Business Overview: Coherent Corp.

NYSE: COHR · Fiscal year ended 30 June 2026

7 September 2026

Evidence base: Coherent’s own filings from the provided research folder — annual reports FY2016 through FY2026, quarterly reports through FY2026 Q3, current reports and proxy statements through August 2026 — supplemented by independent industry and competitor sources from the web, including competitors’ own filings, LightCounting and Dell’Oro market data, USGS mineral statistics and the Federal Register. Coherent’s FY2026 fourth-quarter earnings press release and earnings call, which the research folder does not contain, were taken from the company’s own published release of 12 August 2026 and are cited as such.

This is a business overview for investment research. It is not a valuation and not a recommendation. It contains no price target, no multiple and no view on the shares.

1. Executive Snapshot

2. What the Company Does

The customer problem is physical. Inside an AI datacentre, the chips that do the computing communicate electrically, but an electrical signal degrades over distance: above roughly one to two metres at modern per-lane speeds, copper stops working. Beyond that distance the signal has to become light, travel down a fibre, and become electricity again at the other end. Somebody has to build the device that performs that conversion, tens of millions of times over, reliably, in a package the size of a thumb that draws as little power as possible. That device is an optical transceiver, and Coherent is one of a small number of companies that can make one from its own raw materials upward.

The unit of this business is one high-speed optical transceiver. Upstream of it sits the unit that actually governs Coherent’s economics: one 6-inch indium phosphide wafer, and the laser dies cut from it. A transceiver is largely a housing for a laser, a modulator, a detector and a signal-processing chip; the laser is the part that is hard to make and currently impossible to buy in sufficient quantity. Coherent grows the indium phosphide crystal, processes the wafer, fabricates the laser die, packages it into a module, and ships the module to a customer who pays in 30 to 90 days (FY2026 10-K, Note 1). Revenue is recognised at a point in time, generally on shipment. There is no subscription, no maintenance annuity of consequence and no percentage-of-completion accounting: contract liabilities stood at $63M against $7,118M of revenue. This is a manufacturing business that sells goods outright and gets paid once.

Coherent does not disclose how many units it ships or what they sell for (NOT FOUND). The 10-K states plainly that revenue is not disaggregated by product type because the chief operating decision maker does not use that view (FY2026 10-K, Note 3). For a sense of scale, the nearest disclosed unit economics in the industry come from a competitor: Zhongji Innolight sold 21.09 million optical modules for RMB 37,457M in calendar 2025, an average selling price of roughly RMB 1,776, or about $248 per module (Innolight 2025 annual report). That figure describes the industry’s unit, not Coherent’s realisation.

Two segments, and what each sells

Effective 1 July 2025 Coherent collapsed three reporting segments — Networking, Materials and Lasers — into two, Datacenter & Communications and Industrial, recasting prior periods (FY2026 10-K, Note 20). The two reportable segments are stated to be aggregations of an undisclosed larger number of operating segments.

• Datacenter & Communications — $5,275M of FY2026 revenue, 74.1% of the total, up 40% year on year. It sells 800G and 1.6T datacom transceivers, co-packaged optics, optical circuit switches, ZR and ZR+ coherent transceivers for linking datacentres to one another, and the merchant components underneath all of them — VCSELs, EMLs, InP lasers, photodiodes, pump lasers, isolators and photonic integrated circuits. Segment profit was $1,330M, 25.2% of segment revenue.
• Industrial — $1,844M, 25.9% of revenue, down 10%. Excimer and CO₂ lasers for display annealing and wafer processing, ultrafast lasers for semiconductor inspection and advanced packaging, high-power lasers for materials processing, and engineered materials including ZnSe, ZnS, GaAs, GaSb and silicon carbide substrates. Segment profit was $423M, 22.9% of segment revenue.

Segment profit is the company’s own measure and it is generous: it excludes share-based compensation, acquisition and integration costs, amortisation and impairment of acquired intangibles, restructuring, held-for-sale impairment, gains on business sales, and unallocated corporate research and administration. The gap between the sum of segment profit ($1,752M) and pre-tax earnings ($848M) is $905M, of which intangible amortisation is $280M, corporate cost $297M, share-based compensation $194M and interest $190M. Segment margins should not be read as operating margins.

