AAOI Holds $425M of 1.6T Orders, Can Ship $60M — and Filed to Sell $550M

Applied Optoelectronics just posted the best quarter in its history, returned to GAAP profit, and signed a three-year supply deal with a Tier 1 hyperscaler worth up to $1.5 billion. Then, on 17 November, it filed to sell $550 million of its own stock. The two facts sit awkwardly together, and that is the point: the demand is real, the constraint is physical, and management chose to fund the gap by diluting shareholders into a rally. This piece asks the only question that matters — can $Applied Optoelectronics(AAOI)$ turn a signed 1.6T contract into shipped revenue before the cash cost of waiting eats the opportunity?

Figure 1 — Record revenue, a $1.0–1.5B 1.6T deal, a $50–70M Q4 shipment cap and a $550M ATM, all in one week. Sources: AAOI Q3 FY26 8-K, 5 Nov 2026; Q3 call via BigGo, 5 Nov 2026; ATM filing, 17 Nov 2026.

The number that does not add up

Here is the contradiction in its plainest form. On the 5 November earnings call, chief executive Thompson Lin said the company held between $400 million and $450 million of 1.6T transceiver orders from customers. He also said DSP supply constraints meant only $50 million to $70 million of that could actually ship in the fourth quarter. Roughly $425 million of demand, roughly $60 million of deliverable product. That is not a demand problem. It is a components problem wearing a demand problem's costume.

The market initially liked the quarter anyway. Revenue reached $269.9 million, up 127.5% from $118.6 million a year earlier and 40.7% sequentially from $191.9 million in Q2, according to the company's 8-K exhibit and Yahoo Finance. Non-GAAP earnings were $0.17 a share against a $0.09 loss a year ago, a 21.43% beat versus the Zacks consensus. Sixth consecutive record revenue quarter. First GAAP net profit in years, $6.3 million, against a $17.9 million loss in Q3 2025. Every headline number pointed up.

But the headline is not the thesis. The thesis is that a company cannot recognize revenue it cannot build, and the part it cannot build is the fastest-growing, highest-value part of the mix.

Figure 2 — Orders on hand versus Q4 shippable volume, both at the midpoints of management's ranges: about $425M against about $60M. Sources: AAOI Q3 FY26 earnings call, CEO Thompson Lin, via BigGo, 5 Nov 2026.

Two businesses, one margin problem

The revenue mix explains both the excitement and the caution. Datacenter contributed $155.61 million in Q3, up 254.2% year on year and 44.5% sequentially, and now represents about 58% of the total, according to Yahoo Finance and Zacks. CATV was a record $110.88 million, up 57% year on year, helped by 1.8 GHz amplifier shipments and the millionth amplifier shipped. Within datacenter, 800G revenue nearly quadrupled sequentially to $47.5 million from $12.8 million, while 400G almost doubled to $93.4 million from $48.4 million.

That last detail is the tell. Zacks reported that 800G growth came in below management's own expectations because a major hyperscaler shifted demand toward 1.6T faster than anticipated. Customers are pulling forward, and the company cannot follow. It holds $400–450 million of 1.6T orders but can deliver only $50–70 million of that in the quarter — material availability, not demand, sets the pace.

Figure 3 — Q3 FY26 revenue by segment: datacenter $155.6M, CATV $110.9M, other $3.4M. Sources: AAOI Q3 FY26 results via Yahoo Finance/Zacks, 5 Nov 2026.

The margin line shows the cost of that mix shift. Non-GAAP gross margin was 28.7% in Q3, down 230 basis points year on year and 110 basis points sequentially, and slightly below the 29–30.5% guidance range, Zacks reported. The reason is a higher share of lower-margin 400G product. Management still targets roughly 40% non-GAAP gross margin over the long term, a gap of more than eleven points that only a 1.6T-heavy mix can close.

The guidance the supply chain will not allow

For Q4, the company guided revenue to $385–420 million with non-GAAP gross margin of 29.5–31% and non-GAAP net income of $45.9–60.4 million, or $0.48–0.63 a share on about 96 million shares, per the 8-K. The midpoint implies another large step up. The problem is that the same call that produced the guide also capped 1.6T shipments at $50–70 million. The revenue has to come from somewhere — 400G and CATV — which is precisely the mix that keeps margins near 29%.

Figure 4 — Quarterly revenue trajectory with Q4 plotted at the $385–420M guidance midpoint. Sources: AAOI Q1/Q2/Q3 FY26 results; Q4 FY26 outlook in the 8-K filed 5 Nov 2026.

Capacity is the other half of the story. Chief financial officer Stefan Murry said installed manufacturing equipment capacity approached 330,000 units a month at quarter end, and that the company now expects to install enough equipment to produce over 500,000 800G and 1.6T units a month by end-2026 — below the original 650,000 target — because of delays in obtaining electrical power at the Sugar Land, Texas facility. The target for end-2027 remains over one million units a month. Every one of those figures is a plan, not a shipment.

Bull Case

  1. The 1.6T contract with a Tier 1 hyperscaler is signed and worth an estimated $1.0–1.5 billion over June 2026 to May 2029, with the customer committed to at least 10% of its 1.6T demand — named, sourced revenue visibility.

  2. Datacenter revenue more than tripled year on year and is now the majority of the business; the AI optical cycle is showing up in reported numbers, not just slides.

