Written by: U.S. Stock Kitten

The most popular new stock god Serenity with a 45x return in the past six months, Kitten has researched and distilled all his tweets, organizing a model consistent with his research analysis methods to filter out the next 10x stock from the entire market! If you missed Serenity's previous recommendations, no worries, subscribe to Kitten's membership, and Kitten will tell you the next three 10x stocks!

During the earnings call on May 7, 2026, Applied Optoelectronics (AAOI) — one of the few publicly traded suppliers of optical transceivers with its own indium phosphide (InP) laser wafer fab in the U.S. — revealed something I was eager to hear during the demand cycle. Capacity is the key constraint limiting 2026 revenue, while demand far exceeds what the company can ship. CEO Thompson Lin stated that the demand for 2026 is about $1.4 billion to $1.5 billion, while the company's FY2026 revenue guidance is around $1.1 billion.

It's important to clarify that in optical theme investments, AAOI is by no means a deeply valued stock that has yet to be discovered by the market. 【Insert image here】As shown in the guidance terminal, the stock has already experienced a sharp increase, rising from less than $2 at the beginning of 2023 to a high of $233.67 during intraday trading on May 13, 2026, becoming an extremely rare hundredfold stock in just three years. After the latest earnings report was released, there has been a divergence of views from sell-side firms. Rosenblatt listed AAOI as a preferred target on May 8 and raised the target price to $220. However, B. Riley took a different stance, raising its target price from $54 to $129 after the Q1 earnings report but maintaining a "neutral" rating, citing that the ramp-up of 800G capacity is mainly concentrated in the second half of the year, while the company relies on client forecasts.

I am not surprised at all. This is a stock that has increased by 420% year-to-date. In addition, insider selling has concentrated during the stock's significant rise, although some of it seems to be related to Rule 10b5-1 trading plans or vesting of equity, rather than concentrated sell-offs following the surge in stock price. 【Insert image here】Moreover, let’s not forget that on May 14, the company announced a new $600 million ATM issuance program with Raymond James and Needham. It is essential to clarify that the company is not obligated to sell shares, but this mechanism provides management with the opportunity to issue equity at a favorable time, especially given the current high stock price. So far, this sounds like a set of bearish logic. Let me be clear: I am by no means a bear on AAOI.

Why? Between March and early April 2026, AAOI announced new 800G and 1.6T orders exceeding $324 million related to major hyperscale customers. You might say I am crazy, but it’s hard to reconcile this with judgment about demand slowing. The market still lacks high-speed optical products, and I do not believe this bottleneck will be resolved quickly. I hold a mid-single-digit position in my portfolio. What I will evaluate next is whether this supply-constrained pattern is structural and whether the capacity ramp-up in the second half of 2026 can effectively convert capacity guidance into revenue.

Understanding the Bullish Logic

What sets Applied Optoelectronics apart from most U.S. public optical companies is that it sits at both ends of the bottleneck. AAOI produces its own InP laser chips in Sugar Land, Texas, which are then used in the company’s own high-speed optical transceivers. This gives AAOI stronger control over supply, costs, certification, and ramp-up time compared to other module manufacturers like Innolight, Eoptolink, or Accelink, who must procure lasers from a relatively limited pool of suppliers, including Lumentum, Coherent, Mitsubishi Electric, Sumitomo Electric, and Broadcom.