What is being exited, and why the economics differ

Coherent has been shrinking the Industrial side deliberately. It sold its aerospace and defence business to Advent International on 2 September 2025 for approximately $400M, booking a $115M gain, and sold a Munich materials-processing tools division on 30 January 2026 at a net loss of $96M once $105M of held-for-sale impairments are counted (FY2026 10-K, Note 7; Coherent press release, 13 August 2025). A third Industrial entity was classified held-for-sale in the fourth quarter of FY2026, with no buyer, price or timing disclosed. The company attributes the whole 10% Industrial revenue decline to these disposals, and says both gross margin and segment profit in Industrial benefited from them — which tells you the businesses sold earned less than the ones retained.

One thing has not been sold, and it is widely misunderstood. Coherent still owns roughly 75% of its silicon carbide business. In December 2023 Denso and Mitsubishi Electric each bought a 12.5% stake in Silicon Carbide LLC for $500M, valuing the whole at an implied $4.0bn (Coherent press releases, 10 October and 4 December 2023). The proceeds are ring-fenced: $604M of restricted cash sits inside that subsidiary and can be used only by it (FY2026 10-K, Note 15). The subsidiary is loss-making — the 25% minority absorbed $18.1M of losses in FY2026, implying roughly $72M at the whole-entity level. Silicon carbide remains a product line, a use of capital, and no longer a separately reported one.

3. Industry, Competitive Position & Moat

The optical networking industry sells electrical-to-optical conversion, and the striking fact about it is where the money sits. Reading across the reported results of every listed participant in the chain, gross margin is highest at the two ends and lowest in the middle. Broadcom, which designs the switch silicon and the digital signal processors and owns no fabs, earned a 67.8% GAAP gross margin and a 39.9% operating margin in its fiscal 2025 (Broadcom FY2025 release, 11 December 2025). Fabrinet, which assembles modules to order in Thailand, earned 12.0% gross and 10.0% operating on $4,641M of FY2026 revenue (Fabrinet FY2026 release, 17 August 2026). Coherent, which owns the most of the chain of anyone, earned 37.5% gross and 12.6% operating.

That last comparison is the one that matters, and it does not flatter vertical integration. Fabrinet converts a 12.0% gross margin into a 10.0% operating margin because its operating costs are about two points of sales. Coherent converts 37.5% into 12.6% because its operating costs are roughly twenty-five points of sales, and it spends 15.5% of revenue on capital equipment while Fabrinet builds sheds. On the operating line the contract manufacturer is within three points of the vertically integrated technology owner, on a fraction of the capital and none of the technology risk.

Who is actually winning

The two Chinese module assemblers are winning, and they own neither wafers nor chips. Zhongji Innolight reported calendar-2025 revenue of RMB 38,240M, up 60.3%, at a 42.6% gross margin and a 28.2% net margin; Eoptolink reported RMB 24,842M, up 187.3%, at a 47.8% gross margin and a 38.4% net margin (2025 annual reports filed on the Shenzhen exchange). Both figures exceed Coherent’s 37.5% gross and 11.3% net. Innolight’s first-half 2026 revenue of RMB 41,778M was larger than its entire 2025. On LightCounting’s vendor tracking, Innolight booked $1.87bn in the fourth calendar quarter of 2025 against Coherent’s $1.2bn of datacentre and communications revenue in the same quarter (LightCounting Quarterly Market Update, March 2026). Innolight and Eoptolink between them are reported to hold more than 60% of modules at 800G and above, and Chinese firms hold seven of the top ten positions in the global module market (Caixin Global, 5 August 2026, reporting a drafted US restriction).

Coherent’s revenue is larger than either, because Coherent sells components, systems and materials as well as modules. But on the specific question of who earns the most per dollar of module revenue, the companies with the least technology in the chain are currently earning the most.

