Coherent’s beat is not the story; the repricing risk is that 1.6T and InP supply are turning a low-quality recovery into an optical-cycle margin reset

The print matters because the setup was not for a demand shock; it was for confirmation that Coherent could keep converting AI optical demand into earnings without letting capacity, pricing, or legacy mix dilute the recovery. What was priced in was revenue of $1,554.2 million and EPS of $1.04, which implied investors already expected another sequential step after $1,529.4 million in Q4 FY2025 and a gross margin base that had recovered to 36.6%. What actually surprised was not the top line alone: revenue of $1,581.4 million was only +1.7% ahead of the Street, while EPS of $1.16 was +11.5% ahead. That spread between the modest sales beat and the larger earnings beat is the core of the investment debate. The market can dismiss +1.7% as normal AI optics noise; it should not dismiss +11.5% in EPS when the company is also pointing to 1.6T-driven growth in the current quarter and talking about capacity beyond the next 12 months.

The revenue trajectory is now clean enough to separate cyclical recovery from structural mix change. Coherent’s quarterly revenue bottomed at $1,053.1 million in Q1 FY2024, then climbed to $1,581.4 million in Q1 FY2026, with the most recent quarter showing +3.4% sequential growth and +17.3% year-over-year growth. That is not a single-quarter snapback: the year-over-year stack moved from +28.0% in Q1 FY2025 to +26.8% in Q2 FY2025, +23.9% in Q3 FY2025, +16.4% in Q4 FY2025, and +17.3% in Q1 FY2026. The market may have anchored on decelerating year-over-year growth from Q1 FY2025 through Q4 FY2025, but the current print interrupts that narrative and does so while the company’s reported gross margin held at 36.6%. The quality of growth is therefore improving: revenue is above estimate by $1,581.4 million versus $1,554.2 million, EPS is above estimate by $1.16 versus $1.04, and the company is no longer buying revenue with gross margin erosion.

The margin bridge is the reason the beat deserves a higher multiple than a normal optical-components upside quarter. Gross margin was 29.1% in Q1 FY2024, 34.1% in Q1 FY2025, and 36.6% in Q1 FY2026; the company’s own call basis shows non-GAAP gross margin at 38.7%, and CFO Sherri Luther was explicit that the path is not just mix luck: “I am especially pleased with the progress we have made on gross margin expansion, driven by the cost reduction and pricing optimization initiatives that we continue to focus on as we drive to our target model of greater than 42%.” The wording matters because it ties the margin lift to controllable levers rather than simply to hyperscale demand, and it puts a numerical marker on management accountability. If investors were underwriting Coherent as a revenue beta to AI networking, the print argues for a different framing: this is a margin self-help story being accelerated by AI optical mix.

That interpretation is reinforced by the expense and leverage data, which explain why a +1.7% revenue surprise became a +11.5% EPS surprise. On the company’s call basis, non-GAAP operating expenses were $304 million compared to $307 million in the prior quarter, while operating expenses as a percentage of revenue declined to 19.2% from 20.1% in the prior quarter and 20.6% in the year ago quarter. Non-GAAP operating margin reached 19.5%, compared with 18% in the prior quarter and 16.1% in the year ago quarter. The market was right to expect an improving revenue base, but it appears to have under-modeled how much of that growth would drop through once cost reduction, pricing optimization, and mix all moved in the same direction. Luther’s debt commentary adds a second layer to EPS durability: the company paid down $400 million in debt, reduced the debt leverage ratio to 1.7x from 2.4x in the year ago quarter, refinanced debt with a 60 basis point lower interest rate, and doubled the revolving credit facility to $700 million. That does not change optical demand, but it reduces the financing overhang that previously made gross margin improvement feel fragile.

The demand signal is most investable in Datacenter and Communications because management gave enough segment color to make the next-quarter test concrete. CEO James Anderson said Datacenter and Communications, “our largest and fastest-growing business,” grew by 7% sequentially and 26% year-over-year in Q1, and then added that data center growth should accelerate to approximately 10% sequential growth in the current quarter. The important point is not merely that AI data center is growing; it is that the sequential growth rate management expects for the current quarter is higher than the segment’s 7% in Q1, and management identified 1.6T adoption as a significant portion of that sequential growth. In this subsector, 1.6T is not a product footnote. It changes the mix toward higher-speed optics, consumes advanced laser and InP capacity, and can tighten the supply chain at exactly the time Coherent is trying to expand gross margin. When Anderson says the 1.6T product line adds “over $2 billion of addressable market opportunity over the coming years,” the phrase matters because it extends the debate beyond Q2 guide mechanics and into how much of the company’s margin target can be pulled forward by portfolio mix.

The pro forma revenue disclosure is also a cleaner signal than headline revenue because it strips out a lower-margin drag that investors may not fully adjust for. Luther said that excluding $33 million of Aerospace and Defense revenue for Q1, revenue increased 6% sequentially and 19% year-over-year, and separately noted that over the past 4 quarters this business contributed average quarterly revenue of $25 million with a gross margin well below Coherent’s corporate gross margin. That disclosure changes how to read the 36.6% reported gross margin and the 38.7% non-GAAP gross margin. The company is not just scaling into better utilization; it is also removing or deemphasizing revenue that sits below the corporate margin line. If the Street focuses on $1,581.4 million versus $1,554.2 million, it misses the more important mix point: the higher-quality revenue base, excluding $33 million of Aerospace and Defense, grew 6% sequentially and 19% year-over-year, while operating expenses were held to $304 million. That is the setup for positive earnings revisions even if headline revenue beats remain modest.

