GLW — Corning Inc

As of 2026-10-15, two forces are pulling in opposite directions. Sell-side sentiment is firmly bullish: consensus rating 'Buy' (21 buys, 14 holds, 2 sells), price targets clustered at $171.25 consensus / $177.33 recent average (high $201, low $129), and the late-July Q2 2026 print beat with core sales +17.1% and core EPS +30%, Optical Communications +32% to $2.07B on AI data center demand. The tape says the opposite: GLW fell 11.17% in the week to 2026-10-15 (-3.70% on the day alone, previous close $140.89), is down 5.44% over one month, 14.28% over three months and 18.25% over six months, and sits 11.5-12.3% below its 20/50/200-day moving averages. A stock down ~50% from a 52-week high of $271.78 while earnings estimates rise is the classic profile of an AI-momentum name being de-rated, not a name the market is currently embracing - hence Neutral, not Bullish.

The three-year trajectory is a clear, structural inflection - not a blip. Revenue was stuck at $2.97-3.50B per quarter from Q3 2023 to Q4 2024, then climbed sequentially for six straight quarters: $3.45B (Q1'25), $3.86B (Q2'25), $4.10B (Q3'25), $4.22B (Q4'25), $4.14B (Q1'26), $4.51B (Q2'26). Q2 2026 revenue of $4.505B is +16.6% YoY versus $3.862B a year earlier, and the sequential step-up from Q1 to Q2 (+8.7%) is the largest in the dataset. Margins have moved even more: operating margin went from 2.6% in Q4 2023 and 5.7% in Q2 2024 to 14.8% in Q2 2025 and 15.5% in Q2 2026, while gross margin expanded from 29.2% (Q2 2024) to 36.1% (Q2 2026) - about 700bps of mix-driven expansion. Operating income nearly quadrupled from $186M in Q2 2024 to $698M in Q2 2026. EPS followed: $0.12 (Q2'24) to $0.64 (Q2'26), with a loss of -$0.14 in Q3 2024 now firmly in the rear-view. Free cash flow is the loudest signal: $1,295M in Q2 2026 versus $400M in Q2 2025 and $279M in Q2 2024. The one caveat is seasonality - Q1 is structurally weak (FCF of just $30M in Q1 2026 and -$57M in Q1 2025), so sequential comparisons should always be read Q1-to-Q1. Verdict: accelerating revenue, expanding margins, and rising cash conversion - a structural up-cycle, with the prior trough (2023-2024) clearly behind it.

Expensive on trailing numbers, merely rich on forward numbers. As of 2026-10-15, GLW trades at 61.4x TTM EPS of $2.209, 9.29x book, and 6.89x TTM sales against a five-year average P/S of 2.81 - a 2.5x premium to its own history. On the $3.29 next-FY consensus EPS, the forward P/E is roughly 41.2x (from $135.68/$3.29), which is still a premium multiple for a hardware/equipment company with a 7.5% ROIC and a 15.5% operating margin. FMP's quoted PEG of 0.463 looks artificially low and I would not lean on it. So: the 'Rich vs. its own history' tag is deserved, but it is partly an artifact - the five-year average P/E of 44.17x and P/S of 2.81x were set when Corning was a display-glass business earning trough margins, not a 32%-growth optical interconnect supplier. The honest read is that the multiple has already compressed materially from the peak (the stock hit $271.78 inside the last 52 weeks, which at ~$2.50 of 2026 EPS implies a triple-digit multiple at the top), and at 41x forward it still requires roughly 25%+ compounded EPS growth to be sustained. Not a value trap, but not a bargain either.

The price is doing something the earnings haven't yet justified on a trailing basis but may justify on a forward one. At $135.68, the market is paying 41x next-FY EPS of $3.29 while the sell-side publishes targets of $171-177 - which implies roughly 52-54x that same estimate. In other words, the analyst community's own price targets are pricing in earnings estimates that are materially higher than the ones they are publishing today; the Springboard Plan's $20B annualized revenue run-rate (versus ~$17.0B TTM) and the Amazon multiyear optical agreement are the presumed bridge. That is a real setup, but it is exactly the kind of setup that punishes any deceleration: the stock already showed what happens when the growth narrative wobbles, halving from $271.78. The practical playbook is that Q3 2026 (2026-10-27) is a binary read on whether the Q2 optical acceleration was the start of a staircase or a one-quarter spike. A beat-and-raise compresses the forward multiple and makes $135 look reasonable; an in-line quarter with soft guidance leaves the stock at 41x forward with no catalyst, which is where multiples get cut.

