Rothschild Redburn initiates Celestica stock with buy on AI growth By Investing.com

Celestica was initiated at Buy by Rothschild Redburn with a $460 price target, while CIBC and RBC lifted targets to $480 and $440, respectively. The company also beat Q1 2026 EPS expectations at $2.94 versus $2.82 and revenue at $5.51B versus $5.49B, despite component shortages. The bullish case centers on >50% revenue growth, AI/rack-scale demand, and new 1.6TbE switch orders powered by Broadcom silicon.

Analysis

The market is starting to price Celestica less like a contract manufacturer and more like a levered picks-and-shovels play on the next wave of AI network infrastructure. The key second-order effect is that standardized, open-software switching architectures compress the moat of traditional OEM networking vendors while expanding the addressable share for whoever can industrialize fast enough around Broadcom silicon and hyperscaler specs. That favors CLS near term, but it also means the value pool may migrate from software differentiation to execution, supply chain priority, and program management.

The bigger upside driver is not this quarter’s beat; it is the option value embedded in 2026-2027 hyperscaler capex and a probable step-up in rack-scale integration demand. If that ramp lands, CLS could see a multi-quarter mix shift toward higher-complexity, higher-dollar programs, which should support both gross margin and multiple expansion. The risk is that this becomes a crowded “AI infrastructure” ownership trade where any delay in customer ramps or a normalization in network capex can punish the stock quickly because expectations are now stretched several quarters forward.

For Broadcom, the indirect benefit is underappreciated: the more the market standardizes around its switching silicon, the more it cements pricing power and ecosystem lock-in even if OEM hardware economics get competed away. AMD benefits only at the margin through rack adoption validation; the cleaner read-through is that any material Helios delay would hit sentiment on end-market demand rather than AMD fundamentals. The contrarian view is that investors may be extrapolating one cycle of hyper-scaler spend into a durable step-function, when in reality networking ramps often come in bursts followed by digestion periods.

The main reversal catalyst is not macro; it is execution friction — component allocation, customer qualification slippage, or a lower-than-expected conversion rate from design win to volume shipment. If that shows up, CLS can de-rate rapidly because the stock is likely trading on forward narrative more than near-term earnings power. In that scenario, the winners would be the most vertically integrated incumbents with software/control-plane stickiness, not the hardware assembler closest to the demand spike.

Key Decisions for Investors

• Long CLS on a 3-6 month horizon, but treat it as a momentum-with-fundamentals trade: add on pullbacks only if volume remains strong; upside remains attractive into 2026 ramp, but stop tightly if guidance fails to confirm customer conversion.
• Pair trade: long CLS / short a basket of slower-moving network hardware names with weaker AI exposure over the next 1-2 quarters; thesis is share shift toward execution-oriented suppliers as standardized Ethernet programs scale.
• Add AVGO as a lower-beta beneficiary of the same architecture shift over 6-12 months; the risk/reward is better than CLS if you want ecosystem exposure with less single-customer execution risk.
• Avoid chasing AMD here unless you have a separate view on compute demand; the read-through is incremental and sentiment-driven, so use any strength to trim rather than add.
• For event risk, buy short-dated CLS downside protection into any sharp post-upgrade rally; if the market is pricing the 2027 story too early, a 10-15% drawdown on a missed ramp would be a high-probability reset.