alluded to earlier, government approvals related to national security?"
— William Peterson, JPMorgan

A: "We're very unique in that we are building our own centrifuges. We have the capability, obviously, to build our centrifuges. We have partners that are suppliers. And so a lot of what we do is really within our control and we have a pretty wide range of motion in being able to generate the opportunities... To me, this is fundamental, elementary manufacturing excellence."
— Amir Vexler, President and Chief Executive Officer

Assessment: Vertical integration in centrifuge manufacturing is the answer to the question and it is a real structural advantage over a project that has to buy its core equipment from the same constrained suppliers everyone else uses. The CFO's addition was the more revealing part: that a well-capitalized balance sheet with line of sight over 24 months exists specifically so the company is not forced to raise expensive capital in a weak market. Read against the November at-the-market program, that is a coherent policy rather than opportunism, and it explains why the raise happened before the award rather than after it.

LEU feed for HALEU, and the 2028 interaction

The final question tied the two products and the supply cliff together.

Q: "Longer term, what is your sourcing strategy for the LEU feed for the HALEU? And maybe put another way, how much of your LEU production will be captive to your HALEU production? And how much of that 3.5 million SWU then will be available to satisfy LEU primary demand? And then further, if the LEU capacity isn't available to meet the HALEU feed, what are the options, particularly given that the Russian supply will be going away after 2028?"
— Lawson Winder, Bank of America

A: "It is our intention to be able to provide both LEU enrichment and HALEU enrichment to the optimal and maximum capacity that we can... if, for example, hypothetically speaking, if you have a HALEU contract for which you don't have sufficient LEU feed, I mean, that usually, contractually speaking, is not an issue. There is contracting mechanisms either by ourselves or the customers that are able to utilize other suppliers for that."
— Amir Vexler, President and Chief Executive Officer

Assessment: The technical point the CEO made earlier in the call is the one to hold: HALEU requires substantial LEU enrichment as feed, so HALEU demand is a force multiplier on LEU capacity rather than a competing use of it. The answer to the split question was that there is no split, only maximisation, which is commercially sensible and leaves the capacity allocation unmodellable. The fallback of contracting feed from other suppliers is a reasonable answer in a loose market and a weak one in the tight market management itself forecasts for exactly those years.

What They're NOT Saying

1. The capital cost of the Piketon build-out. Still no figure, now alongside a named EPC contractor, a dated first cascade and a $425M midpoint of one year's capital deployment. Management explicitly stopped short again: "As always, I'm very careful. We're not disclosing full cost."
2. What nth-of-a-kind cost actually is. The company disclosed that it will reach the threshold below 3 million SWU and did not disclose the cost at that threshold, so the competitive claim cannot be tested against any competitor's economics or against a market price.
3. Any trigger for converting the $2.3B contingent backlog. Asked directly, management declined on contractual grounds. Prepared remarks claimed strong progress on removing contingencies with no measure of progress and no date.
4. The cash timing of the $900M award. Payments are milestone-based and no milestone schedule, first-payment date, or annual profile was given, in a year guided to deploy $350M to $500M of capital.
5. The accumulated HALEU Phase 2 fee. Costs incurred since November 2024, now more than fourteen months of them, still carry no fee, with the amount at issue unquantified and no analyst asking.
6. How much of the 2026 capital deployment is prepayments rather than capital expenditure. The CFO flagged that some spend will not appear in capex but will hit free cash flow, and gave no split, which makes the guided reported capex number unknowable.
7. Contract pricing structure. Asked what portion of 2026 deliveries has cost of supply fixed, the answer was that the company does not comment on the contractual makeup of its contracts. The two foreign supply sources were confirmed and not named.
8. Whether the at-the-market program continues in 2026. The 2025 draw and its average price were disclosed. Remaining capacity, intended pace and any price discipline for 2026 were not.
9. The size of the NNSA opportunity. Now confirmed as an intent to sole-source certain enrichment activities from Centrus, and still with no value, scope or date, one quarter after first disclosure. Management called it a potential source of low-cost capital without sizing it.
10. Any hyperscaler counterparty or amount. Prepayment and offtake-like agreements are now named as a funding avenue and there is no named counterparty, structure or figure, one quarter after the same conversations were first described.
11. Why the fourth quarter consumed operating cash. Full-year operating cash flow of $51.0M against $99.4M through nine months implies roughly $48.4M consumed in Q4 on $17.8M of net income. Netting the nine-month statement against the full year puts the cash into settling $137.4M of payables under inventory purchase agreements while inventory fell. Neither document addresses the quarterly cash reversal, and no analyst asked.
12. The $170M of options on the DOE award. Disclosed as potentially taking the award above $1B, with no description of what triggers them, who exercises them, or when.