  3. The company returned to GAAP profitability and Q4 guidance implies non-GAAP net income of up to $60.4 million, a sharp operating-leverage inflection if mix cooperates.

  4. Applied Optoelectronics has beaten the EPS estimate in each of the last 14 quarters, with an average surprise above 11%, according to GNG Research.

Bear Case

  1. Only $50–70 million of $400–450 million of 1.6T orders can ship in Q4; a deal that cannot be fulfilled is not yet revenue.

  2. Non-GAAP gross margin fell 230 basis points to 28.7% and missed guidance, with 400G mix the stated cause and a target of 40% still far away.

  3. The $550 million at-the-market facility, filed 17 November with Raymond James and Needham as sales agents at a 2% commission, is dilution into a share price that had run hard; the stock fell 5–8% after hours on the news.

  4. US capacity targets were cut because of power constraints, and phase two at Sugar Land slipped several months — execution risk is concrete, not hypothetical.

Figure 5 — Non-GAAP gross margin across three reported quarters: 31.0%, 29.8%, 28.7%. Sources: AAOI Q3 FY26 8-K and Yahoo Finance/Zacks, 5 Nov 2026.

Strategic Slam

$AAOI(Applied Optoelectronics)$ is not a broken story; it is a timing story with a financing tax. The bull case rests on a signed, named 1.6T contract and a datacenter franchise that tripled. The bear case rests on the fact that DSP supply, not customer demand, decides how much of that can be invoiced this quarter, and on a margin mix that gets worse before it gets better. The $550 million ATM is management's admission that it needs capital to bridge the gap. For $NVIDIA(NVDA)$ supply-chain watchers, Applied Optoelectronics is a real, sourced read on AI optical demand — just not a clean one. Peers such as $Coherent(COHR)$ and $Lumentum(LITE)$ rallied on the same optical optimism on 17 November, but AAOI is the one carrying the dilution and the DSP constraint.

Figure 6 — A 2×2 of supply resolution versus hyperscaler ramp, mapping what confirms or breaks the thesis. Sources: AAOI Q3 FY26 call and 8-K, 5 Nov 2026; author's scenario framing.

Catalyst Calendar

  • 5 Nov 2026 — Q3 results: $269.9M revenue, first GAAP profit, 1.6T contract disclosed.

  • 17 Nov 2026 — $550M at-the-market equity facility filed; shares fell after hours.

  • Dec 2026 (expected) — Sugar Land phase one begins operation.

  • Feb 2027 (expected) — Q4 results: whether 1.6T shipments reached the $50–70M guide.

  • End-2027 (target) — capacity above one million 800G/1.6T units per month.

Figure 7 — Dated catalysts from Q3 results through the 2027 capacity target. Sources: AAOI Q3 FY26 call, 5 Nov 2026; ATM filing, 17 Nov 2026.

Verdict

The signed deal is real and the demand is real, but the market is being asked to pay today for revenue that DSP supply defers. Until 1.6T shipments actually print, Applied Optoelectronics is a capacity-and-margin story dressed as an AI-demand story, with a $550 million dilution overhang on top. I would wait for one quarter of shipped 1.6T revenue and a margin that turns back toward 30% before treating the contract as earnings. The stock's next move will be decided by a supplier's delivery schedule, not by another record revenue headline.

Key Takeaways

  1. Q3 revenue was a record $269.9 million, up 127.5%, with the first GAAP profit in years.

  2. A $1.0–1.5 billion, three-year 1.6T deal with a Tier 1 hyperscaler was disclosed.

  3. Only $50–70 million of $400–450 million of 1.6T orders can ship in Q4 — DSP supply is the binding constraint.

  4. Non-GAAP gross margin fell 230 basis points to 28.7% on a worse 400G mix; Q4 margin is guided 29.5–31%.

  5. A $550 million ATM was filed on 17 November, and US capacity targets were cut on power delays.

If a company holds more than $400 million of orders it cannot yet ship, is that a demand story or a supply story — and which one are you paying for?

Tickers: $AAOI(Applied Optoelectronics)$ $NVDA(Nvidia)$ $COHR(Coherent)$ $LITE(Lumentum)$

Sources: Applied Optoelectronics Q3 FY26 results, Form 8-K Exhibit 99.1, 5 Nov 2026; AAOI Q3 FY26 earnings call transcript via BigGo, 5 Nov 2026; Yahoo Finance and Zacks Q3 earnings coverage, 5 Nov 2026; GNG Research AAOI earnings data; Applied Optoelectronics at-the-market equity distribution agreement filed 17 Nov 2026 via Blotter and Finwire; Finwire after-hours movers, 17 Nov 2026; Tencent gu.qq AAOI price history, 17 Nov 2026.

Disclaimer: The author holds no position in any security mentioned and receives no compensation for this article. This is analysis, not investment advice. Do your own research before making any investment decision.

# 💰Stocks to watch today?(16 November)

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Report

Comment(2)

  • Top
  • Latest
  • BirdieO
    ·11-18 11:23
    TOP
    AAOI. 425M sitting there with only 50 to 70M shippable says demand is real. The real bet is whether capacity ramps before that ATM overhang drags sentiment
    Reply
    Report
    Fold Replies
    • Xaddy_Analyst: 
      Exactly the right tension — the demand is signed, but the DSP bottleneck decides the timing. If Sugar Land phase one lands on schedule, that $50–70M cap should start lifting, and the ATM is really just buying time to get there.
      11-18 20:09
      Reply
      Report