Companies like Lumentum, Coherent, Mitsubishi, Sumitomo, and Broadcom are EML manufacturers, which means they control the upstream laser chip layer. In simpler terms, these companies are suppliers. AAOI also produces EML, so just looking at that alone isn’t particularly interesting. What is intriguing about AAOI is that it is both a supplier in this bottleneck segment while also being an internal consumer, meaning the company uses its own produced InP laser chips to manufacture its finished optical transceivers. This detail is crucial. Why is this noteworthy? First, the company is not very well known among analyst circles. For example, in December last year, TrendForce published an article discussing the upcoming wave of laser shortages, and notably left Applied Optoelectronics out of the EML supplier list. 【Insert image here】TrendForce

Regarding the optical bottleneck, I have discussed my opinions in a recent report covering the broader photonics theme. In short, the indium phosphide shortage, caused by AI data center construction and limiting the production of optical lasers and transceivers, is structural rather than cyclical. At the analyst level, some sources predict that by 2027, the output of 800G optical transceivers will be 40% to 60% below demand; by 2029, the output of 1.6T optical transceivers will be 30% to 40% below demand, driven primarily by the scarcity of EML.

NVIDIA actually confirmed this on March 2026 when the company committed to provide Lumentum and Coherent with a total of $4 billion in investments and multi-year procurement commitments specifically for laser supply. As for AAOI, since March 2026, the company has already secured over $324 million in new 800G and 1.6T orders: March 9, 2026: AAOI announced it received a bulk order of over $200 million for 1.6T from a major hyperscale customer it has been working with for a long time. March 23, 2026: AAOI revealed it received an order of over $53 million for 800G from the same major hyperscale customer. April 2, 2026: AAOI announced it further received $71 million in 800G orders from that customer, bringing the total 800G orders from that client since mid-March to $124 million.

To me, this indicates that order momentum has not yet peaked. Management stated in the Q1 call that the demand for 2026 is around $1.4 billion to $1.5 billion, but supply chain constraints mean that the scale of what the company can ship is about $1.1 billion. This leads to my fundamental logic.

Fundamental Logic

Given that products are transitioning from 800G to 1.6T, the certification window is currently open, and AAOI is one of the three major U.S. domestic suppliers of optical transceivers participating in this competition, the other two being Coherent and Lumentum. When it comes to 1.6T, AAOI secured its first bulk order for 1.6T data center optical transceivers in March 2026. This is an order from a major hyperscale customer exceeding $200 million, which is expected to start shipping in early Q3 2026 once product certification is completed.

【Insert image here】In Q1 2026, AAOI's 800G revenue was only $4.6 million, accounting for 5.6% of data center revenue, as AAOI was still in the initial stages of ramping up its 800G capacity. Based on my reading of the earnings call transcript, AAOI has begun to ship in bulk, but revenue conversion is still hampered by manufacturing cycles, customer audits, certification requirements, and the scheduling of new installed capacity coming online. Management anticipates that Q2 800G shipments will be nearly four times the level of Q1, with 1.6T products starting to deliver as early as Q3, while the announced 800G/1.6T deliveries are expected to be completed by the end of the year.

To meet the company's goal of over $1.1 billion in annual revenue, I estimate that Q3 and Q4 need to contribute about $751 million to $769 million in revenue, depending on where Q2 revenue ultimately falls within the guidance range of $180 million to $198 million. This means that the revenue structure in the second half of the year will be more aligned with Q3 achieving about $340 million to $360 million, Q4 achieving about $410 million to $430 million, or alternatively structured in other ways that allow second-half revenue to exceed approximately $760 million.

This would imply a quarter-over-quarter growth of about $140 million to $180 million from Q2 to Q3, depending on where Q2 ultimately lands within the guidance range, roughly equivalent to the full Q1 revenue base. Naturally, most of this growth will be driven by the data center division. 【Insert image here】On the production side, AAOI had about 100,000 units per month of 800G/1.6T product capacity at the end of Q1, expected to reach about 150,000 units per month in Q2, and aims to exceed 650,000 units per month by the end of 2026. On the InP laser manufacturing side, the company is expanding capacity by roughly 350%, targeting completion by 2027 and transitioning from 4-inch wafers to 6-inch wafers.

In terms of margin, Q1 Non-GAAP gross margin was 29.2%, down from 31.4% in Q4 2025, due to the product mix being in a transition phase. Lin provided guidance during the Q1 call that by the end of 2026, the gross margin is expected to reach 35%, and by Q3/Q4 2027, it will exceed "40%," explicitly referencing "especially driven by the laser business." Naturally, this is where the bullish logic of the company really lies.

