Buy 3 AI-Powered Photonics Stocks to Tap Solid Short-Term Price Upside

Zacks highlighted Lumentum, Coherent and MACOM as AI-data-center optics beneficiaries, citing accelerating demand for 400G/800G optical networking, optical circuit switching and silicon photonics. Lumentum guided fiscal Q1 2027 revenue to $1.225-$1.275 billion, with the $1.25 billion midpoint implying more than 130% year-over-year growth; Coherent guided to $2.2-$2.4 billion, implying roughly 45.6% annual growth at the midpoint. MACOM's data-center revenue rose 81% year over year and 40% sequentially in fiscal Q3 2026, while analyst average targets imply 16.3% upside for LITE, 37.4% for COHR and 44.4% for MTSI.

Analysis

The investable issue is not generic AI-optics demand but where value accrues as architectures migrate from 800G pluggables toward 1.6T, optical circuit switching and eventually co-/near-packaged optics. LITE and COHR have greater systems and laser-content exposure, creating stronger revenue torque but also greater customer-concentration and qualification risk; MTSI is the cleaner component-content compounder with less dependence on any one packaging architecture. The second-order beneficiary is NVDA: lower network power and latency expand the usable scale of GPU clusters, supporting system-level pricing and reducing a key bottleneck to incremental accelerator deployments.

Near term, consensus estimate revisions and hyperscaler capex commentary should sustain momentum, but this is a crowded, high-expectation sub-sector rather than a discovery trade. A 1-3 month de-risking catalyst is evidence that 1.6T orders are converting from bookings/samples into shipments without gross-margin dilution; the key negative catalyst is a customer qualification delay, which can shift revenue by quarters even if end demand remains intact. Over 6-18 months, CPO is both upside optionality and a risk: successful adoption raises optical content per AI cluster, but could compress the addressable market for conventional pluggable-module suppliers and redistribute margin toward integrated switch vendors and silicon-photonics platforms.

The contrarian view is that optical bottlenecks may be priced too linearly against AI capex. Hyperscalers can improve cluster utilization through topology, software and accelerator interconnect changes before committing to broad architectural replacement, making quarterly optical demand lumpy. Prefer companies with diversified non-datacenter cash flows and demonstrated margin conversion over those valued solely on optical revenue run-rate extrapolation; an NVDA capex pause would likely hit LITE/COHR multiples before reported revenue.

A relative-value expression is preferable to outright beta: COHR offers the highest operating leverage if its margin trajectory holds, while MTSI is the higher-quality defensive leg if CPO adoption is delayed. Falsify the constructive thesis on a material cut to forward optical growth, gross-margin guidance below prior midpoints, or evidence that 1.6T demand is replacing rather than adding to 800G volume.

Key Decisions for Investors

• Initiate a 3-6 month long COHR / short LITE pair, sized dollar-neutral. COHR has greater upside to execution on systems and integrated optics, while LITE is more exposed to a high-expectation ramp and customer timing; reassess if COHR gross-margin guidance fails to improve or LITE demonstrates materially faster component growth.
• Accumulate MTSI on post-earnings volatility rather than chase momentum; use a 6-12 month horizon. The thesis requires continued data-center growth with stable consolidated margins; exit or reduce on two consecutive quarters of decelerating data-center bookings or a meaningful guide-down in next-year EPS.
• Maintain NVDA as the liquid core AI exposure and use optical names as satellite positions, not substitutes. Add only if hyperscaler capex guidance and networking attach-rate commentary remain intact; a broad capex slowdown would favor reducing LITE and COHR first.
• Set an earnings watch item for disclosed 1.6T shipment mix, optical-circuit-switching revenue conversion, and customer prepayment/backlog terms. Do not add exposure solely on analyst target-price dispersion; confirmation should come from shipment volume and incremental gross-margin performance.