Which barriers actually bind

One barrier genuinely binds today: qualified indium phosphide epitaxy capacity. Coherent’s chief executive stated on the FY2026 fourth-quarter call that “indium phosphide capacity continues to be our primary constraint” while transceiver assembly and test capacity is not constrained (Coherent FY26 Q4 call, 12 August 2026). Lumentum independently described itself as “effectively sold out for the foreseeable future despite our rapid capacity expansion” (Lumentum FY26 Q4 call, August 2026). Two competing issuers describing the same shortage is strong evidence the shortage is real. The upstream metal is constrained too: China refines about 70% of world indium production, imposed export controls in February 2025, and unwrought indium exports fell 72% year on year in the twelve months to September 2025, with the price rising from $351 to $370 per kilogram (USGS Mineral Commodity Summaries 2026, indium chapter).

A second barrier is measurable and is Coherent’s own: the 6-inch wafer learning curve. The company claims 6-inch InP delivers four times the production capacity and a 60% reduction in die cost against its legacy 3-inch lines, and the chief financial officer stated that 6-inch yields already exceed 3-inch yields across all product devices with the bulk of the ramp still ahead (Coherent FY26 Q4 call). The US Commerce Department describes the Sherman, Texas plant as the first and largest high-volume 150mm indium phosphide facility, and issued a letter of intent for up to $50M of CHIPS Act support in June 2026 (NIST, 16 June 2026).

The barriers that do not bind are the ones most often cited. The claim that Chinese vendors cannot build high-speed optics has been falsified within four years: LightCounting judged them two to three years behind on 100G-per-lane components in January 2024, and by 2026 they held the majority of the 800G-and-above market. Hyperscaler qualification slows share shifts by quarters, not years — Applied Optoelectronics announced a $200M-plus 1.6T order in March 2026 with volume shipment beginning that summer. And capital intensity, which looks like a barrier, is only a barrier when capital is scarce; Coherent raised $2.0bn of equity from a single buyer in a single day in March 2026.

Testing the company’s account of itself against outside evidence

Coherent makes four claims about its position. Outside evidence supports one and a half of them.

• Vertical integration into InP is a differentiator — partially supported. The rarity is real: Lumentum buys InP substrate externally and said so on its own call, and Innolight derives 98% of revenue from module assembly with no disclosed substrate or epitaxy operation. Coherent appears to be the only participant owning substrate through module. But the margin evidence contradicts the value of it. The most integrated company in the set earns a lower gross margin than the two least integrated, and the two other chip-owning integrators, Accelink and Ligent, earn the worst margins of any branded module maker — Ligent’s own listing prospectus discloses an optical chip segment running a negative 121% gross margin in 2025 because volume did not cover fixed cost. Integration buys supply access, not margin.
• Its capacity is scarce — supported, but the scarcity is the industry’s, not Coherent’s. Every participant is adding capacity into the same shortage simultaneously: Coherent doubled internal InP output a quarter ahead of plan and intends to more than double it again by the end of calendar 2027; Lumentum is expanding two Japanese InP fabs and converting a US fab; Applied Optoelectronics targets 650,000 modules a month by the end of 2026; Fabrinet is building a two-million-square-foot facility adding $3.0–3.5bn of revenue capacity. LightCounting expected the shortage to ease from mid-2026.
• It has pricing power in AI datacom — contradicted. Coherent’s own risk factor reads: “our large customers have in the past sought price concessions from us, and we expect that they will continue to do so in the future”. Its defence is long-term agreements, but the chief executive described those agreements as carrying “agreed upon pricing” running three years and in some cases through the rest of the decade — which means Coherent cannot reprice upward into a shortage. It kept the volume and gave away the scarcity rent. The realised numbers say the same thing: Coherent’s non-GAAP gross margin rose 152 basis points in FY2026 while Lumentum’s rose 1,130 and Innolight’s 796.
• The materials business is a moat — contradicted, and the disclosure has been withdrawn. Materials was the highest-margin segment Coherent reported in FY2024, at $1,017M of revenue and a 21.0% non-GAAP operating margin. It no longer exists as a reported segment; it now sits inside Industrial, which fell 10% in the strongest capital-spending year the industry has ever seen. A moat that shrinks in a boom, and whose disclosure is then removed, cannot be tested.

