Cohu, Inc. (COHU)

Executive Summary / Key Takeaways

• The Handler-to-Holistic Pivot: Cohu is attempting to transform from a cyclical test handler specialist into a diversified semiconductor equipment provider with AI-exposed growth vectors, but its $126 million quarterly revenue base remains a fraction of giants like Teradyne (TER) ($769M) and Advantest (ATE) (~$500M+), creating persistent scale disadvantages that threaten margin recovery.

• Recurring Revenue Resilience vs. Systems Cyclicality: While recurring revenue (55% of Q3 sales) provides stability with mid-50s gross margins, systems revenue (45% of Q3) remains below normalized levels despite four consecutive quarters of sequential growth, exposing the company to capital expenditure volatility in automotive, industrial, and mobile markets that are still working through inventory corrections.

• AI Opportunity: Real but Modest: Management projects approximately $40 million in AI-related system revenue for 2025, with computing potentially reaching low-teens percentage of total revenue in 2026, representing genuine traction in HBM inspection and high-power handler applications, yet this remains a small fraction of the company's overall revenue and faces intense competition from integrated test cell providers.

• Balance Sheet Fortification with Dilution Risk: The $287.5 million convertible notes offering in September 2025 strengthened liquidity to $278.9 million net proceeds, but the 1.5% coupon and potential conversion create future dilution risk, while $31.4 million in capped calls only partially mitigates this exposure up to a $41.02 cap price.

• Execution Risk on Restructuring and Market Timing: The strategic restructuring program targeting $2 million quarterly savings by consolidating manufacturing to Asia and reducing operating expenses is necessary but risky, as it coincides with a non-linear market recovery where Q4 guidance already anticipates a seasonal slowdown, testing management's ability to cut costs without impairing product development.

Setting the Scene: A Legacy Handler Company at an Inflection Point

Cohu, incorporated in 1947 and headquartered in Poway, California, spent most of its 78-year history as a test handler specialist serving the back-end semiconductor market. This heritage defines both its strengths and vulnerabilities today. The company makes money through two distinct streams: capital equipment sales (test handlers, semiconductor testers, HBM inspection systems) that follow cyclical capex cycles, and recurring revenue from consumables and software that provides stability. Its place in the industry structure is as a niche player in the $15.11 billion semiconductor test equipment market, where it competes against integrated giants like Teradyne and Advantest that offer complete test cell solutions combining handlers, testers, and software.

The core strategy emerging from recent acquisitions and product launches is differentiation through specialization. The 2018 Xcerra acquisition expanded Cohu's test handler portfolio, while the 2023 EQT team acquisition brought the Ultra-S contactor technology now generating a $20 million revenue opportunity. The January 2025 Tignis acquisition ($36.6 million purchase price) added AI process control software to deepen the DI-Core analytics offering. These moves signal a pivot from pure hardware to a more balanced model, but they also reveal the company's fundamental challenge: it must innovate on multiple fronts simultaneously while its larger competitors can leverage scale advantages in R&D, manufacturing, and customer relationships.

Industry trends create both headwinds and tailwinds. The semiconductor equipment industry is experiencing a cyclical downturn, with automotive, industrial, and mobile segments undergoing inventory corrections that have pressured Cohu's systems revenue for several quarters. However, the AI boom is creating new demand for high-power device testing and HBM inspection, where Cohu's new Eclipse handler (capable of 3,000-watt power dissipation) and Neon inspection tools are gaining traction. The company's ability to capture this AI opportunity while navigating the broader downturn will determine whether its transformation succeeds.

Technology, Products, and Strategic Differentiation

Cohu's product strategy centers on solving specific pain points that larger competitors may overlook, but this approach creates both opportunities and limitations. The Eclipse handler platform represents the company's most significant technology bet, designed to scale across diverse power applications with proprietary active thermal control. Its selection by a leading U.S.-based semiconductor manufacturer for next-generation AI processor testing validates the technical approach, but the $28 million design win ships over multiple quarters and represents just one customer relationship. The platform's ability to handle up to 3,000 watts with ultrafast temperature ramp rates addresses a real market need as AI processors grow hotter, yet competitors like Teradyne and Advantest are also developing high-power solutions within their integrated test cells, potentially limiting Cohu's addressable market to handler-only replacements rather than new test cell wins.

