Celestica: Top AI Infrastructure Stock As Tech Earnings Heat Up

Five Mag 7 companies are set to report earnings next week, with investor focus centered on AI capex and return on investment. Four of the world's largest hyperscalers are projected to spend $650B on data centers this year, underscoring continued infrastructure buildout. Celestica is highlighted as a Strong Buy supplier to watch ahead of its Monday earnings release.

Analysis

The market is still underestimating that AI infrastructure is evolving from a narrative trade into a working-capital and balance-sheet competition. The real winner set is not just the hyperscalers, but the vendors with the fewest single-customer concentration constraints and the fastest ability to convert backlog into cash; that favors the strongest EMS/optical/networking suppliers over pure-play semiconductor exposure because lead times, integration complexity, and deployment sequencing create a longer revenue tail. Celestica sits in the part of the stack where order visibility can compound before the market gets comfortable underwriting the next leg of capex, which is why it can outperform even if top-line growth is merely in line.

The key second-order effect is that AI spend is increasingly judged on installed-base utilization, not just deployment scale. If cloud providers signal any moderation in capex growth or a shift toward monetizing existing racks, the first names to derate will be the higher-beta infrastructure suppliers priced for uninterrupted acceleration; conversely, any indication of continued demand pull from inference workloads should trigger another leg higher in the entire supply chain. In that regime, the most vulnerable shorts are companies relying on a near-term digestion phase in networking or server build cycles, because the market is likely to extrapolate AI spending for longer than fundamentals would justify.

The near-term catalyst window is the next 1-3 earnings prints: guide-up/guide-down asymmetry is large, and one good order book read-through can reset multiple quarters of expectations. The contrarian point is that consensus may be too focused on capex magnitude and not enough on mix; if spend shifts toward custom silicon, power, or software optimization, some hardware suppliers will see slower incremental margin expansion even while AI spend stays elevated. So the trade is not simply long AI exposure, but long the subset with durable share and pricing power versus the names dependent on broad-based capex beta.

Key Decisions for Investors

• Long CLS into earnings or on any post-print dip if the guide supports backlog conversion; target a 6-12 week horizon with asymmetric upside if AI order flow remains firm, but cut quickly if management signals customer digestion or margin pressure.
• Pair trade: long CLS / short a higher-beta AI-hardware beneficiary that has already priced in flawless capex acceleration; use this to express relative strength in supply-chain execution rather than outright market direction.
• Sell 1-2 month out-of-the-money puts on CLS only if implied volatility spikes materially into the print; structure for defined risk, since the setup is more about revenue visibility than a one-day event trade.
• If hyperscaler commentary shows any capex moderation, rotate out of the most stretched AI infrastructure names and into suppliers with clearer backlog and program visibility; the payoff is likely over 1-2 quarters, not days.
• Avoid chasing the broad AI basket ahead of earnings unless the thesis is specifically tied to revenue conversion rather than spend intensity; the downside scenario is a multiple reset even if AI remains strategically important.