Buy 3 AI Infra Giants Flourishing Past Month With More Room to Run

The article highlights strong AI-infrastructure demand, citing McKinsey’s estimate of roughly $7 trillion in global AI data center capex by 2030 and $650 billion of hyperscaler AI capex in 2026, up 71.1% year over year. It recommends Ciena, Corning, and Lumentum, each supported by Zacks Rank #1 or #2, with recent gains of more than 30% and improving earnings estimates. The piece is broadly constructive for AI hardware and optical networking names, but it is primarily an analyst-pick article rather than a new company-specific catalyst.

Analysis

The market is still underestimating how much of the AI spend cycle is shifting from compute scarcity to connectivity scarcity. That is a second-order positive for optical/networking vendors because hyperscaler capex only translates into revenue once bottlenecks move from GPUs to the fabric that binds clusters together; this favors names with exposure to DCI, metro, subsea, and co-packaged/photonic architectures. In other words, the next leg of the AI trade is less about who sells the most chips and more about who monetizes the traffic explosion created by those chips.

CIEN and LITE look like the higher-beta expressions of that transition. Their upside is driven by a multi-quarter mismatch between demand and supply, which usually supports pricing power and margin expansion before it shows up in consensus revisions; that gives the stocks room to rerate even if broader AI sentiment cools. GLW is a more defensive version of the same theme: lower operating leverage, broader product diversification, and more incremental benefit from a multi-year migration away from copper make it a steadier compounder but with less torque.

The contrarian risk is that the market may already be discounting a very clean 2026 build-out, so any pause in hyperscaler ordering, timing slippage on deployments, or normalization in lead times could hit these names hard despite intact long-term demand. Another overhang is that AI capex concentration can become a customer-concentration problem: if one or two hyperscalers digest prior spend, optical order growth can decelerate faster than end-demand narratives suggest. NVDA’s low direct readthrough here is also a tell that this is not a broad AI beta trade; it is a narrower supply-chain rotation where the winners are the picks-and-shovels underneath the platform layer.

Key Decisions for Investors

• Go long LITE into any 3-5% post-rally pullback; best risk/reward is a 3-6 month hold for operating leverage to consensus revisions, with upside if hyperscaler ordering stays ahead of supply.
• Buy CIEN on weakness versus strength in NVDA/mega-cap AI; the setup is a multi-quarter share-gain story, but size smaller than LITE because near-term expectations are more elevated.
• Use GLW as the lower-volatility AI infrastructure hedge: long GLW / short a higher-beta AI hardware basket if you want exposure to optical capex without full single-name execution risk.
• For event-driven exposure, buy LITE call spreads 2-4 months out rather than stock; the market is likely to reward continued estimate revisions, but options cap downside if AI order timing slips.
• Avoid chasing the trade after another sharp leg higher; if the group stalls while consensus keeps rising, trim 25-30% of exposure and wait for a revenue-confirmation point next earnings cycle.