Risks

Chinese module suppliers like Innolight and Eoptolink have already secured a dominant share of AI-related 800G demand, with some sources estimating that together they supply about 60% of NVIDIA's 800G optical module demand. Meanwhile, Chinese module suppliers reportedly price comparable optical transceiver products about 20% to 25% lower than existing Western suppliers. Bears might argue that as Chinese EML capacity expands, AAOI's moat will narrow.

This argument has its merits. That said, I believe the bears underestimate the fact that the transition to 1.6T will reset the certification cycle. In my view, hyperscale customers will not simply purchase the lowest-cost modules on the market. They will certify specific suppliers and specific factories. I believe this could create an entry point for AAOI, as the company is building a supply system reducing the components sourced from China at the same time customers are re-certifying the next generation of optical products.

Specifically, AAOI disclosed during the Q4 2025 call that the component value from China in its 800G and 1.6T designs is currently less than 10%, and there exists a pathway to reduce this exposure to nearly zero. In terms of valuation, this is an expensive stock, currently trading at over 200 times forward earnings. Any underperformance at the execution level could lead to a downward reevaluation of the stock price, especially considering the stock has already risen over 400% year-to-date and nearly 1,000% over the past 12 months.

Looking Ahead

I believe the Q3 2026 earnings report will be a significant test to validate or break the bullish logic. Three core clues will converge in the same quarter: 800G shipments ramping up to about four times the Q1 level; the first deliveries of 1.6T after hyperscale customer certification is completed; and gross margins moving from 29.2% towards Lin's promised end-of-year target of 35%. In my view, if the company can cleanly achieve over $340 million in revenue while gross margins reach just over 30%, it would be extremely favorable for the stock price. Additionally, there is a factor in the bullish logic that I have not mentioned before: ELSFP, or External Laser Source Form Factor Pluggable. To put it plainly, this is a replaceable module that places the laser outside of the switching chip and delivers the light source to a co-packaged optical (CPO) engine. In March of this year, AAOI showcased a 25 dBm external laser source module.

【Insert image here】AAOI Why am I bringing this up? Because Lumentum received a $2 billion investment from NVIDIA in March 2026 specifically for CPO laser modules, and one competitive ELSFP product produced by Lumentum is rated at 24 dBm. 【Insert image here】Lumentum Thus, the product showcased by AAOI is 1 decibel higher. However, that $2 billion was given to Lumentum. It is important to clarify that Lumentum has NVIDIA's preferred position for CPO laser supply; I am not suggesting that this will change. What I mean is that AAOI may have the opportunity to become a second source for Nvidia. In fact, there is evidence that AAOI may be negotiating for a potential supply relationship.

Management disclosed during the Q1 2026 call that the company is collaborating with two unnamed large customers on ELSFP, and CEO Thompson Lin stated that the company is signing long-term agreements with multiple customers for products including lasers and ELSFP over a three-year term. Interestingly, according to CFO Stefan Murry’s comments during the Q1 call, management has provided a capacity target: by the end of 2027, ELSFP capacity will reach approximately 400,000 units per month.

I can't help but wonder how the stock price would react if Nvidia announced a strategic investment in AAOI. That said, the probability of this announcement happening is low; but if it does, the upside potential would be substantial. On the risk side, the $600 million ATM issuance program submitted on May 14 has the opposite effect. The Q2 10-Q document will show how many shares the company issued, if any, and what the aggregate issuance price is.

Overall, I currently hold a mid-single-digit position. If Q2 confirms that the 800G capacity ramp-up is proceeding as planned, I would be willing to increase my position during the Q3 earnings report process. The situation that would force me to exit would be if the 1.6T certification is delayed until after Q4 or if Q2 gross margins fall below the guidance range of 29% to 30%.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。