Margins and pricing

Coherent’s gross-margin trajectory through FY2024 → FY2026 is the cleanest expression of the AI-photonics thesis at the unit-economics level. Non-GAAP gross margin moved from 34.3% in FY2024 to 37.9% in FY2025 to 38.7% in Q1 FY2026 to 39.0% in Q2 FY2026 — a +470 bps cumulative expansion that is more than mix-shift can explain alone. The structural drivers: (1) 6-inch InP wafer-fab cost advantage (60% die-cost reduction vs 3-inch as the ramp progresses), (2) supply-tightness ASP discipline on 200G EMLs, (3) 1.6T module margins exceeding 800G margins (Coherent management framing), and (4) operating-leverage on the larger revenue base.

Gross margin trajectory

Period	GAAP GM	Non-GAAP GM	Driver narrative
FY2024 (full)	30.9%	34.3%	Pre-AI-cycle baseline; lots of acquisition-related amortization in COGS
Q1 FY2025	32.3%	36.7%	Mix shift starting; supply tightness emerging
Q2 FY2025	35.5%	38.2%	Networking ramp accelerating
Q3 FY2025	35.2%	38.5%	6-inch InP cost benefit starting
FY2025 (full)	35.2%	37.9%	+358 bps YoY non-GAAP GM expansion
Q1 FY2026	36.6%	38.7%	200G EML ASP discipline visible
Q2 FY2026	36.9%	39.0%	1.6T module ramp; +77 bps YoY non-GAAP
Q3 FY2026 (guided)	n/a	38.5–40.5%	Continuing trajectory

Source: quarterly press releases ✓.

The Q2 FY2026 +77 bps YoY non-GAAP GM expansion is smaller than prior quarters because the FY2025 Q2 base was already strong (38.2%). The structural read: margin expansion has not stalled — it’s pacing into a tougher comp.

InP EML ASP dynamics

The InP EML chip is the load-bearing pricing variable. Three speed-grade tiers matter:

Speed grade	Use case	ASP per chip ($, indicative)	Production maturity
100G EML	Legacy datacom (400G-DR4, 400G-FR4)	~$30–50	Mature; declining ASP
200G EML	800G-DR4, 800G-FR4, 1.6T-DR8	~$80–150	Production ramp; rising ASP under supply tightness
400G EML (R&D)	3.2T (future)	n/a	R&D phase

⚠ ASP figures are illustrative — Coherent does not disclose chip-level ASPs publicly. Industry trade-press analyst commentary suggests double-digit ASP increases on 200G EMLs in CY2026 under supply tightness. The 200G EML to 100G EML transition roughly doubles per-chip ASP, and 1.6T modules consume more chips per unit of bandwidth than 800G modules — both effects compound to lift Coherent’s revenue per unit of customer demand.

1.6T module margin economics

Per management commentary on Q2 FY2026 (Futurum Group analysis) ◐:

1.6T module gross margins are expected to exceed 800G levels “particularly early in the lifecycle”
The cost advantage is partly driven by the 6-inch InP wafer-fab die-cost reduction (~60% vs 3-inch baseline)
1.6T modules consume more chips per module than 800G; the chip-content uplift more than offsets module-assembly-cost increases

This is counter-intuitive vs typical product-cycle margin patterns — usually new products start at lower margins and improve through the lifecycle. Coherent’s frame argues that 1.6T starts at higher margins than 800G because (a) the 6-inch fab cost advantage is realized in the new generation, (b) ASP discipline holds through supply-tight new-product launch, and (c) the customer mix is more concentrated toward fewer high-volume hyperscaler + NVDA buyers, reducing channel/distribution friction.

If sustained, this lifts gross margin durably above 39% through the 1.6T ramp through FY2027.

Operating margin and operating-leverage

Period	Non-GAAP OM	YoY bps Δ
FY2024 (full)	~13%	flat
Q1 FY2025	~13%	flat
Q2 FY2025	~18%	+500 bps
Q3 FY2025	~18%	+480 bps
FY2025 (full)	17.8%	+472 bps
Q1 FY2026	~19%	+600 bps
Q2 FY2026	19.9%	+147 bps
Q3 FY2026 (implied from EPS guidance)	~19–22%	continuing trajectory

The operating-leverage trajectory has been steeper than the gross-margin trajectory — because Coherent has held operating expenses broadly flat while revenue grew. Q3 FY2026 operating-expense guidance of $320–340M (non-GAAP) is roughly flat with prior-quarter levels at meaningfully higher revenue. The implied operating-margin path through FY2027 is into the 22–25% range under base-case revenue assumptions.

