AXT Halved From Its Peak — Why This Isn't a Value Dip

AXT has been one of the most violent stocks of the AI-optics trade: up roughly 18 times off its 52-week low of $3.70, to a record high around $143, and then cut nearly in half to about $68 in a matter of weeks — all while the business underneath got stronger. That price-versus-fundamentals divergence is why people keep calling this a dip worth buying. The honest answer is that it depends entirely on which report card you are reading.

The trailing report card says "expensive" — and screens pass on it

Start where every conventional valuation check starts: the numbers you can already verify. At a market value of roughly $4.46 billion, AXT trades near 36 times trailing sales. That is not cheap by any sector-relative standard. Put it next to the profitable names that share its corner of the semiconductor supply chain — Amkor at about 1.7 times sales, ACM Research around 5 times — and the premium is an order of magnitude.

Trailing profitability tells the same story in a different register. Over the past twelve months AXT's operating margin is still negative, free cash flow is negative, and return on equity is under 1%. Trailing earnings are effectively break-even, which is why the trailing P/E is a meaningless three-figure number. On any value screen, profitability screen, or cash-flow screen, this stock fails. "Cheap" is simply not a word the factor stack supports, no matter how far the price has fallen from its peak.

The honest buy case lives entirely in the forward report card

None of that is a reason to dismiss it — it is a reason to be clear about what is actually driving the rating. This is a stock whose case is all forward-looking, which makes those trailing numbers more decoy than verdict.

The forward picture is real. In the quarter reported in late July, AXT posted record revenue of $47.6 million, up 164% year over year, driven by record indium-phosphide substrate sales of $30.7 million — nearly two-thirds of the total. The heart of the trade is that indium phosphide is the wafer material in the optical transceivers that shuffle data inside AI data centers, and there are only about three credible suppliers worldwide. With the industry flagging the most severe indium-phosphide shortage on record and Q4 wafer prices expected to rise, AXT went from a near-zero gross margin a year ago to about 45% last quarter, and from a loss to its first recurring profit ($0.17 a share, GAAP).

The scale of the bet is what makes the stock worth a hard look. Management guided to roughly $66 million of third-quarter revenue and guided production capacity to nearly triple by the end of this year and roughly double again to about $130 million per quarter in 2027, with a backlog already above $100 million. Long-term supply agreements and prepayments from Coherent, Casella, and Lumentum are reserve capacity before it is even built. AInvest's aggregate signal currently labels the stock a Buy, consistent with the revision momentum — the direction of change is unmistakably up.

Where the risk actually sits

Now do the forward math the same way the market is implicitly doing it. At a $4.46 billion market value, the guided third-quarter revenue annualized is only about $264 million — roughly 17 times annualized run-rate sales. Even if the full 2027 capacity ramp to $520 million a year materialized, the stock would still sit near nine times sales. In other words, the current price already underwrites the entire shortage-plus-scaling story at once. Nothing about the improving business makes this cheap; it makes it a momentum-and-revisions trade on a single-product lever.

That concentration is where a disciplined investor should be skeptical. More than half of revenue now comes from China, and management itself flagged that export permits gate how much of the guided quarter can actually ship. The shortage is the bull case's fuel, so its flip is the bear case: if demand normalizes or capacity lands ahead of sell-through, a business priced for near-perfect execution reprices violently — and a stock with this much daily volatility (roughly double-digit percentage swings are routine here) will do it fast. Add the roughly $632 million raised in a secondary this spring, which put a net-cash balance sheet behind the expansion but also diluted existing holders, and the setup is a high-octane single-position story, not a value dip.

The factor discipline here is the barbell, not the conviction. AXT belongs at the aggressive pole of a portfolio — a spec-sleeve holding sized so that the ~50% drawdowns that are normal for it don't break the plan — paired on the other side with durable cash-flow and dividend names that carry the barbell when the optics trade wobbles. The strengthening report card is real and the revisions support it, but this is a bet on a forward that is already mostly priced. Buy the dip only if you are buying the shortage thesis and can hold it without reaching for leverage. Anyone underwriting it as "cheap" is reading the wrong report card.