Cohu sr VP & chief customer officer Christopher Bohrson sells $57,920 in stock

Cohu shares have surged 179% YoY, and a 10b5-1 insider sale by Christopher Bohrson totaled 1,000 shares at $57.92 on July 15, 2026 ($57,920). Despite mixed Q1 results, the company guided for 20%–25% YoY revenue growth in 2Q and for CY2026, and won ~$5M of DiamondX testing platform orders tied to gallium nitride for AI data-center power. Analyst price targets were lifted (Stifel to $70, TD Cowen to $80, Baird to $65 with an Outperform), supporting a mildly positive read-through, though InvestingPro flags the stock as overvalued vs fair value.

Analysis

The meaningful signal here is not the insider sale; it is that COHU is now a story stock trading on a narrow AI test narrative while the business still has to prove repeatability. After a 179% run, incremental upside depends on bookings turning into sustained revenue and margin leverage, not on one-off orders or sell-side target resets. The market is implicitly capitalizing a 2028 outcome today, which leaves little room for any slip in AI eval spending or second-half order timing.

The bigger second-order effect is on the semiconductor test ecosystem: if TSMC’s capex intensity stays high but becomes more selective, the winners are likely the most diversified exposure to advanced-packaging and high-volume production test, not the smallest, most narrative-driven names. That argues for relative underperformance in COHU versus broader ATE exposure if the AI capex cycle normalizes or if customers delay qualification cycles. Conversely, if AI power architectures broaden into GaN and other specialty materials, the benefit should diffuse across the chain, which limits COHU’s ability to maintain a scarcity premium.

Contrarian view: the consensus is treating AI-related test demand as a clean secular step-up, but the near-term cash-flow impact is still noisy because evaluation work inflates opex before it scales revenue. The true falsifier is not the insider trade; it is a downgrade to CY26 growth, missed gross margin expansion, or bookings that fail to accelerate relative to revenue over the next 1-2 quarters. In that case, the stock can re-rate down quickly because expectations are already elevated.

Key Decisions for Investors

• No action on the 10b5-1 sale itself; treat it as noise. Use it only as a reminder that COHU is extended and should be validated by next-quarter bookings/gross margin, not insider flow.
• Relative-value trade: short COHU / long TER for 1-3 months. Thesis: COHU has more valuation compression risk if AI test orders prove episodic, while TER offers broader exposure to the same cycle with less single-name narrative premium. Risk/reward improves if COHU keeps outperforming on headlines but fails to convert into guide raises.
• If long COHU, reduce size and use a defined-risk structure: buy 1-2 month put spreads or sell out-of-the-money calls against the position. This fits a stock that is already up sharply and vulnerable to any booking disappointment.
• Watch TSMC capex commentary as a sector catalyst over the next 2-6 weeks. If capex growth moderates or mix shifts away from test-heavy nodes, expect a pullback in ATE sentiment; if capex stays elevated, SOXX should hold better than COHU.
• Set a falsification alert on COHU: a CY26 growth guide below 20% or gross margin missing by >150 bps would argue the AI thesis is not yet monetizing. Under that outcome, exit longs and rotate into broader semiconductor equipment exposure.