Clearly weak, and not yet washed out. As of 2026-10-15, RSI(14) stood at 37.86 - approaching oversold but above the 30 threshold - while price was 12.31% below the 20-day SMA, 12.34% below the 50-day SMA and 11.50% below the 200-day SMA. Being below all three with a one-week return of -11.17% is a textbook downtrend, not a pullback within an uptrend; the 200-day breach in particular marks a loss of the longer-term trend. The 52-week range of $77.39-$271.78 describes a stock that more than tripled and then gave back roughly half. Entry-timing implication: with RSI at 37.9 there is no oversold bounce signal yet, and the -11% week into the 2026-10-27 earnings print argues for patience rather than averaging down now. A more constructive setup would be either RSI below 30 with the $77-90 zone holding as support, or a decisive reclaim of the 50-day and 200-day averages on volume after the print.

• AI/hyperscaler capex cycle - the single largest swing factor, since Optical Communications is now roughly 44-46% of segment mix and grew 32% YoY on enterprise and generative-AI data center interconnect demand; any slowdown in hyperscaler spending hits Corning's fastest-growing line directly.
• China exposure and trade policy - roughly 30% of revenue is tied to China, so tariff escalation or retaliation cuts both ways even as Section 232 measures favor Corning as a US-based critical-materials manufacturer.
• Consumer electronics and auto display demand - smartphones, PCs, TVs and automotive displays remain cyclical and constitute the more mature display/cover glass base that funds the optical build-out.
• Rate sensitivity - $7.756B of long-term debt against $2.504B of cash (net debt ~$5.25B) plus a high-multiple growth equity means both interest expense and the discount rate on future earnings matter more than for a typical industrial.
• Sector rotation away from AI-adjacent hardware - the -18.25% six-month return and the 50% drawdown from the 52-week high suggest GLW has become a source of funds when AI-momentum positioning unwinds, a flow dynamic independent of its actual fundamentals.

• Q3 2026 earnings on 2026-10-27 - the first hard read on whether the Q2 optical acceleration (Optical +32% to $2.07B) sustained, plus management commentary on the Springboard Plan's $20B annualized run-rate target.
• Continued hyperscaler deal announcements and scaling updates (building on the Amazon multiyear optical agreement) plus next-generation data center interconnect architecture wins.
• Progress toward the Springboard margin target - operating margin has already gone from 5.7% (Q2 2024) to 15.5% (Q2 2026); a credible path to 20% would justify the current multiple.
• Further analyst target and estimate revisions - UBS at $201 and Truist at $182 already sit well above consensus, and upward EPS revisions toward/above $3.29 would mechanically compress the forward P/E.
• Free cash flow durability - Q2 2026 FCF of $1,295M versus $400M a year earlier; sustained $1B+ quarterly FCF would support buybacks/dividend growth and underwrite the valuation.

• Valuation risk is the dominant one: 61.4x trailing and ~41x forward earnings for a company with 7.5% ROIC and 11.2% net margin leaves no room for error, and the 50% drawdown from $271.78 is proof of how violently this multiple can reset.
• Customer/end-market concentration in AI infrastructure - Optical is Corning's growth engine and its most cyclical-new line; a pause in hyperscaler orders would hit the fastest-growing 40%+ of the business first.
• Roughly 30% China revenue exposure plus tariff and trade retaliation risk, layered on top of an already wide 52-week range ($77.39-$271.78) that shows how much headline risk this name carries.
• Cyclical consumer electronics and automotive display demand could offset data center gains if global consumer spending softens - Corning's display and cover glass base is still a large share of revenue.
• GAAP versus 'core' earnings gap - Q2 2026 core EPS of $0.78 versus GAAP diluted EPS of $0.64 means roughly 18% of reported profitability depends on adjustments; the forward consensus figure of $3.29 is closer to a core number, so the true GAAP-based forward multiple is higher than it appears.

Corning is a good business that has gotten meaningfully better and a stock that has gotten meaningfully cheaper - but not yet cheap. The operating story is real: six straight quarters of sequential revenue growth, operating margin from 5.7% to 15.5% in two years, FCF from $279M to $1,295M a quarter, and a genuine AI-optical franchise growing 32%. The problem is the price still embeds that story plus more - at $135.68 (2026-10-15) the stock trades at 61x trailing and ~41x forward earnings, versus my calculated fair value of roughly $115, and the tape is in an unambiguous downtrend with price below all three major moving averages and a 50% drawdown from the 52-week high. My read is that this is a name to accumulate into weakness rather than chase, particularly given the 2026-10-27 print is a genuine binary on optical momentum: a beat-and-raise makes today's price defensible, while an in-line quarter at 41x forward with a broken chart would invite another leg down.

A 50/50 blend of two approaches: (1) a justified forward P/E of 38x (PEG ~1.45 on ~26% expected EPS growth from the $2.209 TTM base, and a discount to Corning's own 5-year average P/E of 44.17x since that average was set on a lower-earnings, display-glass-era mix) applied to next-FY consensus EPS of $3.29 = $125; and (2) 20x forward EV/EBITDA on ~$4.8B of FY2027E EBITDA, less ~$5.25B net debt and divided by ~861M shares = ~$105. The average is $115.19, roughly 15% below the 2026-10-15 price of $135.68 and well under the $171.25 analyst consensus, which itself implies ~52x the same EPS estimate.