Cohu (COHU) Q3 2026 Earnings Call Transcript

Cohu reported Q3 2026 revenue of $173.0 million, up 37% year over year, and non-GAAP EPS of $0.43; GAAP net income was $12.1 million versus a $4.1 million loss a year earlier. Bookings increased 126% year over year, while the high-performance computing pipeline reached approximately $1 billion and the company raised its FY2026 HPC revenue outlook to approximately $120 million from $105 million. Management expects 2027 revenue growth of 30%–40% and plans to expand Eclipse handler output to about 3x the 2026 level, while citing longer lead times and higher costs for some components as near-term risks.

Analysis

COHU’s upside case is increasingly about converting customer qualifications into shipments, not merely expanding the opportunity funnel. That makes the next 1–3 months unusually execution-sensitive: announced qualification milestones from late November through Q1, plus the November 10 Investor Day, can clarify whether customer plans support the proposed Eclipse ramp. But pipeline value is not backlog, and management said improved visibility—not a broad influx of new customers—drove much of the latest pipeline increase. The market may therefore be underestimating the risk of timing slippage even as it rewards AI exposure.

For 6–18 months, successful scale-up could improve product mix and pull through thermal heads, services, and software as devices change generations. Conversely, a threefold output target raises supplier-readiness and quality risks; specialty-component tightness could absorb operating leverage through higher costs or delayed deliveries. The gap between the 2027 growth ambition and the still-developing qualification base is the key risk to the multiple. Traditional industrial demand adds some diversification, but does not yet remove dependence on the compute ramp. Cash provides a buffer against execution volatility, while convertible shares remain a potential dilution consideration.

The favorable earnings signal supports a constructive stance, but avoid paying for the full funnel before conversion is demonstrated. Falsifiers: qualification delays, Eclipse shipment shortfalls, deterioration in gross margin versus the 2027 outlook, or customer forecasts being deferred. Conversely, confirmed production conversions and stable component availability would strengthen the growth-and-leverage thesis.

Key Decisions for Investors

• Maintain a measured long bias in COHU rather than chasing an earnings-driven move; add only as late-November-to-Q1 qualification milestones convert into production orders or shipment evidence. Reassess if major milestones slip or customer demand dates move out.
• Use the November 10 Investor Day to test the bridge from pipeline to revenue: seek customer/program conversion detail, Eclipse capacity ramp cadence, and evidence that supplier capacity is scaling in step. Treat pipeline expansion without firmer production commitments as a watch item, not incremental earnings power.
• Track quarterly gross margin and specialty-component lead times alongside bookings. The thesis weakens if input-cost pressure or delivery delays prevent margin progression toward management’s 2027 range; stronger margins with rising shipments would validate operating leverage.
• Do not underwrite recurring-revenue growth from the installed base as proven yet: monitor thermal-head, service, and software attach rates as the first larger Eclipse cohorts reach upgrade cycles. Faster repeat orders would improve earnings quality; weak attach would leave results more exposed to lumpy system sales.