Bear case
The Coherent Corp bear case is more textured than the Lumentum bear case because Coherent’s vertically-integrated breadth introduces additional failure modes that don’t apply to a more-focused competitor. The five primary failure paths below would each individually compress the bull-case price-target and could combine to produce a meaningfully worse outcome. Each pillar has invalidating signals (which would refute the bear case) and confirming signals (which would advance it).
Bear pillar 1 — NVDA customer concentration risk
NVDA’s multi-billion-dollar purchase commitment to Coherent is structural concentration risk. The bear-case framing:
- Concentration is real even if committed — even with NVDA’s purchase commitment, a single customer accounting for 25–35% of Coherent’s D&C revenue (~20–25% of consolidated revenue) is a structural concentration that would be problematic if the relationship deteriorated.
- Purchase commitments are not equivalent to take-or-pay — the multi-year scope of NVDA’s commitment may have flexibility provisions that allow renegotiation if NVDA’s own demand softens.
- NVDA could in-source CPO source lasers in CY2028+ — even with the March 2026 equity investment, NVDA’s long-term strategic interest is in supply-chain control. Equity ownership in merchant suppliers is consistent with transitional alignment but not necessarily terminal alignment.
- NVDA dependency narrows the customer-acquisition energy — Coherent’s sales team is increasingly focused on serving NVDA’s specific roadmap, potentially at the cost of multi-customer relationship breadth.
Confirming signals (advance bear case):
- NVDA quarterly capex guidance moderation
- NVDA optical-supplier announcement (e.g., NVDA in-house InP fab plan)
- NVDA-Coherent renegotiation of purchase commitment terms
- NVDA preferring CPO architectures that bypass Coherent’s vertical-integration scope
Invalidating signals (refute bear case):
- NVDA expanding the multi-billion-dollar purchase commitment
- NVDA-Coherent follow-on agreements expanding scope (CPO, telecom transport)
- NVDA capex sustained 30%+ growth through CY2027
Bear pillar 2 — CPO timeline slip
The 2027 first-volume CPO and 2028 NVLink-scale-up CPO timing is industry-consensus aspirational rather than confirmed. Multiple historical photonics-architecture transitions (PSM4 → 400G, coherent-DWDM → pluggable, etc.) have slipped 12–24 months from initial-public-roadmap timing.
The bear-case framing:
- CPO assembly is unprecedented at scale — co-packaging optics with the switch ASIC is technologically aggressive. Yield, reliability, thermal management, and serviceability challenges are real.
- Pluggable-form-factor improvements (LPO, OBO) could absorb growth that the industry expected CPO to take — if LPO/OBO win 1.6T / 3.2T share faster than expected, CPO penetration is delayed.
- NVDA’s switch-vendor relationships could pivot toward Broadcom or other CPO-engine vendors that exclude Coherent’s silicon-photonics franchise from the engine layer.
- Coherent’s vertical-integration scope into CPO modules is unproven at volume — the franchise has not yet shipped CPO modules at hyperscaler scale.
Confirming signals:
- NVDA pushes Spectrum-X / NVLink CPO timeline by 12+ months
- LPO captures more 1.6T / 3.2T share than expected
- Marvell or Broadcom CPO-engine vendor wins outsized share, excluding Coherent SiPh
- Hyperscaler-direct CPO procurement RFPs delayed beyond 2028
Invalidating signals:
- Spectrum-X / Quantum-X Photonics shipment volumes hit announced milestones
- Coherent CPO-product announcements with NVDA-named scope
- CPO module assembly capacity additions at Coherent
Bear pillar 3 — NVDA in-sourcing risk
Even with the bilateral $4B equity structure preserving the duopoly through 2028, the long-term direction of NVDA’s optical-supply strategy could pivot toward in-sourcing. The bear-case framing:
- NVDA’s silicon roadmap moves toward heterogeneous integration — at scale, monolithic InP-on-Si or quantum-dot-on-Si laser approaches could displace flip-chip-bonded merchant InP supply.
- NVDA owns the silicon stack — DSPs, CPO engines, and switch ASICs are increasingly NVDA-controlled. The optical layer is the only remaining external-supply layer.
- NVDA could acquire merchant supplier capacity — the equity stakes in Coherent and Lumentum are consistent with eventual acquisition optionality (though no specific signal exists today).
- NVDA’s CY2030 architecture — beyond the current visible roadmap (Spectrum-X 2026 → CPO 2028), NVDA’s next-generation strategy could change the optical-component dependency.
Confirming signals:
- NVDA announces heterogeneous-integration breakthrough (monolithic InP-on-Si)
- NVDA acquires a captive optical-supply asset
- NVDA reduces purchase commitments to Coherent / Lumentum
- NVDA’s CY2030 roadmap emphasizes captive optics
Invalidating signals:
- Continued bilateral commitment to Coherent + Lumentum
- NVDA-Coherent partnership scope expanding
- No public signal of NVDA captive-optics development
Bear pillar 4 — SiC pricing pressure + Industrial weakness
Coherent’s SiC + Industrial-laser businesses face structural pricing pressure that compresses Industrial-segment margin expansion potential:
- Chinese SiC substrate suppliers (TankeBlue, SICC) are pricing aggressively on 150 mm and ramping 200 mm capacity that will pressure ASPs.
