Jefferies initiates Cohu stock with buy rating on AI test demand

Jefferies initiated Cohu with a Buy and a $55 price target versus the current $46.93 share price, citing 16% revenue CAGR from fiscal 2025-2028, 49% gross margin by fiscal 2028, and 105% EPS CAGR from fiscal 2026-2028. The bullish case is driven by AI and high-bandwidth memory test demand, plus recovery in automotive and industrial end markets. Cohu also disclosed $30 million of follow-on Eclipse platform orders and additional U.S. semiconductor manufacturer orders, reinforcing demand visibility.

Analysis

The important read-through is not just that COHU is benefiting from AI spend, but that test is becoming a higher-content bottleneck as device complexity rises. That shifts bargaining power toward differentiated testers and away from commoditized capital equipment vendors, especially if customers need more insertions per wafer and more software/recurring support to manage yield learning. The second-order winner is likely the broader semi test ecosystem, while laggards are smaller legacy ATE vendors that lack exposure to HBM, HPC, and datacenter qualification cycles.

The market is probably still underestimating the durability of the mix shift. If AI-related demand is real, it should extend beyond a single order announcement into multi-quarter pull-through as new packages move from development to ramp, which means the revenue inflection could last through the next 4-6 quarters rather than one quarter. The key swing factor is gross margin leverage: once high-complexity platforms get installed, software and recurring revenue should scale faster than hardware, but any inventory correction or delayed factory acceptance can quickly compress sentiment because the stock already discounts a strong recovery.

Consensus may be treating this as a cyclical rebound story, when it may actually be a structural content-per-chip story. That matters because cyclical auto/industrial upside is easier to fade, while AI test intensity can support a higher terminal multiple if it proves repeatable across multiple customers. The contrarian risk is that this becomes a “good news, already priced” setup: after a 178% run and near-peak trading levels, even a modest slowdown in order cadence could trigger multiple compression before fundamentals roll over.

The cleanest market expression is to stay long COHU on pullbacks, but only if you can define risk tightly around order conversion and margin execution. For investors needing cleaner relative value, the better trade is a pair that isolates AI test content versus beta to the broader semiconductor equipment cycle.

Key Decisions for Investors

• Long COHU on a pullback toward the low-$40s with a 3-6 month horizon; target a retest of the mid-$50s if follow-on orders convert into visible backlog growth, but trim if the stock fails to hold prior breakout levels after the next print.
• Buy COHU call spreads for 3-6 months out rather than outright stock; the setup is momentum-positive but already extended, so defined-risk upside participation is better than paying full delta into a possible consolidation.
• Pair trade: long COHU / short a more cyclical semiconductor-capex proxy over the next 2-4 quarters to isolate AI test intensity versus generic equipment beta; the thesis works if COHU’s order book proves stickier than the broader group.
• Avoid chasing here if you are valuation-sensitive; at these levels, the stock needs continued execution on margins and recurring revenue to justify further multiple expansion, so a failure to confirm in the next earnings cycle is the main near-term downside catalyst.
• Use any post-earnings weakness caused by margin noise as an entry point, not a reason to exit the long thesis, provided order cadence remains intact and management commentary supports multi-quarter demand visibility.