What could weaken the position

The most concrete threat is a customer moving upstream. At the OFC conference in March 2026 Broadcom announced not only Taurus, a 400G-per-lane optical DSP, but “first-to-market 400G electro-absorption modulated laser and photodiodes”, alongside 200G-per-lane VCSEL, EML and CPO technologies (Broadcom, 12 March 2026). Those are Coherent’s core product categories, and they are being entered by a company with a 67.8% gross margin that can fund the attempt out of petty cash. Broadcom does not disclose optical revenue, so the scale is unknown — but a fabless business with two-thirds gross margins subsidising entry into a capital-intensive one with a third is the textbook shape of a profit pool migrating.

On the evidence, the companies that win in this industry own a scarce input the buyer cannot second-source, or sit at a stage where their part is a small share of system cost but determines system performance, or are the low-cost producer at scale. Coherent qualifies on the first and is trying to qualify on the third. It does not qualify on the second, and LightCounting’s diagnosis of why explains most of the industry’s history: optics is “a tiny part of a massive industry” yet an “ever-larger portion of the bill of materials” — big enough to be worth squeezing, small enough to have no leverage. Coherent is a good version of a structurally difficult business.

4. Growth Engine

FY2026 revenue grew 22.5% to 7,118M. That figure is composed of Datacenter & Communications up 40% (1,519M) and Industrial down 10% ($211M). The Industrial decline is attributable primarily to the two divestitures, which means reported growth understates organic growth in FY2026 — the mirror image of the usual problem. Coherent gives no organic-versus-inorganic revenue bridge and does not disclose the revenue of the businesses sold (NOT FOUND). Removing an estimated Industrial base is therefore not possible from the filings.

The reverse distortion is present in the history and matters more. FY2023 revenue of $5,160M was 56% above FY2022, and essentially all of that came from the acquisition of Coherent, Inc., which closed on 1 July 2022 for total consideration of $7,096M and contributed $1,469M of revenue and a $412M net loss in its first year (FY2023 10-K, Note 3). The same is true of FY2020: Finisar closed in September 2019 for $2,909M and contributed $938M of revenue and a $94.6M net loss in nine months. Over the last decade Coherent’s revenue rose from $827M to $7,118M, but the diluted share count rose from 62.9 million to 195.4 million and total debt from $235M to $3,222M. A long-run revenue chart for this company is substantially a chart of acquisitions.

What is actually driving growth now, ranked

• 1. AI datacentre transceiver volume (structural in direction, cyclical in amplitude). Hyperscaler capital spending is the whole mechanism. Amazon’s trailing-twelve-month capital expenditure to June 2026 was $169.0bn, up 64%; Microsoft spent $115.9bn in its fiscal 2026, up 80%; Meta guided calendar 2026 to $130–145bn; Alphabet spent $80.6bn in the first half of 2026 alone (company earnings releases, 2026). LightCounting puts optics at 2.7% of top-five cloud capital spending in 2025, rising to 3.1% in 2026. The direction is structural because copper does not reach; the amplitude is a spending cycle set by four balance sheets.
• 2. Indium phosphide output and 6-inch yield (management-driven). Coherent produced roughly 80% more InP lasers in the June 2026 quarter than a year earlier and doubled internal capacity a quarter ahead of plan. This is the driver most within management’s control and the one that converts scarcity into revenue rather than into someone else’s revenue.
• 3. Mix shift from Industrial to Datacenter & Communications (management-driven). D&C went from 55.9% of revenue in FY2024 to 74.1% in FY2026, partly by growing and partly by selling the other side. Because D&C carries a 25.2% segment margin against Industrial’s 22.9%, and because Industrial carries 58% of the group’s depreciation and amortisation on 26% of revenue, the mix shift moves reported margin on its own.
• 4. Pricing optimisation (temporary). Management names pricing as one of five gross-margin drivers in FY2026, without quantifying it. Current pricing strength is a scarcity artefact; the last time this industry saw one, in 2016–17, price erosion resumed and then accelerated within two quarters of the shortage ending (LightCounting, May 2020). Section 6 sets out what that did to supplier margins.
• 5. Datacentre interconnect and traditional telecom (structural, early). Management cites elevated demand for new ZR and ZR+ transceivers plus sustained growth in telecom transport. Dell’Oro forecasts the optical transport equipment market to grow 16% in 2026 and exceed $18bn for the first time since 2000, and notes lengthening lead times.
• 6. Semiconductor capital equipment (cyclical). The one part of Industrial management describes as strong. It is a capital-equipment cycle and behaves like one.

