Shares of Celestica surged as investors piled in for a second straight session, extending a rally that has added roughly 16.7% since May 27. The catalyst: management raised its 2026 annual outlook to $19.0 billion in revenue and $10.15 in adjusted earnings per share — a massive jump from the $17.0 billion revenue and $8.75 adjusted EPS target set just 90 days earlier in January . At $409.53, the stock is pricing in near-flawless execution on an AI-infrastructure thesis that has reshaped this former contract electronics assembler into a Wall Street darling.

  • A 53% Revenue Surge Shows the AI Pipeline Is Real. Q1 revenue hit $4.05 billion, up 53% year-over-year , driven by the Connectivity & Cloud Solutions segment, which grew 76% to $3.24 billion . That segment builds the networking switches, server racks, and cooling systems hyperscalers need for AI data centers. Adjusted EPS of $2.16 beat the top of its own guidance range and rose 80% from a year ago . This isn't a guidance bump on hope — orders are arriving.

  • Margins Are Expanding, Not Just Revenue. Adjusted operating margin reached a record 8.0%, up from 7.1% a year earlier . For a hardware manufacturer historically earning low-single-digit margins, that improvement signals Celestica is capturing higher-value work — designing complex liquid-cooled AI racks and custom networking platforms — rather than merely assembling commodity boards.

  • The Valuation Already Reflects a Lot of Good News. Celestica's trailing P/E ratio — the price investors pay per dollar of past earnings — stands at 46.58 , meaning the market values each dollar of earnings at nearly 47 dollars of stock price. Its 10-year average is just 16.78, making the current level 176% above that historical norm . On a forward basis, using the new $10.15 EPS target, shares trade at roughly 40 times this year's expected earnings. Analysts' consensus price target sits at $408.73 — essentially where the stock trades today.

  • Big Spending Comes With Big Risk. Celestica plans to lift 2026 capital expenditure to roughly $1 billion to build out North American manufacturing capacity for AI hardware. High customer concentration — a handful of cloud giants drive the bulk of orders — means any slowdown in hyperscaler spending would hit hard. The stock has quadrupled from its 52-week low; the margin for error at this altitude is razor-thin.