The capacity story explains why the read-through extends beyond Coherent and why suppliers matter more after this print than they did before it. Anderson said that in Communications, Coherent has seen 5 quarters of sequential growth, including 11% sequential and 55% year-over-year last quarter, and he tied demand visibility not only to next calendar year but to ’27 and ’28, with a plan to keep ramping indium phosphide capacity beyond the next 12 months. For AXT Inc, which supplies InP substrates and compound semi wafers, the direct implication is that Coherent is not treating InP demand as a one-year capacity bridge; “beyond the next 12 months” points to a multi-year substrate pull. For 5N Plus, the read-through is tied to high-purity semiconductor metals including Ge, In, Se, and Te, with In demand specifically supported by the indium phosphide expansion. For Fabrinet, the contract manufacturing read-through is more mixed: higher optical transceiver demand helps volumes, but the value capture depends on whether 1.6T ramps increase manufacturing complexity without pressuring outsourced assembly economics. On the customer side, Zhongji Innolight’s exposure to EML and CW laser chips for transceivers is validated by Coherent’s 1.6T commentary; NVIDIA and Broadcom, both tied in the data pack to optical or SiPho components for CPO, get a supply-chain confirmation that advanced optics capacity is being expanded in response to demand visibility into ’27 and ’28 rather than only current-quarter orders.

The peer comparison keeps the stock-specific conclusion honest: Coherent is not the fastest-growing optics name, but it is now pairing scale with margin improvement in a way that should matter for PMs sizing exposure. In the latest peer table, Coherent revenue is $1,805.6 million with gross margin of 37.7% and revenue YoY of +20.5%. LITE shows $808.4 million of revenue, 42.4% gross margin, and +90.1% revenue YoY, while AAOI shows $151.1 million, 29.1% gross margin, and +51.4% revenue YoY, and LASR shows $80.2 million, 33.1% gross margin, and +55.2% revenue YoY. The right comparative point is not that Coherent has the highest growth; it does not. It is that Coherent has far larger revenue scale than LITE, AAOI, and LASR in the table while moving gross margin toward the higher-quality end of the group. Against Japanese photonics peers, 6925.T at 36.2% gross margin and +5.5% revenue YoY is the closer margin comp, but Coherent’s +20.5% revenue YoY in the peer table is a different growth profile. The market may still group Coherent with slower legacy photonics, while the data increasingly put it between high-growth optical AI names and scaled diversified components.

The call tone supports the thesis, but the nuance is that management sounded more directed in guidance than universally more upbeat. The tone history shows Q1 FY2026 sentiment of 0.41, guidance_tone of 0.58, tone_confidence of 0.30, prepared_sentiment of 0.61, qa_sentiment of 0.30, ai_optimism of 0.67, uncertainty of 42.1, and qa_evasiveness of 18.0. Compared with Q4 FY2025, guidance_tone rose from 0.44 to 0.58, ai_optimism rose from 0.60 to 0.67, and uncertainty fell from 64.3 to 42.1, even though qa_evasiveness rose from 9.2 to 18.0. That mix is important: prepared remarks and guidance became more assertive, while Q&A still contained enough friction to keep the stock from being priced as a frictionless AI optics story. The later tone history also shows that by Q3 FY2026, guidance_tone reached 0.59, uncertainty was 45.9, and qa_evasiveness was 6.4, with call-over-call delta showing guidance_tone +0.07, uncertainty -6.4, and qa_evasiveness -39.4. Delivery is not the thesis, but it corroborates it: management’s language around guidance and AI demand became more precise at the same time the financials showed operating leverage.

The guide is where the variant perception should either earn its keep or get falsified. For the current quarter, Luther guided revenue between $1.56 billion and $1.7 billion, non-GAAP gross margin between 38% and 40%, operating expenses between $300 million and $320 million, tax rate between 18% and 20%, and EPS between $1.10 and $1.30. That range embeds the key tension. The low end of revenue, $1.56 billion, would sit below Q1’s $1.581.4 million print basis, while the high end of $1.7 billion would confirm that the approximately 10% sequential data center growth and 1.6T adoption are translating into consolidated upside. The gross margin range is the cleaner test: after Q1 non-GAAP gross margin of 38.7%, the midpoint logic of 38% to 40% leaves room for sustained progress toward greater than 42%, but a print at 38% would weaken the margin-reset argument. EPS guidance of $1.10 to $1.30 also tells investors that the $1.16 quarter was not an isolated beat unless revenue lands near the low end and expenses push toward $320 million.

The practical stance is to treat this event as a positive revision to earnings quality, not just to revenue expectations. What was priced in was a company participating in AI data center demand with $1,554.2 million of expected revenue and $1.04 of expected EPS. What surprised was the conversion of $1,581.4 million of revenue into $1.16 of EPS, helped by non-GAAP gross margin of 38.7%, operating expenses of $304 million, and operating margin of 19.5%. The next confirmation point is the next quarter’s revenue range of $1.56 billion to $1.7 billion, with particular focus on whether management delivers the approximately 10% sequential data center growth it described, whether non-GAAP gross margin stays inside 38% to 40% rather than slipping to the bottom, and whether EPS holds within $1.10 to $1.30 with operating expenses no worse than $320 million. The thesis breaks if 1.6T adoption fails to show up in the current-quarter data center acceleration, if gross margin falls back toward 38% despite the low-margin Aerospace and Defense context, or if Q&A evasiveness moves back toward 45.8 rather than the later 6.4 reading in the tone history. It is confirmed if revenue tracks toward $1.7 billion, non-GAAP gross margin prints near 40%, and management keeps framing indium phosphide capacity as extending beyond the next 12 months rather than as a short-cycle fix.