The NVIDIA arrangement

On 2 March 2026 NVIDIA bought 7,788,161 Coherent shares at $256.80 for $2.0bn, with a six-month lock-up, alongside a multi-year strategic agreement carrying a “multi-billion-dollar purchase commitment”, future access and capacity rights, and access for NVIDIA to five additional Coherent product families related to co-packaged optics (8-K, 2 March 2026; FY2026 10-K). The dollar value of the purchase commitment is not disclosed (NOT FOUND). Coherent also entered a capacity agreement that “may require incremental investments in equipment, labor, and working capital to support future production volumes through 2030” and which “may result in material future cash requirements and could affect revenue concentration, gross margin, and capital expenditures as volumes ramp”. The equity funded the capital expenditure; the commitment is the reason for the capital expenditure. Whether the arrangement is worth more than it costs cannot be answered from the disclosure available.

5. Margin, Cash & Capital Allocation

Gross margin rose 233 basis points to 37.5% in FY2026. Management attributes it to five things: lower product input costs, better manufacturing cycle times, yield improvement in Datacenter & Communications, pricing optimisation, and lower intangible amortisation — with Industrial additionally helped by having divested two lower-margin businesses (FY2026 10-K, MD&A). None is quantified individually. Two of the five are mechanical rather than operational: intangible amortisation fell from $303M to $280M simply because acquired assets age, and the divestiture benefit is a subtraction of bad margin rather than an addition of good.

Below gross margin, the operating structure is fixed to an unusual degree, and the filing says so: “because certain of our sales, research and development, and internal manufacturing overhead expenses are relatively fixed, a reduction in customer demand likely would decrease our gross margins and operating income”. Research and development runs a stable 10.2% of revenue; selling and administrative costs fell from 18.1% to 14.7% over three years on operating leverage rather than absolute cuts. Share-based compensation was $194M, 2.7% of revenue and 24% of net earnings.

Comparability. These four years are not a like-for-like series. Finisar closed in September 2019 and Coherent, Inc. in July 2022, so FY2023 onward describes a different company from FY2019. Reportable segments were redefined in FY2020, FY2023, FY2025 and FY2026, and the FY2025 change moved $59.6M of FY2024 cost between segment and corporate. The company presents no operating-income subtotal on the face of its income statement, so the operating margins above are derived from disclosed lines and reconcile to reported pre-tax earnings. FY2026 revenue is not like-for-like with FY2025 because two divested businesses were treated as held-for-sale within continuing operations rather than as discontinued operations, so prior years were not restated.

The quality of FY2026 earnings

Reported earnings of $805M and diluted earnings per share of $4.12 contain a large amount that will not repeat. A $124M gain on the sale of businesses appears as a new income-statement line, zero in both prior years. A $74M gain on the sale of an unnamed equity investment sits in other income. The effective tax rate was 7%, against 68% in FY2025 and a 21% US statutory rate; the reconciliation shows the drivers are discrete — a $62.8M release of unrecognised tax benefits, a $47.5M German tax-law change, and share-based payment benefits that depend on the share price. At a statutory rate the tax charge would have been roughly $178M rather than $61M. Restructuring charges also fell $97M year on year, worth about 1.4 points of operating margin, and management expects the current programme substantially complete by the end of FY2026 — so the year-on-year benefit stops too. Stripping the gain and adding back the held-for-sale impairment gives operating income of $838M, an 11.8% margin rather than 12.6%.

