COHRCoherent Corp.

Investment Thesis

Coherent Corp. is the most vertically integrated photonics company in the Physical AI investable universe. The investment thesis is that as AI GPU clusters scale, optical interconnect becomes the hard constraint — and Coherent owns more of that constraint than any other public company. The company spans the full photonics stack: it grows indium phosphide (InP) and gallium arsenide (GaAs) substrates, fabricates laser diodes and photonic integrated circuits, and assembles finished 800G and 1.6T optical transceivers for hyperscaler data centers. NVIDIA validated this strategic position with a $2 billion investment in Coherent as part of a broader $4 billion photonics commitment.

The financial scale is in a different league from smaller optics pure-plays. Q3 FY2026 total revenue reached $1.81 billion, up 21% year-over-year. The Datacom & Communications segment — approximately 75–77% of revenue, or roughly $1.4 billion in the quarter — grew 37% YoY and carries a book-to-bill ratio exceeding 4x, indicating multi-year demand visibility. The company is doubling internal InP manufacturing capacity ahead of schedule and shifting to 6-inch wafers to improve throughput and cost.

The non-datacom segments — industrial lasers, LiDAR/sensing components, defense photonics, and silicon carbide (SiC) substrates — contribute approximately $400–450 million per quarter and provide diversification. LiDAR is a meaningful but secondary revenue driver within this segment; the primary LiDAR exposure is as a picks-and-shovels supplier of laser diodes, fiber lasers, and optics to LiDAR manufacturers, rather than as a LiDAR system vendor.

Physical AI / Value-Chain Relevance

Coherent spans three Physical AI layers simultaneously: Perception & Sensing (laser diodes and photonics for LiDAR, sensing, and imaging), AI/ML Infrastructure (optical transceivers for GPU-to-GPU interconnect in AI clusters), and Critical Materials (InP and GaAs substrate manufacturing — the raw material layer of the photonics supply chain). This vertical integration is the structural moat. Competitors that buy substrates from third parties, fabricate lasers, or assemble transceivers — but not all three — lack the margin stacking and supply assurance that Coherent can offer in a capacity-constrained market.

The Physical AI relevance of Coherent's optics franchise is twofold. First, every AI training and inference cluster depends on optical interconnect to function at scale; as clusters grow from tens of thousands to hundreds of thousands of GPUs, optical bandwidth requirements grow faster than linearly. Coherent is one of a small number of suppliers qualified to deliver at hyperscale. Second, the sensing and LiDAR component business supplies the perception layer of autonomous systems — the laser diodes and optics that enable robots, drones, and autonomous vehicles to see and range their environments.

The NVIDIA $2 billion investment is the ecosystem's explicit endorsement of Coherent's centrality. It signals that the dominant platform orchestrator considers Coherent's photonics to be a strategic supply-chain node deserving of balance-sheet commitment.

Catalysts

Q4 FY2026 earnings, expected mid-August, is the most immediate catalyst. Key data points include: datacom revenue growth trajectory, the book-to-bill ratio for optical transceivers, InP capacity expansion progress, and any update on the NVIDIA partnership scope. The shift from 4-inch to 6-inch InP wafers is a manufacturing catalyst — larger wafers improve throughput and unit economics at scale. Any named LiDAR/sensing design wins in the defense or autonomous-vehicle markets would add a second growth narrative beyond AI datacom. The SiC substrate business could re-rate if EV and industrial power-electronics demand accelerates, though this is a secondary driver.

Positioning / What the Market May Be Missing

At a $55.25 billion market capitalization with forward P/E of 34.06, Coherent is priced as a large-cap growth compounder — appropriate given the scale, but leaving less margin for error than smaller optics plays. The stock is down 32.5% over the last month and 16.1% over the last week, reflecting broader optics and semiconductor sector weakness rather than company-specific deterioration. The 50-day moving average at $347 sits above the current price of approximately $324, while the 200-day moving average at $255 provides a deeper support floor.

What the market may be missing: Coherent's book-to-bill ratio above 4x in datacom is a leading indicator that significantly understates near-term revenue acceleration. A book-to-bill above 1.0 means revenue is growing; above 4x means demand is overwhelming supply by a factor of four, implying that reported revenue is constrained by manufacturing capacity, not by end-demand. As InP capacity doubles, that constrained demand should convert to reported revenue — a mechanical tailwind that is more predictable than new design wins.

The crowding flag is YELLOW, and conviction is 3 — moderate. Coherent has the highest tech_moat_score (6.7) in our sensor stack. However, the current price of $324 sits below the entry range of $330, and the consensus price target of $92.74 appears stale (pointing more to an old data pull than a current analyst view). The investment case is strong on bottleneck quality but requires patience on entry timing.

Risks and What Invalidates the Thesis

The primary risk is the AI capex cycle. If hyperscaler spending decelerates, the 4x book-to-bill ratio could compress faster than revenue, removing the forward-demand signal that underpins the growth thesis. Conglomerate complexity is a secondary risk: Coherent operates across materials, lasers, transceivers, and SiC — different business lines tied to different cycles, making the aggregate story harder to model cleanly. Legacy industrial lasers are cyclical and lower-growth, potentially acting as a drag on consolidated margins.

The thesis would be invalidated if: (1) co-packaged optics (CPO) — which integrate optics directly onto switch ASICs — shift transceiver demand away from pluggable modules where Coherent is strongest; (2) NVIDIA shifts its optical sourcing to a competing vendor or internalizes more of the photonics stack; or (3) the InP capacity doubling encounters execution delays that prevent the book-to-bill conversion from materializing. At current scale, a thesis break would be more about competitive displacement than demand disappearance.

What to Watch Next

Q4 FY2026 earnings (mid-August) — datacom revenue growth, book-to-bill ratio, and InP capacity timeline. Any disclosure on the scope and duration of the NVIDIA partnership beyond the initial $2 billion investment. The 1.6T transceiver transition — Coherent's position relative to competitors in the next speed tier. LiDAR/sensing design wins that could re-rate the non-datacom segment from cyclical industrials to structural growth. Silicon carbide substrate demand as a secondary growth driver.