Celestica: Time To Bet Bigger

Celestica was upgraded from Buy to Strong Buy as fundamentals improve and a recent price dip makes valuation more attractive; forward FY2028 P/E is now well below 20 and non-GAAP PEG is 0.7. The company is positioned to benefit from sustained hyperscaler AI infrastructure spending and is expanding its moat via a strategic partnership with AMD on the Helios rack-scale AI platform, supporting upside to the stock.

Analysis

CLS's move into rack-scale and hyperscaler-driven systems shifts the economics from one-off box builds to higher-ASP, more recurring engineering and services revenue; that change increases gross margin leverage but also front-loads working capital and yield risk during ramp. The tighter integration with a major chip vendor materially raises switching costs for customers — over 18–36 months that can translate into multi-year revenue visibility for the systems integrator but also concentrates demand risk around a handful of hyperscalers.

Downstream, suppliers of high-density power, cooling, and advanced interconnects should see outsized order growth as BOM density rises: expect faster order cadence for power-supply and thermal vendors and meaningfully higher HBM/advanced DRAM content per chassis if designs favor GPUs/accelerators. The competitive losers in the near term are smaller EMS players that lack scale to absorb engineering and warranty volatility; their margins will compress if they chase volume with higher fixed-cost footprints.

Key reversals that would unwind the thesis are not macro price moves but execution and customer-concentration events — a single hyperscaler design pivot or a sputtering yield ramp could flip margin tailwinds to headwinds within 2–6 quarters. Over 2–4 years the real test is whether design wins translate into recurring service contracts and >100bps incremental EBIT margin expansion rather than temporary ASP uplifts; monitor customer diversification and component lead times as the earliest hard evidence.

Key Decisions for Investors

• Core long CLS (2–3% portfolio) — accumulate on 5–10% pullbacks; target 30–50% upside over 6–12 months if quarterly revenue mix shifts toward higher-margin rack systems; tighten to 12–15% stop-loss on entry to limit execution risk.
• Pair trade: long CLS / short FLEX or JBL (equal notional, 12 months) — directional play on scale advantage and margin expansion; expected payoff if CLS captures rack wins and smaller EMS lags, asymmetric reward ~2:1 with pair reducing macro beta.
• Options hedge: buy 12-month ATM CLS LEAP calls sized at 0.5–1% portfolio (replace part of outright equity); fund premium by selling 3–6 month covered calls on a portion to monetize near-term optimism; goal: >3x upside if market re-rates on tangible design-win cadence.
• Tactical long AMD (small weight, 6–18 months) — a modest exposure to the silicon partner captures any share gains in accelerator CPU/GPU content; take profits on a 25–35% move and re-evaluate on quarterly design-win disclosures.