Cash is the part that has not recovered

Operating cash flow was $79.5M in FY2026, against $633.6M in FY2025 and net earnings of $805M. Capital expenditure was $1,103M. Free cash flow was therefore negative $1,023M, against positive $193M a year earlier. The mechanism is a working-capital build of extraordinary size: inventories rose 79.5% to $2,581M against 22.5% revenue growth, with work in progress alone up 89%; receivables rose 39.3%. Days of inventory went from 139 to 212. Against this, accounts payable rose 125% to $1,905M and days payable outstanding stretched from 82 to 156 — without that extension operating cash flow would have been sharply negative, and $372M of the payables balance is unpaid capital expenditure rather than trade credit. The gap was funded by NVIDIA’s $2.0bn equity purchase, $437M of divestiture proceeds and $89M from an equity investment sale.

For contrast, Lumentum — running the same demand at 42% of Coherent’s revenue — generated $751M of operating cash flow and $300M of free cash flow in its FY2026 (Lumentum FY2026 10-K). The divergence is not the end market. It is the scale of the build Coherent has committed to: capital expenditure went from 7.6% of sales to 15.5% in one year, construction in progress more than doubled to $777M, and $11.8bn of purchase commitments to vendors sit behind it, of which $3.4bn falls due in FY2027 against $1,092M a year earlier. Reported capital expenditure understates the commitment further, because $372M of additions were unpaid at year end and CHIPS Act investment credits reduced the recorded amount by $39M. Depreciation actually fell in FY2026 despite a 60% increase in net property, because most of the new capacity is not yet in service; it will step up sharply.

Where the cash has gone, and what that says

Over the four years FY2023 to FY2026, ranked: acquisitions $5,489M — effectively all of it the Coherent, Inc. close on the first day of FY2023; gross debt repayment $4,466M, of which $1,269M was voluntary prepayment; capital expenditure $2,327M, front-loaded into FY2026 at $1,103M; dividends $50M, all on preferred stock, with no common dividend ever paid; and share repurchases of exactly zero, with no repurchase programme in effect in either FY2025 or FY2026.

The pattern is legible. Management bought two large businesses on debt, spent four years paying that debt down ahead of schedule, refinanced repeatedly to cut the spread — the term loan margin came down from 2.75% to 2.50% in April 2024, to 2.00% in January 2025, and to 1.75% in September 2025, when the revolver was also upsized to $700M and a new $1.25bn term loan added — and then, once the balance sheet was repaired, turned the capital taps on for capacity. Total debt fell from $4,310M at June 2023 to $3,222M at June 2026 while interest expense fell from $288M to $190M. The equity side went the other way: the Series B preferred, $2.15bn of face value issued to Bain Capital in 2021 and 2022, converted into 30.1 million common shares in the December 2025 quarter after the holder waived its dividend rights, and NVIDIA bought 7.8 million more. Shares outstanding rose 25.8% in one year. This is management that has consistently chosen dilution over leverage, and growth over returning cash.

Two things post-date the reported figures. On 12 August 2026 a foreign subsidiary entered an unsecured facility of approximately $945M, undrawn as at the filing date and earmarked partly for working capital. On 27 August 2026 the board approved $100M of one-time performance share units across five executives, $50M of it to the chief executive, vesting only against absolute share-price compound growth rates of 10% to 25% over four years and gated on relative total shareholder return above the median of an electronic-components index.

6. Cyclicality, Constraints & What to Monitor

Coherent now discloses only two end markets, which are the same as its two segments; the four-market split of communications, industrial, instrumentation and electronics was discontinued from 1 July 2025 and was not restated. By customer location, North America is 65.1% of revenue and rising, Europe 11.4%, China 11.4% and Japan 5.5%. Customer concentration is severe and worsening: one customer at 20% of revenue in FY2026 against 10% in each of the two prior years, and a second at 12% — roughly $2.28bn between them, both in Datacenter & Communications. Neither is named.

