Execution and demand visibility look strong into 2026–2027, but the 2025 re-rating raises the bar: to outperform, the company must sustain ramps and attach higher-margin services while avoiding cycle and working-capital shocks.

Celestica can keep compounding through 2030 by being the execution layer for AI infrastructure (co-design → qualify → ramp → integrate → lifecycle), but shareholder upside is increasingly a multiple-management story after the 2025 re-rate: deliver 2026–2027 ramps, add higher-margin attach services, and avoid capex/working-capital accidents.

AI buildouts shift value to reliable industrial execution, supply assurance, and security/traceability—areas where Celestica is strong; moats are still relationship/execution, not platform network effects.

Celestica’s near-term fundamentals are strong (management guided to a large 2026 step-up driven by AI networking/compute), but the market has already re-priced the stock from “EMS” to “AI infrastructure partner.” Into 2030, upside should come from staying inside the highest-complexity programs (next-gen switching, custom compute, rack integration) and attaching repeatable services (validation/warranty, security/traceability, lifecycle). Offsetting that, investors will likely demand a lower sales multiple as revenue scales and end-demand looks more cyclical. Net: solid growth, but not a clean hypergrowth multiple expansion story.

The dominant risk is a regime shift from AI buildout to digestion that causes utilization swings and rapid multiple compression. Second-order risks are hyperscaler concentration, working-capital whiplash, and taking on warranty/JV structures that increase balance-sheet and geopolitical exposure without durable pricing power.

Upside is now more constrained by valuation after the 2025 re-rate; we weigh multiple compression more heavily than in the prior period.

$383.61