Celestica (CLS) Q3 2026 Earnings Call Transcript

Celestica reported Q3 2026 revenue of $5.83 billion, up 83%, and adjusted EPS of $3.30, up 109%; both exceeded the high end of guidance. It raised its 2026 revenue outlook to $21.0 billion and adjusted EPS outlook to $11.55, and forecast 2027 revenue of $40.0 billion, up 90%, with adjusted EPS of $21.50. Management cited AI infrastructure demand, while warning that 2027 investment in capacity and working capital will mean net cash use; 2026 free-cash-flow guidance was reduced to $460 million from $600 million.

Analysis

The key underwriting question is conversion, not demand: CLS is adding capacity and inventory ahead of a sharply second-half-weighted 2027 ramp, while management expects cash use and only targets positive free cash flow again in 2028. That makes execution and working-capital discipline the near-term equity catalysts; revenue growth alone may not translate into comparable cash earnings. The modest planned operating-margin expansion relative to the scale of the forecast also leaves the thesis dependent on mix and utilization, not just volume.

The moat is most valuable where design, qualification, and scarce powered factory space are bundled. That can raise switching costs and favor CLS over less integrated manufacturers such as Flex and Jabil, but capacity is not a durable advantage if customer programs slip or customers rebalance production. Top-three customer exposure makes a single hyperscaler’s schedule, financing, or sourcing decision a material company-level variable. Power, land, and geopolitical constraints could delay ramps even with end demand intact.

For AMD, Broadcom, and Alphabet, the read-through is directionally supportive of custom compute and Ethernet infrastructure demand, but CLS’s program disclosures do not establish material incremental earnings for any of them. Longer term, 1.6T and rack-scale integration widen CLS’s content opportunity; CPO remains an execution option rather than a near-term base-case driver.

Consensus may underweight cash conversion and ramp timing while extrapolating headline growth. Conversely, a miss against the 2027 plan could trigger multiple compression because expectations now rely on a large, concentrated buildout. Watch quarterly inventory and receivables versus revenue, operating margin, and evidence of customer shipment schedules. Thesis weakens materially if awarded-program ramps slip, margins fail to progress toward guidance, or working-capital use exceeds available liquidity assumptions.

Key Decisions for Investors

• Prefer staged long exposure to CLS rather than chasing an earnings-driven gap; consider a defined-risk call spread only after checking implied volatility and liquidity. The upside case is further confirmation that design-led programs convert to shipments and margins; the principal downside is forecast de-rating on ramp delays or cash absorption. No numeric target is justified without current valuation and options data.
• For the next 1–3 months, track reported revenue against the ramp cadence, operating margin, and inventory/receivables growth. Add only if growth is accompanied by stable-to-improving margins and working capital begins to normalize; reduce if sequential inventory and receivables outpace revenue for multiple quarters or guidance is cut.
• Treat AMD, Broadcom, and Alphabet as watchlist read-throughs, not automatic pair-trade shorts or longs: verify program attribution and their own earnings sensitivity before trading. Key thesis falsifiers are delayed 2027 customer ramps, a margin shortfall versus the 9% outlook, or a meaningful increase in financing needs beyond stated liquidity.