Where the business sits in its own cycle

At the top, on volume; not at the top, on margin. Fourth-quarter FY2026 revenue of $2,045M (derived by subtracting nine-month figures from the full year) is 49% above the previous peak quarter of $1,370M in December 2022 and 94% above the trough of $1,053M in September 2023. Full-year revenue of $7,118M is 38% above the prior peak year. On gross margin the picture is different: the quarterly GAAP figure of 38.5% remains 250 basis points below the all-time peak of 41.0% set in December 2020, and the annual 37.5% is 160 to 240 basis points below the FY2017 and FY2021 peaks. Annual GAAP operating margin of 12.6% is 30 basis points below the FY2021 peak of 12.9%, and 11.8% excluding the one-off gain and impairment. Peak volume at below-peak margin is what a business looks like when it is growing into a cost base rather than harvesting one.

The last downturn is instructive because it was not, as usually described, purely telecom. FY2024 revenue fell 8.8% with declines in all four markets then reported: electronics fell 43% on a single customer’s design change, instrumentation 17% on inventory digestion, industrial 5%, and communications only 1% — because a genuine telecom capital-spending collapse was already being masked by datacom growth. GAAP operating margin went from negative 0.7% to positive 2.0% across the trough, and gross margin bottomed at 30.9%.

What a downturn would do

The cleanest evidence is Lumentum, whose mix made it the pure-play victim of the same cycle. Revenue fell 23.1% from FY2023 to FY2024; GAAP gross margin fell 1,370 basis points to 18.5% and non-GAAP gross margin 1,020 basis points; GAAP operating margin went from negative 6.5% to negative 31.9%. Roughly 44 basis points of gross margin was lost for every 1% of revenue decline. Coherent’s more diversified mix cushioned it to about 27 basis points per 1% over the same window. Apply the milder of those elasticities to a company whose depreciation base is about to step up materially and whose cost structure is now far larger than it was in FY2024, and the outcome is worse than FY2024 was, not better. This does not require an AI spending pause. It requires only that the indium phosphide shortage ends — which is what LightCounting expects and what every participant’s capacity expansion is designed to cause.

The industry base rate is the other half of the argument. Average optical component supplier margins ran 9% in 2016–17, 4% in 2018 and 1% in 2019, and LightCounting attributed the collapse to price erosion that accelerated as volumes fell. There is no episode in this dataset where volumes declined and price held.

Leading indicators, and where they are published

7. Risks, Unknowns & Questions for Deeper Work

Cyclicality is covered in Section 6. What follows is what cyclicality does not capture, ordered by how badly each compounds with the others.

• Customer concentration is now compounding with an inventory and commitment position built for those customers. One customer is 20% of revenue and a second is 12%. Coherent’s own filing concedes that such customers seek price concessions and may impose most-favoured-customer or exclusivity terms that foreclose selling to others. If forecast orders do not arrive, the 10-K names the consequences precisely: excess or obsolete inventory, underutilised capacity, liabilities under supplier arrangements, reimbursement obligations for supplier capital expenditure, and non-cancellable purchase commitments. The exposure is $2,581M of inventory at 212 days and $11.8bn of vendor purchase commitments. A single large customer deferring a ramp would produce a cost-of-goods charge and a balance-sheet write-down simultaneously.
• A $3.1bn goodwill balance sits in the shrinking half of the company with an 8% cushion. At the 1 April 2026 test date the Lasers reporting unit’s fair value exceeded its carrying value by approximately 8% on an 11.0% discount rate, and the company states the unit “remains sensitive to changes in assumptions and future operating performance”. That unit sits inside Industrial, whose revenue fell 10% in FY2026, and the valuation assumes profitability rises as volumes and site-consolidation benefits arrive. Goodwill and intangibles together are $7.26bn, 39.7% of assets and 64.6% of equity. A modest forecast reduction produces a nine-figure non-cash charge.
• Broadcom is entering Coherent’s component categories from a structurally higher margin. The March 2026 announcement of 400G EMLs and photodiodes, alongside 200G-per-lane VCSELs, puts a 67.8%-gross-margin fabless company into the products where Coherent is currently capacity-constrained and therefore most profitable. Coherent has no equivalent cross-subsidy. The scale is unknown because Broadcom does not disclose optical revenue.
• An unreserved export-control contingency remains open. In January 2025 Coherent received a Bureau of Industry and Security inquiry concerning past product sales to Huawei. It has stopped shipping to Huawei, is cooperating, and states it cannot determine an estimate or range of loss. Penalties, suspension of export privileges or debarment are the disclosed possibilities. There is no update in the FY2026 10-K. Separately, forward export controls on products enabling AI infrastructure would, on the company’s own description, force charges for excess inventory and non-cancellable purchase obligations, because the capacity has already been committed.
• The debt is small but the maturity is concentrated. $2,136M of principal matures in FY2030 and $986M in FY2031 — 96% of the total in a two-year window, including the $990M of 5.00% senior notes due December 2029. The revolver and term A carry a springing maturity 91 days before those instruments if liquidity falls below $250M plus the outstanding principal. Most debt floats; a 100 basis point move is $25M of interest, and the interest rate cap that saved $17M in FY2026 has decayed in value from $17M to $3M.
• The scarcity thesis is falsifiable and dated. The investment case rests on indium phosphide scarcity being structural rather than cyclical, and on Chinese module supply being restricted. As of September 2026 no US rule in force restricts imports of Chinese optical transceiver modules; the Section 301 semiconductor action carries a 0% additional rate until June 2027 and does not name modules, and the reported FCC action remains a draft with no primary document located. Both legs of the thesis are testable, and both would fail in the scenario the industry’s own tracker considers most likely.

