Celestica: Irrational Pullback Gives Another Buying Opportunity

Celestica posted strong Q1 results, with revenue up 53% and EPS up 80%, underscoring robust hyperscaler demand. The most important upside driver is CCS and HPS, where 1.6T switching, liquid cooling, and future CPO programs could expand revenue, improve mix, and lift margins further. While margins may be near a near-term peak, accelerating capacity ramps and richer product content support continued earnings growth through 2027 and beyond.

Analysis

CLS is increasingly behaving like a scarce-capacity levered proxy on AI infrastructure intensity rather than a generic electronics manufacturer. The second-order bull case is mix shift: as liquid cooling, higher-speed networking, and future CPO programs take share, revenue should compound faster than unit volume because content per rack is rising faster than the installed base. That means the market may be underestimating how much of the growth can come from richer BOMs, not just more racks shipped.

The key competitive dynamic is that hyperscaler capex is forcing suppliers with credible thermal and interconnect capabilities into a winner-take-more structure. Smaller peers and contract manufacturers without the same engineering depth should see worse pricing power and lower attach rates, while upstream component vendors with proprietary optics, connectors, and thermal hardware may get pulled through in CLS-led programs. The flip side is that this can compress the ecosystem’s available margin pool over time if customers use scale to renegotiate once the new platforms become standardized.

The main risk is not demand but timing: margin expansion can look strongest during the transition, then flatten once capacity catches up and customers press for price concessions. On a 3-6 month horizon, any hyperscaler digestion pause or delayed ramp on next-gen platforms could trigger multiple compression because the stock is likely discounting several years of above-trend earnings growth. Longer term, the bear case is that CPO and liquid cooling become more modular than expected, allowing OEMs to re-source and normalize margins by 2026-2027.

Consensus still seems too anchored on near-term gross margin pressure and not enough on the durability of earnings power once mix shifts. The stock can still work even if margins peak soon, because the bigger driver is numerator growth from higher content and faster capacity utilization. The asymmetry is that the market may be pricing a cyclical peak, while the business is actually moving into a structurally higher-revenue, higher-barrier segment of the AI supply chain.

Key Decisions for Investors

• Long CLS on any post-earnings dip over the next 1-3 weeks; use weakness as entry if the selloff is driven by margin-peaking concerns rather than order commentary. Risk/reward favors a 12-18 month hold because the next leg is mix-driven EPS acceleration, not just multiple expansion.
• Buy CLS Jan-2027 call spreads to express the view that earnings can compound through the platform transition while limiting downside if hyperscaler spending pauses. Best structure: slightly out-of-the-money debit spread with 2-3x payoff if the market re-rates duration on AI infrastructure.
• Pair trade: long CLS / short a lower-quality EMS or generic hardware name with weaker AI attach rates. The thesis is that capital will keep flowing to suppliers that can capture thermal and interconnect content, while commoditized names face margin compression as pricing power shifts upstream.
• If already long CLS, hedge with a short-dated collar into the next quarter: near-term valuation is vulnerable if investors focus on margin normalization, but the longer-dated upside remains intact. This preserves participation in the 2026-2027 thesis while reducing event-risk drawdown.