Transceiver gross-margin profile

The pluggable transceiver is the vertically-integrated finished product that captures the mix of EML chip + silicon photonics + assembly value. Margin layers:

Layer	GM% (indicative)
EML chip (sold merchant)	50–70%
Silicon photonics engine	35–50%
Module assembly + test	15–25%
Composite finished pluggable transceiver (vertically integrated)	35–45%

⚠ Indicative ranges; Coherent does not disclose layer-level margins.

The vertically-integrated transceiver captures more dollar-margin per unit of customer demand than the pure-EML-chip merchant sale. Coherent’s vertical integration depth (deeper than Lumentum’s component-plus-Cloud-Light footprint) is a structural advantage — and is one reason Coherent’s blended GM is competitive with Lumentum’s despite Coherent’s broader Industrial-segment dilution.

Mix-shift to AI

The mix-shift narrative drives the headline GM expansion:

Period	D&C share of total	Composite GM uplift estimate
FY2024	49%	baseline
FY2025	59%	+200 bps just from mix shift
FY2026 H1	71%	+400 bps cumulative from mix shift
FY2027 (forecast)	~75%	+500 bps cumulative

⚠ Mix-shift estimates assume D&C GM ~5–7% richer than Industrial GM. The other +50–150 bps of historical margin expansion is from ASP discipline, 6-inch InP cost benefit, and operating leverage — independent of mix.

SiC pricing pressure

The SiC substrate market is the negative-pricing-pressure outlier in Coherent’s portfolio:

Chinese substrate suppliers (TankeBlue, SICC) are pricing aggressively on 150 mm wafers for the domestic-China EV-inverter market.
Wolfspeed’s 2024 distress caused some near-term pricing dislocation in the merchant SiC substrate market.
EV cycle slowdown in CY2024–2025 weakened SiC demand growth, pressuring ASPs.

Mitigations:

DENSO + Mitsubishi long-term supply agreements lock in floor pricing on 150 mm and 200 mm substrates and epi wafers.
200 mm wafer-size leadership preserves a generation-ahead premium against Chinese 150-mm-only competitors.
JV-funded capex structure removes SiC capex burden from Coherent’s parent-level capital allocation.

The SiC GM is not a Coherent-public disclosure but is structurally lower than Datacom GM. As Industrial mix declines (Industrial share drops from 51% in FY2024 to ~28% in FY2026 H1), the SiC-pricing-pressure drag on consolidated GM diminishes.

Industrial-laser pricing

Industrial fiber lasers (legacy Coherent Inc. franchise) face the standard cyclical-pricing dynamics:

Trumpf and IPG Photonics competition keeps ASP discipline modest.
Chinese fiber-laser entrants (Raycus, Maxphotonics) pressure low-end pricing.
End-market capex weakness in CY2024–2025 reduced operator-capex willingness, pulling down ASPs.

The Industrial segment posted –9.9% YoY revenue decline in Q2 FY2026 — partly volume, partly ASP pressure. The August 2025 Aerospace & Defense divestiture removed approximately $400M of higher-margin defense-grade revenue from the segment. Pro forma, Industrial would be closer to flat-to-modestly-positive.

Margin sensitivity scenarios

Scenario	FY2027 non-GAAP GM	FY2027 non-GAAP OM	Driver
Bull	41–43%	22–25%	Sustained 1.6T margin advantage; D&C share to ~75%+
Base	39–41%	20–23%	Continued mix shift; 6-inch InP cost benefit fully realized
Bear	36–38%	17–19%	NVDA capex digestion; ASP compression on EMLs; Industrial flat

Cross-link

Quarterly trend — historical detail
Segment revenue mix — D&C vs Industrial
Capex cycle — 6-inch InP fab investment driving cost advantage
04_market InP EML duopoly — pricing-discipline mechanics
04_market datacenter optics TAM — TAM / ASP framework
04_market SiC market — SiC pricing pressure detail
DCF assumptions — margin scenarios in valuation
LITE — margins and pricing — duopoly-partner comparison