The Neon HBM inspection tools illustrate both the opportunity and scale challenge. Cohu raised its 2025 revenue forecast to $10-11 million after securing repeat orders, and shipped its first HBM4-configured system in Q3. While this represents genuine progress in a critical AI infrastructure segment, the revenue magnitude is modest relative to the company's $400+ million annual revenue base. The PD3x instrument upgrade for Diamondx testers offers 320-channel simultaneous measurement for display driver ICs, targeting the OLED display market, but this remains a niche application compared to the broader SoC and memory test markets dominated by competitors.

The Ultra-S contactor's $20 million revenue opportunity in precision analog testing demonstrates Cohu's ability to leverage acquired technology, but the development timeline—initiated by the EQT team before the 2023 acquisition—highlights the long gestation periods required for new products to contribute meaningfully. The Tignis acquisition aims to grow software revenue at 50% annually over three years, but 2024 revenue was sub-$1 million and 2025 is expected to be similar, with breakeven still years away. This creates a timing mismatch: the company is investing in software capabilities that won't materially impact profitability for several years while facing immediate margin pressure from the systems business.

Financial Performance & Segment Dynamics: Mixed Signals

Cohu's Q3 2025 results show a company growing revenue but struggling with profitability, a pattern that directly impacts the investment risk/reward profile. Consolidated net sales increased 32.4% year-over-year to $126.2 million, driven by stronger demand for mobile and AI-based computing applications. This growth rate outpaces many competitors and suggests market share gains in targeted segments. However, gross margin compressed to 43.8% from 46.8% in the prior year, impacted by product mix and $1 million in incremental costs from the Tignis acquisition. The company posted a net loss of $4.1 million, contrasting sharply with profitable competitors like Teradyne (15.5% profit margin) and Advantest (26.8% profit margin).

The segment mix reveals the strategic tension. Systems revenue improved sequentially for the fourth consecutive quarter to 45% of total revenue, driven by mobile and AI computing demand. However, management explicitly states it "remains below normalized levels," indicating the cyclical recovery is incomplete. Recurring revenue at 55% of Q3 sales provides stability and carries gross margins in the mid-50s, but its growth is modest compared to the systems segment's volatility. This creates a barbell risk profile: the higher-margin recurring business isn't growing fast enough to offset the cyclicality and margin pressure from systems revenue.

Operating expenses of $48 million were $2 million below guidance due to timing shifts, but the full impact of the restructuring program won't be realized until early 2026. The company is targeting quarterly operating expenses of approximately $49 million at $130 million quarterly revenue, implying significant operating leverage if achieved. However, current revenue of $126 million is already close to this target, suggesting limited room for error. R&D expenses increased due to $1 million in Tignis costs and higher material costs for new product development, while SG&A rose $0.4 million from Tignis operations. These investments are necessary for the transformation but are depressing current profitability.

Cash flow tells a concerning story. Operating cash flow was negative $14.0 million in Q3, driven by a $33 million increase in accounts receivable to support 17% quarter-over-quarter sales growth. While this reflects working capital investment in growth, it also highlights the cash conversion challenges facing smaller equipment companies. The company held $126.8 million in cash at foreign subsidiaries as of September 27, 2025, with repatriation potentially triggering withholding taxes, limiting financial flexibility.

Outlook, Management Guidance, and Execution Risk

Management's guidance for Q4 2025 reveals both confidence and caution. Revenue is projected at approximately $122 million, a 3.5% sequential decline driven by expected seasonal slowdown in systems, partially offset by continued market recovery. This guidance implies full-year 2025 revenue around $500 million, representing solid growth but still leaving the company well below the scale of larger competitors. The gross margin forecast of approximately 45% suggests improvement from Q3's 43.8%, driven by a more favorable mix including higher-margin recurring revenue, but this remains well below Teradyne's 58.9% and Advantest's 60.7%.

