COHU · Cohu, Inc. · Stock research

Current Thesis

Cohu's re-rate from a cyclical handset/auto-test laggard into an AI-compute test picks-&-shovels name has moved from mid-discovery to a broad, multi-desk sell-side chase. The leg an investor is buying is unchanged: the AI xPU test-handler ramp (Eclipse + Active Thermal Control) plus HBM inspection (Neon) inflecting through H2 2026, with order velocity backing it (orders +57% YoY on the 2026-04-30 Q1 print). What has changed is that the sell-side has now caught up in force three price-target actions in under a month: TD Cowen $60→$80 (2026-06-16), Baird initiating Outperform at $65 (2026-07-09), and Stifel $54→$70 (2026-07-10). The stock printed a fresh all-time high of $70.47 (2026-06-18) off roughly $55 at the start of June about +25% in two weeks and now sits at or above nearly the entire published target stack. The setup has also degraded on the calendar: the Q2 FY26 print lands 2026-07-30 (after close), now ~18 days out and inside the catalyst window, converting a slow-burn narrative into a near-term binary. The theme is still ACCELERATING, but a fresh entry here pays up the day sell-side finished re-rating, into a print that will confirm or crack the H2 HPC ramp. High-conviction story, late-stage entry.

Bullish and bearish views on Cohu, Inc.

The model's bull view on Cohu, Inc. (COHU), in brief: Q1 FY26 (2026-04-30): revenue $125.1M vs consensus ~$122.1M; orders +57% YoY; gross margin 46.5% a beat-and-raise with a high-quality recurring/consumables mix (~60% of revenue). The bear view: Extended into a fresh ATH: ~$70 with a $70.47 high (2026-06-18) sits at the top of the 52-week range ($17.71–$70.47) after a +25% two-week sprint. Both cases follow in full.

Bull Case

• Q1 FY26 (2026-04-30): revenue $125.1M vs consensus ~$122.1M; orders +57% YoY; gross margin 46.5% a beat-and-raise with a high-quality recurring/consumables mix (~60% of revenue).
• Q2 FY26 guide $144M ±$7M = +15% QoQ, ~+34% YoY, driven explicitly by HPC/AI demand rather than a one-quarter pop.
• Broadening sell-side conviction: TD Cowen $80 (2026-06-16, drawing the "Teradyne a year ago" AI-re-rate analogy), Baird initiating Outperform at $65 (2026-07-09), Stifel to $70 (2026-07-10). The cluster has migrated from $50–60 post-Q1 to $65–80.
• FY26 revenue-growth guide raised to +20–25% on the Q1 call; AI-compute pipeline lifted to ~$750M ($650M test handlers + $100M HBM inspection); FY26 HPC revenue outlook $80–100M.
• Eclipse AI xPU test handler: ~$650M SAM, $60–80M FY26 revenue, ~$30M follow-on orders booked; dissipates 3kW today with a 10kW roadmap into a GPU TAM compounding ~23%.
• Neon HBM inspection/metrology: ~$100M SAM, ~$20M FY26 revenue, +80% YoY direct leverage to the memory build-out.
• Pipeline depth (Q1 call): ~12 AI/HPC customers (5 in qualification, 7 in early engagement); ~$5M new DiamondX GaN power orders add data-center-power exposure.
• TTM revenue $481.3M, +23.1% YoY; market cap ~$3.3B an institutional re-rate rather than a float squeeze.

Bear Case

• Extended into a fresh ATH: ~$70 with a $70.47 high (2026-06-18) sits at the top of the 52-week range ($17.71–$70.47) after a +25% two-week sprint. Buying here is the range top, ~18 days from a binary print.
• Price has caught up to the target stack: after the July raises the tape is roughly in line with Baird's $65 and Stifel's $70, with only TD Cowen's $80 — above it the re-rate is mostly priced, not front-run.
• Still loss-making: Q1 FY26 was a net loss of $12.1M, or −$0.26/share. The bull case is a return to profitability in 2026, a forecast rather than a printed result, so the multiple runs ahead of the P&L turn.
• Margin tailwind is fading: GM guided to ~44% for Q2 from 46.5% as systems mix rises and HPC scales the high-margin consumables slice dilutes as the story monetizes.
• Binary risk stacked on top of positioning: the 2026-07-30 print must show Eclipse follow-on cadence and new xPU design wins to justify the re-rated multiple; silence on order momentum after a +25% run is an asymmetric downside.
• Small-cap beta cuts both ways: on a semicap-theme unwind it gaps harder than the large-cap complex it trades with.

Setup & Price Structure

Price structure is a parabolic advance that ran from ~$55 (early June) to a $70.47 ATH (2026-06-18) and has held near the highs while the sell-side ratcheted targets up through July. The stock is now above the June breakout shelf (~$55) and effectively at the ceiling of its published-target range. Momentum readings on a move of this slope are stretched, and the name has no overhead resistance but also no nearby support until the breakout base near $55, a ~20% air pocket below spot. This is the classic late-entry configuration: peak sell-side attention, price at range top, extended above rising moving averages, and a binary catalyst approaching inside three weeks. Strength is the setup in this playbook, but strength into a print, after the analysts have already moved, is a poor entry location the reward for chasing here is thin versus waiting for either the print to clear or a reset to the base.

Catalyst Calendar (next 30 days)

• 2026-07-30 (after close) Q2 FY26 earnings. The defining binary. Validates or breaks the H2 HPC ramp; watch Eclipse follow-on order cadence (~$30M booked so far), new xPU handler design wins, and the 12-customer pipeline progression (5 qual / 7 early engagement). Avoid fresh entries inside three trading days of it.
• Ongoing analyst cadence three PT actions in the last month (TD Cowen 6/16, Baird 7/09, Stifel 7/10). Further clustered upgrades into the print would confirm narrative acceleration; a pause or a "priced-in" downgrade would mark the sell-side chase peaking.
• No FDA/PDUFA, no scheduled filings, no index events in the window. The print is the only hard dated catalyst.

What Would Change Our Mind

• A reset that restores the entry: a pullback toward the early-June breakout shelf (~$55) that holds and bases would convert a late chase into a clean re-entry on the same intact narrative.
• A post-print continuation: a Q2 beat-and-raise on 2026-07-30 that discloses fresh Eclipse follow-on orders and new xPU design wins, followed by a higher low, would re-arm the setup at a defined-risk entry rather than a blind chase.
• Order-cadence disclosure: explicit new AI/HPC bookings above the ~$30M Eclipse base is thesis acceleration; silence on follow-on velocity is the first crack.
• Downside tell: loss of the breakout base on a weekly basis, or a guide-down that stalls the H2 HPC ramp, flips the theme from ACCELERATING toward crowded-and-fading and removes the reason to own it.

Correlation Notes

Cohu is a back-end semicap-equipment name and trades with that cluster, not as a standalone AI story: TER, ONTO, FORM, KLAC, AMAT, LRCX. Its ~$3.3B cap makes it a higher-beta expression of the same AI-test/back-end theme those larger names carry, so it amplifies both the up-moves and the drawdowns of semicap sentiment. For a concentrated book, sizing here should account for overlap a position in Cohu alongside another AI-test/back-end equipment name (LRCX in particular) is largely one correlated bet on the same H2 2026 HPC/HBM test-capacity ramp, not two independent ideas. Treat it as the small-cap, high-torque sleeve of the semicap-equipment exposure rather than a diversifier.