A revised fair-value read after the revenue guide tripled, the balance sheet healed, and the stock ran 3.6x off our last note. The thesis is no longer "valuation makes no sense" — it's "the dream is now partly in the numbers, but you're still paying for 2027 today."
In February we called fair value at $25–35 with the stock at $48, on a ~$250M revenue base and negative EBITDA. We were right that it was a momentum/narrative name — and that call aged poorly, because the narrative became fundamentals faster than we modeled.
What's different now is not sentiment — it's the income statement and the balance sheet. Management has guided 2026 revenue to north of $1.1B (vs. the $250M base we anchored to) and is targeting a $1.4B optical-transceiver run-rate by Q3 2027. Cash went from ~$216M to $449M. The 800G/1.6T ramp for hyperscalers is real and supply is expected to lag demand into 2027. The serial-diluter critique still holds, but they're diluting at $170 instead of $24 — a very different conversation for legacy holders.
So we are retiring the old number. At ~$173, however, the stock is still trading above any fundamentally-defensible fair value we can build. The market is paying for flawless execution of a ramp that, on management's own Q1 commentary, slipped to the back half of the year. Our posture: respect the trade, don't fight the tape, but trim into strength rather than chase.
The print that mattered: Q1 2026 revenue of $151.1M, +51% YoY, though it was a slight miss vs. the ~$155M consensus and adjusted EPS came in at −$0.07 (vs. −$0.05 expected). Q2 is guided to $180–198M with EPS −$0.03 to +$0.03 — i.e., management is now drawing a line to roughly break-even on the bottom line for the first time. The H2 ramp is the whole ballgame; 800G timing got pushed to the back half, which is the single biggest execution risk in the model.
| Last price (6/7/26) | $173.60 |
| Market cap | ~$14.2B |
| Shares outstanding (approx.) | ~81.8M |
| 52-week range | $15.29 — $233.67 |
| Revenue (TTM) | $455.7M |
| Net income (TTM) | −$38.2M |
| Gross margin | ~30% |
| EPS (TTM, diluted) | −$0.64 |
| Cash | $449.4M |
| Price / TTM sales | ~31x |
| Price / 2026E sales ($1.1B) | ~13x |
| Beta | 3.74 |
| Next earnings (est.) | Late Jul / early Aug 2026 |
The honest problem: AAOI still loses money, so this is a revenue-multiple exercise, not a DCF you'd defend in committee. We anchor on forward sales and let the multiple carry the disagreement. Peers bracket the range — commodity optical names (COHR, LITE) trade at low-to-mid single-digit sales multiples; AI-connectivity hypergrowth names (CRDO, ALAB) command far more. AAOI sits in between: hypergrowth and US-manufactured, but thin-margin and competitively commoditized at the MSA level.
| Scenario | Basis | Sales Mult. | Implied Value |
|---|---|---|---|
| Bear | 2026E rev $1.1B; commodity multiple, ramp disappoints, ATM grinds | 4–5x | $54–67 |
| Base | 2026E rev $1.1B at premium for AI mix + US capacity; ~6–7x on 2027E ~$1.5B | 8–10x | $108–135 |
| Bull | 800G/1.6T ramp lands on time, margins inflect, market keeps paying momentum multiple | 12–14x | $165–200+ |
We center the base case near $120. That already bakes in a generous premium to the optical peer group for the AI tailwind and the hyperscaler relationships. At $173 the stock is effectively trading in the bull scenario — pricing the 2027 run-rate as if it's already de-risked. It isn't.
This was the core of our original skepticism and it remains intact, just at a higher altitude. AAOI announced a $600M ATM in early June and shareholders approved a new 2026 equity incentive plan on June 4. History is unambiguous: when this stock rips, management sells into it. The good news for the equity is that diluting at $170 is dramatically less destructive per dollar raised than the death-spiral raises of the $15–25 era — $600M is only ~4% of the cap here. But it caps upside: every rally toward the ATH invites fresh supply, and every dilution filing trips the algos. That's a structural reason the stock struggles to hold its highs.
| Q2 2026 earnings (late Jul / early Aug) | Ramp & margin proof point |
| 800G / 1.6T volume ramp — H2 2026 | Core thesis validation |
| FAB2 Sugar Land + Taiwan capacity online | Supply for $1.4B run-rate |
| ATM share-count disclosures | Dilution pace, rally caps |
| Hyperscaler capex commentary (NVDA/hyperscalers) | Sentiment beta driver |
| Insider Form 4 activity | Conviction signal |
The thesis upgraded, the price overshot. Fair value moved from $25–35 to roughly $120 base — a real re-rating earned by a real guide. But at $173 you're paying the bull case in full for a ramp that just slipped a quarter. Own the optical/AI theme, respect the momentum, but trim into strength and let the ATM-driven pullbacks come to you rather than chasing the highs.