Operating expenses are expected at $50 million, including $2 million for variable R&D prototype materials, consistent with restructuring targets. Management anticipates quarterly operating expenses will drop to $49 million once the full restructuring impact is realized in early 2026, but this assumes successful completion of manufacturing transfers to Asia without disruption. The company is projecting interest income net of expense at $1.7 million and a tax provision of $4 million, with a diluted share count of 47.1 million shares.

The most telling guidance relates to AI opportunity sizing. Management estimates AI-related system revenue of approximately $40 million for 2025, with the computing segment potentially contributing low-teens percentage of total revenue in 2026. This represents genuine progress but also highlights the modest scale of the AI pivot relative to the overall business. The Eclipse handler's design win for next-generation AI processors and the Neon HBM inspection tools' $10-11 million forecast demonstrate technical traction, but the revenue contributions are insufficient to fundamentally alter the company's cyclical profile in the near term.

Management acknowledges the recovery will not be linear, anticipating seasonal slowdown in Q4 and a "steady, progressive recovery" in automotive and industrial markets rather than a V-shaped rebound. This cautious outlook, combined with the restructuring program's execution risk, creates a fragile near-term setup. The company must simultaneously cut costs, transfer manufacturing, launch new products, and capture AI opportunities—all while navigating a cyclical downturn.

Risks and Asymmetries: What Could Break the Thesis

The central thesis faces several material risks that could impair returns. Scale disadvantage represents the most persistent threat. Cohu's $126 million quarterly revenue is dwarfed by Teradyne's $769 million and Advantest's implied $500+ million quarterly run rate. This size gap creates a vicious cycle: smaller scale limits R&D spending (Cohu invests ~10% of revenue vs. competitors' higher absolute dollars), which slows innovation speed, making it harder to win new designs and grow scale. The severity is evident in gross margins—Cohu's 43.8% trails Teradyne's 58.9% and Advantest's 60.7% by 15-17 percentage points, directly impacting profitability and cash generation. If Cohu cannot achieve scale, it will remain trapped in a low-margin, cyclical business model.

Customer concentration risk amplifies the scale problem. While specific percentages aren't disclosed, the company's commentary references "leading" customers and "existing" relationships for major design wins like the $28 million Eclipse order. In semiconductor equipment, it's common for the top 3-5 customers to represent 40-60% of revenue. If a key customer delays capex, switches to an integrated supplier, or develops in-house solutions, Cohu could experience sudden 10-20% revenue drops that larger competitors could absorb more easily. The company's low single-digit China exposure (explicitly stated as "only a low single-digit percentage") mitigates tariff risk but also limits growth in the world's largest semiconductor market.

Execution risk on the restructuring program could derail margin recovery. The plan targets $2 million quarterly savings from manufacturing consolidation and overhead reduction, but transferring remaining production from U.S. and Europe to Asian factories risks quality issues, customer qualification delays, and employee attrition. The Q3 earnings call notes the manufacturing team is in the "final stretch" of this transfer, suggesting completion is imminent, but any disruption could impair customer relationships and increase warranty costs. The $1.7 million adjustment to Tignis contingent consideration in Q1 indicates acquisition integration challenges.

AI opportunity timing risk creates potential for missed expectations. While Cohu has secured design wins in HBM inspection and high-power handlers, the $40 million AI revenue estimate for 2025 represents less than 10% of projected total revenue. Competitors are also targeting these segments with integrated solutions that may be more attractive to customers seeking single-vendor test cells. If AI processor designs evolve differently than expected—perhaps requiring even higher power dissipation or different thermal management approaches—Cohu's Eclipse platform may require costly redesigns, delaying revenue recognition.