Market Reaction

• Pre-print setup: LEU closed at $264.99 on February 10, up 9.2% year to date from $242.76 against the S&P 500's +1.4%, and up 131.7% over the trailing twelve months. The stock had weakened into the print, down 13.5% over the trailing thirty days from $306.19. The 52-week closing range entering the print was $54.43 to $436.00.
• Reaction session: the stock opened at $253.00 on February 11, 4.5% below the prior close, traded a range of $206.23 to $258.00, and closed at $210.16, down 20.7% or $54.83.
• Intraday pattern: the high of $258.00 marked a 2.6% decline and the low of $206.23 a 22.2% decline. The stock closed 1.9% off its low, in the bottom decile of a 25.1% intraday range.
• Volume: 3.2M shares against a 30-day average of 1.3M, or 2.5x normal.
• Market context: the S&P 500 was unchanged on the session, so the entire move was idiosyncratic.

This is a worse reaction than November's on every measure except the opening gap, which was smaller: a much larger close-to-close decline, a finish 1.9% off the session low rather than about a third of the way up the range, and a higher volume multiple. The release and the call were both public before the open, so the pattern of a modest gap and a close near the low suggests the selling built as holders worked through them, which is the opposite of November, when the stock was already down 10.9% before the call began.

The most likely trigger is the guidance, not the quarter. Revenue was in line and the EPS miss had disclosed mechanical causes. What was new on February 10 was a flat 2026 revenue guide alongside $350M to $500M of capital deployment, in a name whose shareholder base had been buying an accelerating enrichment story. Flat revenue plus a capital-spending step change is the exact combination that removes a growth multiple, and it arrived in the same release that told holders the first new cascade is three years away.

The $900M DOE award was already known, having been announced on January 5, so it was in the pre-print price and could not cushion the reaction. That timing asymmetry matters for how the session should be read: the market discounted a first-time capital plan against an award it had already paid for five weeks earlier.

Some perspective on the base. The stock closed the reaction session at $210.16, still 83.7% above the $114.39 it traded at a year earlier, and now down 13.4% year to date from $242.76 in a market up 1.4% entering the print. Across two consecutive prints the stock has fallen 14.7% and 20.7% on the reaction sessions, and it is down 35.5% from the $325.73 pre-print close in November, while the funding architecture has strengthened materially. That divergence is the upgrade.

Street Perspective

Debate: Does the $900M award change the funding equation or just add to it?

Bull view: The bull case holds that this is categorically different capital. Management was explicit that the award comes through a procurement and involves neither debt nor equity, so unlike the $805M of converts and the $533.6M of at-the-market equity it does not dilute or lever the existing shareholder. With $170M of options potentially taking it above $1B, it funds a large share of the HALEU capacity while the company retains the whole residual. It is also, by itself, one of the three events we said would justify an upgrade.

Bear view: The bear response is that the award is a selection subject to negotiation, not a signed contract, and the company's own guidance lists its finalization and funding as an assumption on which 2026 depends. Payments are milestone-based, so cash arrives after work is done and funded. And $900M covers only about 60% to 85% of a capital programme running at $350M to $500M a year for three years, and less for each year it runs beyond that.

Our take: The bulls have it on kind and the bears have it on sufficiency. Non-dilutive public capital is a step-change in the quality of the funding stack, and that matters more than the amount, because it establishes that the public leg exists at all. The negotiation risk is real and modest: DOE selected a named subsidiary for a specific scope after a competitive process. Sufficiency is the honest concern, and it is why prepayments and the NNSA opportunity remain the things to watch.

Debate: Is a flat revenue guide a disappointment or an irrelevance?

Bull view: The optimistic reading is that 2026 revenue was never the point. The existing resale business is a bridge whose supply is legislated away after 2027, and the equity's value is the plant. Flat revenue with $425M of capital deployment at the midpoint is precisely what converting a reseller into a producer looks like, and management guided conservatively while stating it sees upside.

Bear view: The skeptical reading is that flat revenue in a year of record SWU prices is evidence the company cannot monetize the market it keeps describing, that the first new cascade is not until 2029, and that between now and then shareholders fund roughly $425M a year against a business earning $117.5M of gross profit. A 20.7% single-day decline is the market repricing three years of cash consumption.