What the sources could not answer

Each of these is a disclosure the company chooses not to make, or a figure no independent source publishes. They are the questions to put to management or to resolve before a thesis can be formed.

• Transceiver-only revenue, unit volumes and average selling prices. Coherent states explicitly that it does not disaggregate revenue by product type because the chief operating decision maker does not use that view. Without units and prices, the direction of price cannot be separated from the direction of volume.
• The organic versus divested revenue bridge for Industrial, and the revenue of the aerospace and defence, Munich and Newton Aycliffe businesses that were sold.
• The identity of the two customers at 20% and 12% of revenue, and whether NVIDIA is one of them. The 10-K does not say and the connection should not be inferred.
• The dollar value of the NVIDIA multi-year purchase commitment, described only as multi-billion, and its year-by-year shape against the capacity obligations running to 2030.
• The split of capital expenditure between growth and maintenance, and any capital expenditure plan for FY2027. Neither is disclosed.
• Backlog and book-to-bill. Coherent discloses neither; the $11.8bn of purchase commitments is an obligation to suppliers, not an order book.
• Standalone silicon carbide revenue and profitability, which ceased to be visible when the Materials segment was folded into Industrial — against an implied $4.0bn valuation set in 2023 and $604M of restricted cash inside the subsidiary.
• Whether optical transceivers fall inside the Section 232 covered-product annex published in January 2026. The annex was not obtainable.
• Per-vendor 800G and 1.6T market share, which every independent tracker places behind subscription. The share picture in this memo rests on revenue comparison rather than measured share.

8. Investor Takeaways

• Coherent is a vertically integrated component manufacturer that has become, for now, an AI infrastructure supplier: three-quarters of revenue is Datacenter & Communications, and it sells goods outright with no recurring revenue.
• The economic engine is the indium phosphide wafer. Owning qualified 6-inch InP capacity is what lets Coherent ship transceivers when the industry cannot, and the 6-inch yield curve is the one genuine cost advantage the evidence supports.
• The growth lever is hyperscaler capital spending converted through wafer output. Everything else — mix, pricing, restructuring — is second order or non-repeating.
• What breaks the story is the shortage ending. Prices in this industry have never held when volumes fell, Coherent has fixed its selling prices under multi-year agreements, and it has taken capital expenditure to 15.5% of sales into a shortage that every competitor is spending to eliminate.
• Monitor the gap between earnings and cash. FY2026 produced $805M of net earnings and $80M of operating cash flow; until that gap closes, the reported recovery is an accrual, not a cash, event.