Balance sheet and dilution risk from the convertible notes offering could pressure the stock. The $287.5 million principal amount at 1.5% coupon provides cheap financing, but the conversion feature creates potential dilution if the stock appreciates above the conversion price. The $31.4 million in capped calls only offsets dilution up to $41.02 per share, well above the current $25.09 price but creating an overhang if the stock approaches that level. The net share settlement structure means Cohu must repay principal in cash, potentially straining liquidity if conversion occurs during a cash-constrained period.

Valuation Context: Pricing a Transformation Story

At $25.09 per share, Cohu trades at a market capitalization of $1.17 billion and enterprise value of $1.03 billion, representing 2.76 times trailing twelve-month sales of $401.8 million. This price-to-sales multiple is significantly below Teradyne's 11.14x and Advantest's implied higher multiple, reflecting the market's skepticism about Cohu's profitability and growth trajectory. The discount is warranted given Cohu's -17.2% profit margin and -8.5% return on equity, but it also embeds option value if the transformation succeeds.

For an unprofitable company in a cyclical industry, revenue multiples and balance sheet strength matter more than earnings-based metrics. Cohu's enterprise value to revenue ratio of 2.42x compares favorably to inTest (INTT)'s 0.85x but unfavorably to the larger competitors' higher multiples, suggesting the market prices Cohu as a second-tier player. The company's current ratio of 4.39 and debt-to-equity of 0.06 indicate a strong liquidity position, with $278.9 million in net proceeds from the convertible notes offering providing runway for the restructuring and R&D investments.

The absence of a price-to-earnings ratio (due to losses) and negative operating cash flow (-$14.0 million in Q3) limit traditional valuation approaches. Investors must focus on the path to profitability signals: gross margins are stabilizing around 43-45%, the restructuring aims to reduce quarterly operating expenses to $49 million by 2026, and AI-related revenue is growing. If Cohu can achieve its target of $130 million quarterly revenue with 46.5% gross margins and $49 million operating expenses, it would generate approximately $11 million in quarterly operating income, representing a 8.5% operating margin—still well below competitors but marking a credible inflection point.

The capped call transactions at $41.02 per share provide a reference point for potential dilution but also suggest management believes intrinsic value exceeds current levels. However, until Cohu demonstrates consistent profitability and scales its AI revenue beyond the current $40 million estimate, the stock will likely trade at a discount to integrated competitors, with valuation hinging on execution of the transformation rather than current earnings power.

Conclusion: A Credible Pivot with Formidable Hurdles

Cohu's attempt to reinvent itself from a cyclical handler specialist into a diversified equipment provider with AI exposure and resilient recurring revenue is strategically sound but faces formidable competitive and financial hurdles. The company's 32.4% year-over-year revenue growth in Q3 and sequential systems revenue improvement demonstrate market share gains in targeted segments, while the 55% recurring revenue mix provides stability that pure-play equipment companies lack. The Eclipse handler's selection for next-generation AI processor testing and the Neon HBM tools' $10-11 million revenue forecast validate the technology strategy.

However, the scale deficit relative to Teradyne and Advantest creates a persistent margin and cash flow disadvantage that the restructuring program alone cannot solve. Cohu's 43.8% gross margin trails competitors by 15+ percentage points, and negative operating cash flow in Q3 highlights the working capital intensity of growth at small scale. The $40 million AI revenue opportunity, while real, remains too small to fundamentally alter the cyclical profile in the near term.

The investment thesis hinges on whether Cohu can execute three simultaneous transformations: consolidate manufacturing to Asia without disruption, scale AI-related products faster than competitors, and maintain customer relationships while cutting costs. The balance sheet fortification from the convertible notes provides strategic flexibility, but also creates dilution risk. For investors, the critical variables are the pace of AI revenue growth beyond 2025, the trajectory of gross margin recovery toward management's 46.5% target, and the company's ability to win additional design wins that diversify customer concentration. If execution falter on any front, Cohu risks remaining a second-tier player in an industry where scale increasingly determines survival.