Lucas Downey of Tradesmith lays out the bull case for optical networking stocks as the next wave of AI infrastructure buildout. He frames AI data centers in five layers and argues the fifth — networking/connectivity — is now the bottleneck. He presents three stocks: Amphenol (APH), Corning (GLW), and Ciena (CIEN), each filling a different role in the optical ecosystem. His thesis: a multi-year shift from copper to fiber optics is underway, backed by multi-billion-dollar hyperscaler deals, and analyst estimates are still too low. He flags Ciena's recent pullback as a buying opportunity and points to earnings revisions as the core signal.
Watch on YouTube ›Get the market thesis, key claims, assets, contradictions, and follow-up questions from any financial video — then unlock a version personalized to your portfolio, watchlist, and favorite speakers.
Lucas Downey, a contributor at Tradesmith, frames the AI market through a five-layer model of data center buildout: (1) infrastructure/land, (2) power, (3) cooling, (4) compute and memory, and (5) networking and connectivity. The interview focuses on layer five, where he argues a global shift from copper to optical (glass/fiber) links is creating a multi-year investment thesis. The driver: fiber optics offer lower latency, less heat, and cost savings — advantages that hyperscalers are now scaling aggressively. The three stocks he presents each address a distinct segment within optics. Amphenol (APH) — a $200B market-cap connector and interconnect manufacturer — just posted a standout quarter: $7.62B in revenue vs. $7.08B consensus, with an EPS beat and raised guidance. Downey emphasizes the magnitude of the beat, not just the streak, and thinks analysts are still playing catch-up. …
Near-term: rebalancing-driven weakness in high-flyers like Ciena creates tactical entry points; with earnings season weeks away, the setup favors optical names that are likely to beat and raise, following the pattern established by Micron's repeated upside surprises.
Medium-term: the optical buildout is a multi-year capacity expansion, not a 2026 story — Corning and Amphenol have visibility through at least 2028 via signed hyperscaler deals; the key risk is whether analyst revisions can keep pace with actual demand or whether expectations get front-run.
Long-term: the copper-to-fiber transition is framed as a secular infrastructure upgrade within a trillion-dollar AI capex cycle — if that holds, optical networking becomes a durable structural allocation rather than a cyclical trade, but the thesis depends on sustained hyperscaler spending and no in-house displacement.
As long as analysts continue to revise earnings estimates higher for these AI infrastructure companies, the stocks will be rerated higher after shallow corrections.
The speaker argues that the pattern of upward earnings revisions will drive stock prices higher following pullbacks, citing the historical trend of analysts being wrong about companies like Micron.
Corning's sales run rate will reach $20 billion by end of this year, $30 billion by 2028, and $40 billion by 2030.
Speaker references Corning's own presentation at a JP Morgan conference projecting these revenue milestones.
Ciena's full-year 2026 EPS estimates show strong growth: ~$6.53, rising to $9.65 next year, then ballooning to $14.28.
The speaker provides specific numerical EPS projections for Ciena over the next several years to support the bullish thesis.
How much room is still left in the AI market, and is there still opportunity for investors?
Lucas Downey says AI has been adopted by Wall Street and that there are multiple layers to the data-center buildout. He argues the opportunity is still broad because infrastructure, power, cooling, compute, memory, and networking are all still being built out.
What do the optical stocks in this group have in common, and why is demand so strong right now?
He says the AI infrastructure capex is approaching a trillion dollars and that networking is shifting from copper to optical links. The appeal is lower latency and less heat, which reduces cooling costs, and some firms are also getting multi-year hyperscaler deals.
Are these optical companies actually new technology, or are they older businesses benefiting from renewed demand?
He says some of these companies have existed for more than a hundred years, but they are now seeing renewed demand as data-center builders adopt newer optical technologies. He also breaks the optics ecosystem into layers such as light generation, fiber travel, signaling, physical connections, components, and cleaning.
Unlock the full claims, asset map, scores, related transcripts, follow-up questions, and AI chat — shaped around your portfolio, watchlist, favorite speakers, and risks.