- EV-cycle deceleration in CY2024–2025 has reduced near-term SiC volume growth; if EV adoption continues to plateau, the SiC TAM trajectory weakens.
- Industrial fiber-laser cyclicality — IPG Photonics + Trumpf competition keeps ASP discipline modest; China-capex weakness extends through CY2026.
- Aerospace & Defense divestiture (CY2025 close) removes ~$400M of higher-margin defense-grade revenue from the Industrial segment.
The Industrial segment posted –9.9% YoY revenue decline in Q2 FY2026 — the weak baseline. If Industrial continues to weaken, the consolidated operating-leverage profile compresses.
Confirming signals:
- Continued YoY declines in Industrial segment
- Manufacturing PMI < 50 across multiple geographies
- SiC ASP compression visible in margins / commentary
- Wolfspeed restructured re-emergence as a price-pressure source
Invalidating signals:
- Industrial segment QoQ revenue inflection from negative to positive
- IPG Photonics quarterly results showing stabilization
- DENSO/Mitsubishi JV partners increasing volume commitments
Bear pillar 5 — Vertical-integration operating-leverage compression
Coherent’s vertical-integration breadth is a feature in the bull case but a structural cost-base burden in the bear case:
- More fixed costs across more business lines — legacy industrial-laser overhead, materials processing, life-sciences instrumentation, etc., consume opex that a focused photonics competitor can avoid.
- Slower decision-making and resource allocation — large diversified entities exhibit decision-friction that tighter operations can avoid.
- Capital allocation across competing platforms — Coherent must invest in InP fab + SiPh + transceiver assembly + ROADM/wave-shaper + SiC + industrial lasers + life sciences. A competitor focused only on AI-photonics can deploy capital more efficiently.
- Series B Convertible Preferred dilution overhang — the Bain Capital Series B (held alone via the BCPE Watson SPV; NOT Senator) is potential incremental dilution of up to ~$2B authorized face / ~23M+ common-equivalent shares at the $85.00 optional conversion strike, currently ~3.6× ITM at COHR ~$305. The November 2025 dividend waiver halted PIK accretion but the converted-share base accreted between July 2022 and November 2025 is locked in.
Confirming signals:
- Operating margin expansion stalls below 22%
- Working-capital growth outpaces revenue (signal of execution friction)
- Capex creep beyond guidance
- Series B preferred conversion accelerates (more dilution sooner)
Invalidating signals:
- Sustained operating-margin expansion through FY2028
- Working-capital efficiency improving
- Capex on plan with capacity-utilization rising
Composite bear-case framing
The composite bear case argues:
- AI capex moderates — hyperscaler capex CAGR drops to 10–15% in CY2027–CY2028; Coherent’s D&C revenue stalls at ~$7B by FY2028.
- NVDA renegotiates — purchase commitment scaled to ~70% of original framing; Coherent’s NVDA revenue contribution falls below bull-case path.
- CPO timeline slips 12–18 months — Spectrum-X scale-out CPO pushed to CY2028; NVLink scale-up CPO pushed to CY2029.
- Industrial weakness persists — Industrial segment flat-to-down through FY2028; SiC ASP compression reduces Materials margin contribution.
- GM compresses to 35–37% — partly mix-shift offset by ASP compression on EMLs and Industrial-segment drag.
- Operating margin compresses to 16–19% — reflecting margin pressure plus operating-leverage compression on smaller-than-expected revenue base.
- Series B convert pushes diluted share count to 220M+ by FY2028 — additional dilution overhang.
- Multiple-compression — AI-photonics multiple regime breaks; COHR forward P/E compresses to 18–22×.
This stack drops to FY2028 non-GAAP EPS of approximately $5–7 with 18–22× P/E — implying $90–155 per share by mid-2028.
What would invalidate the bear case
The bear case would be invalidated by:
- Sustained AI capex acceleration (≥30% CAGR through CY2028)
- NVDA partnership expansion (new commitments, expanded scope)
- CPO commercial-volume milestones hit on NVDA-stated timing
- Industrial segment QoQ inflection from declining to growing
- Sustained operating-margin expansion past 22% through FY2027
Cross-link
- Bull case — opposing view
- Risks — formal risk register
- Catalysts — events to monitor
- Open questions
- Cross-thesis implications
- 04_market AI capex cycle, InP EML duopoly, CPO market, SiC market, industrial laser market
- 05_financials DCF assumptions — bear-case scenario detail
- LITE — bear case — duopoly-partner risk comparison
Sources
- All sources documented across 04_market and 05_financials; primary load-bearing sources for the bear case include Q2 FY2026 release (Industrial –9.9% YoY signal), Coherent 8-K Mar 2, 2026 NVDA placement (concentration risk anchor), Bain dividend waiver Nov 2025 (Series B preferred-related context), and FY2025 annual report ✓