Celestica Q3 Earnings: Setting Records, And 3.2T Switches Are The Next Catalyst

Celestica exceeded Q3 estimates, with revenue up 83% year over year and EPS up 109%, driven by surging AI hardware demand. Compute & Storage, now 28% of revenue, grew 242% year over year, well above its annual growth forecast. Management raised 2026 and 2027 guidance; the 2027 revenue target rose from $35B to $40B and the EPS target to $21.50.

Analysis

The key underwriting question is whether AI-related demand converts into durable, cash-generative earnings—not whether Celestica can post another strong quarter. A rising Compute & Storage mix increases sensitivity to hyperscaler investment plans, customer program timing and component availability. It can also raise execution demands: rapid revenue growth may absorb working capital, while competitive bidding by contract manufacturers such as Quanta, Foxconn and Jabil could limit margin expansion. Verify gross-margin progression, cash conversion, backlog quality and customer concentration before treating the higher outlook as fully de-risked.

Over the next few days, positive guidance revisions may support momentum, but the stock’s reaction depends on how much acceleration was already priced in; valuation and positioning data are not provided. Over 1–3 months, earnings-call detail on orders, capacity and margins is more informative than headline growth. Over 6–18 months, the structural risk is that hyperscalers slow or rebalance data-center spending, shift more designs in-house, or favor alternative compute architectures—leaving suppliers with capacity ahead of demand. Component constraints or customer program delays could also defer revenue.

Contrarian view: the raised targets may encourage investors to extrapolate a steep growth curve, while the less visible risk is earnings quality and customer concentration. The bullish thesis weakens if growth is accompanied by deteriorating margins or cash conversion, or if management trims its outlook; sustained order/backlog evidence and stable profitability would strengthen it.

Key Decisions for Investors

• Consider a staged long in CLS on a post-results pullback rather than chasing a gap higher; size against the risk that AI program timing or customer spending resets. Reassess if management cuts its outlook or subsequent results show material margin or cash-conversion deterioration.
• Use the next 1–3 months to verify backlog/order quality, customer concentration, gross-margin trend, working-capital needs and capacity constraints. Treat these as gating data, not assumptions, before increasing exposure.
• Monitor hyperscaler capex commentary and demand signals across AI infrastructure suppliers. A broad spending slowdown would challenge the thesis even if CLS-specific execution remains sound; persistent demand alongside stable margins would support holding through near-term volatility.
• Avoid a relative-value short against a named contract-manufacturing competitor without comparable valuation, margin and customer-program data; the article does not establish which peer is most exposed or mispriced.