GLW: Corning Is the Quiet Monster of the AI Trade

Everyone's been focused on the chips. Fair enough — Nvidia gets the headlines, the memes, and the eye-watering multiples. But while the world was debating whether AI is a bubble, a company that's been making glass since 1851 quietly became one of the best-performing stocks in the S&P 500. Corning (NYSE: GLW) is up nearly 200% over the past twelve months, and the bull case isn't some speculative fever dream. It's grounded in contracts, capacity, and a technology moat that most investors still haven't fully appreciated.

Let me explain why I think GLW still has a meaningful runway from here.

The Setup: What Corning Actually Does

Most people know Corning as the company that makes the glass on your iPhone. That's true — Gorilla Glass is a real business, and Apple has a $2.5 billion commitment for 100% of iPhone and Apple Watch cover glass. But that's not the story. The story is fiber.

Corning's Optical Communications segment is the engine right now. It makes the high-density fiber cables that connect servers to each other inside AI data centers — the physical plumbing that makes a 100,000-GPU cluster actually function as one coherent system. Every time a hyperscaler adds another rack of Nvidia H100s or GB200s, they need more of Corning's fiber to wire it together. There is no substitute at scale. That's the moat.

The Numbers That Matter

Full-year 2025 was a breakout year. Core sales rose 13% to $16.41 billion. Core EPS grew 29% to $2.52, well ahead of what the Street expected coming into the year. The Optical Communications segment alone posted $6.3 billion in revenue — up 35% year-over-year — with enterprise data center revenue specifically growing 61%. That last number is the one to watch. It's not telecom infrastructure or consumer broadband. It's AI data centers, the fastest-growing capital expenditure category in the world right now.

Margins are moving in the right direction too. Core operating margin hit 20.2% in Q4 2025, up 390 basis points over two years. Adjusted free cash flow nearly doubled to $1.72 billion for the full year. This is a company that's not just growing revenue — it's converting that growth into real cash.

| Metric | FY2024 | FY2025 | Change | |---|---|---|---| | Core Sales | $14.5B | $16.41B | +13% | | Core EPS | $1.95 | $2.52 | +29% | | Optical Comms Revenue | $4.67B | $6.3B | +35% | | Enterprise Data Center Revenue | — | +61% YoY | — | | Core Operating Margin (Q4) | ~16.5% | 20.2% | +390 bps | | Adjusted Free Cash Flow | ~$0.9B | $1.72B | ~+91% |

The Meta Deal Changes the Math

On January 27, 2026, Corning announced a multiyear supply agreement with Meta worth up to $6 billion — the largest deal in the company's 175-year history. Meta is building out AI data centers across the U.S. at a pace that requires a guaranteed anchor supplier for optical fiber. Corning is that supplier.

The day after the Meta announcement, management upgraded their internal "Springboard" strategic plan. The original version targeted $6.5 billion in incremental annualized sales by end of 2026. The upgraded version now targets $11 billion in incremental annualized sales by 2028. That's not a rounding error. That's management telling you they can see the demand, they have the contracts, and they're building the capacity to meet it.

CEO Wendell Weeks put it plainly on the Q4 earnings call: the company is expanding its manufacturing footprint in Hickory, North Carolina specifically to serve this demand. When a 175-year-old industrial company starts rapidly expanding capacity, it's because the orders are already in hand.

The Technology Angle: Multicore Fiber

Here's the part of the story that most people are sleeping on. At the Optical Fiber Communication Conference (OFC) in Los Angeles this month, Corning unveiled its Multicore Fiber (MCF) technology. A single strand of MCF packs four separate fiber cores, quadrupling data capacity without taking up any additional physical space in a data center rack.

Data center racks are already running out of room. Hyperscalers are trying to pack more GPU compute into the same square footage, and the fiber interconnects are becoming a physical bottleneck. MCF solves that problem. It also cuts installation time by up to 60%, which matters when you're trying to bring a 100,000-GPU cluster online in six months instead of twelve.

Corning is also working on Hollow Core Fiber (HCF), where light travels through air instead of glass, reducing signal latency by nearly 50%. That's still in the pilot phase, but it represents the next product cycle — and Corning is already ahead of the competition on it.

What Wall Street Thinks (and Where I Disagree)

The Street consensus price target sits around $126, which is roughly where the stock trades today. Most analysts are at Moderate Buy or Hold. Bank of America is the outlier — they raised their target to $144 last week, citing a $10.3 billion scale-out revenue opportunity by 2030 and an EPS path to $2.42 by that year. UBS is even more aggressive at $160.

Here's my take: the consensus is anchored to a version of Corning that no longer exists. Analysts who cover industrial technology companies tend to apply conservative multiples because these businesses historically had lumpy revenue and thin margins. But Corning's margin profile has structurally improved — 20% core operating margins are not a one-quarter fluke — and the revenue is increasingly locked in via multi-year contracts rather than spot orders. That combination deserves a higher multiple than the market is currently giving it.

The $11 billion Springboard target, if achieved by 2028, would represent a company generating roughly $27-28 billion in annual revenue with 20%+ margins. At a 20x earnings multiple — not aggressive for a high-quality compounder with visible growth — that's a stock worth considerably more than $144.

The Risks (Because There Are Always Risks)

I want to be straight with you: this is not a risk-free trade. The stock has already run nearly 200% in a year, so you're not buying a sleeper. Valuation is stretched relative to Corning's historical averages.

The bigger risk is hyperscaler capex. If Meta, Microsoft, Amazon, or Google pull back on AI infrastructure spending — whether because of a macro slowdown, a regulatory shock, or a genuine reassessment of AI ROI — Corning's growth story slows materially. The $6 billion Meta deal is a floor, not a ceiling, and the floor only holds if Meta keeps building.

There's also competitive risk. Chinese fiber manufacturers have been ramping capacity, and while Corning's specialty AI products are technically differentiated, commoditization is always a long-term threat in materials businesses. Hollow Core Fiber and Multicore Fiber keep Corning ahead for now, but "for now" is doing a lot of work in that sentence.

The Bottom Line

Corning is not a momentum trade. It's a company that spent a decade quietly building the technical capability to become indispensable to the AI infrastructure buildout, and now the world is figuring that out. The $6 billion Meta deal, the upgraded Springboard targets, the 61% enterprise data center growth, the new multicore fiber technology — none of this is speculative. It's happening.

At current prices around $130, you're paying a fair price for a company with locked-in revenue, improving margins, and a technology roadmap that keeps it ahead of competitors for the next several years. Bank of America's $144 target is reasonable. UBS's $160 is achievable if the Springboard plan executes on schedule. And if the AI infrastructure buildout continues at its current pace — which every major hyperscaler's capex guidance suggests it will — the upside case is higher than either of those numbers.

GLW is the kind of stock you want to own before everyone else figures out it's not just a glass company anymore.

This is editorial analysis and reflects the personal views of the author. It is not financial advice. Always do your own research before making investment decisions.