Our take: The bears describe the next three years correctly and the bulls describe the asset correctly. The reconciling fact is the balance sheet: $1,957.2M of cash covers roughly four and a half years at the guided midpoint before any milestone payment, prepayment or further raise. This is a funded development programme, not a stretched one. The flat guide is a real disappointment for anyone who owned this for 2026 earnings and close to irrelevant for anyone who owns it for 2029 capacity, and the reaction says the former group is still large.

Debate: Has the dilution been worth it?

Bull view: The bull argument is that management raised $533.6M of gross equity in 2025, including $390.4M at an average of $269.21 per share, which is 28% above the reaction close. It also secured $782.4M net from a 0% coupon convertible. The result is roughly $2.0B of cash, $782.4M of net cash, and a company that intends never to be forced to raise into weakness. The CFO said as much: the objective is to avoid being forced to raise expensive capital in a downward market.

Bear view: The bear response is in the per-share line. Full-year net income rose 6.3% and diluted EPS fell 12.8%. Fourth-quarter net income fell 66.9% and diluted EPS fell 75.3% on a 34.3% increase in diluted shares. Existing holders funded the war chest and their claim on the earnings it protects has shrunk, with no announced end to the issuance.

Our take: Worth it, and the average price is why. Selling equity at $269.21 to fund an asset whose funding path has since improved is good capital allocation, and the disclosure of the realized price is the transparency November's announcement lacked. The bear point is arithmetically correct and it is the price of optionality on a first-of-a-kind industrial project. What would change our view is issuance continuing at materially lower prices, and that is the specific thing we will watch.

Model Update Needed

Valuation: At the $210.16 reaction close on 19,664,565 shares of Class A and Class B common stock outstanding at December 31, the equity is worth roughly $4.13B. Netting approximately $782M of net cash gives an enterprise value near $3.35B. That is 7.5x full-year 2025 revenue, 7.4x the midpoint of the 2026 revenue guide, 28.5x full-year gross profit, 0.88x the $3.8B headline backlog and roughly 2.2x the non-contingent portion. Since our initiation the enterprise value has fallen from about $4.60B to about $3.35B, a decline of 27%, while a $900M non-dilutive public award, an EPC contractor, a build schedule and a cost-curve threshold below full capacity have all been added.

We upgrade to Outperform and set a fair value range of $260 to $320, a midpoint of $290 that implies +38.0% from the $210.16 reaction close. The $260 low end implies +23.7% and the $320 high end +52.3% against the same close. The low end approximates roughly 1.1x headline backlog on an enterprise basis, in line with where the shares traded at our initiation before any of this quarter's disclosures existed. The high end capitalizes the base-case build-out with the DOE award finalized and the first Certified for Construction package released, without crediting expansion beyond 3.5 million SWU, the NNSA opportunity, or any hyperscaler prepayment. A disclosed capital cost remains the single event that would let us narrow the range rather than move it.

Thesis Scorecard Post-Earnings

Overall: The thesis improved on the pillar that was failing and deteriorated nowhere. At initiation the weak point was Bull #3: the only evidence on funding was a $1B equity shelf with no stated use, and we said so. One quarter later the company has a $900M public procurement selected, $170M of options behind it, an EPC contractor signed, a dated first cascade, a capacity threshold for competitive cost below full capacity, the first guidance in our coverage, and equity already raised at a price 28% above the current one. Bear #1 has been partly refuted by the full-year segment figures. Bear #4 has been refuted by a 27% decline in enterprise value against an improving fact set. The two live risks are unchanged: no capital cost, and a 2028 supply cliff that arrives a year before the first new cascade.

Action: Upgrading to Outperform from Hold, fair value $260 to $320. In November we named three disclosures, any one of which would move the rating, and said we would downgrade if at-the-market issuance ran materially ahead of any of them. By its letter, that trigger was met: $390.4M of stock was sold through the November program before the January 5 DOE selection. We do not act on it because the concern behind the trigger was open-ended issuance at any price, and this stock was sold at an average of $269.21, 28% above where it now trades, with the award following five weeks later. The next four quarters finally contain checkable events, and we will grade them: the finalized DOE contract, the first Certified for Construction package, at least 150 net new hires, the deferred Russian shipment showing up as a lower average cost per SWU in the first quarter, and the first trailing-twelve-month presentation. We would revert to Hold if the DOE contract fails to close, and downgrade further if equity issuance resumes materially below $210.