Corning after its Q3 2026 report: sales +18%, optical fiber explains the growth and the stock is 52% below its June high

Corning reported on October 27 GAAP sales of US$4,855 million (+18% year over year) and "core" sales of US$5,081 million (+19%), the strongest growth since it launched its Springboard plan. The stock closed that day at US$126.44, 1.0% lower, and on November 2 at US$123.30: 52% below its June closing high. Three readings, none of them a verdict:

• Growth has a single engine. Of the US$755 million increase in sales, optical fiber contributed US$727 million (96%) and solar US$117 million; the rest of the business, together, subtracted US$89 million. Optical fiber went from 40% to 49% of sales and from 40% to 60% of segment profit.
• The figure the company highlights and the GAAP figure tell stories of different size. "Core" earnings per share were US$0.91 (+36%); GAAP, US$0.78 (+56%). Since April the company changed how it computes "core" and did not restate earlier quarters: the year-over-year comparison mixes two methods.
• The "adjusted" free cash flow of US$661 million includes gains on currency hedges that operating cash flow does not contain, and over nine months operating cash flow received a customer deposit of US$1,000 million. Without that deposit, operating cash flow minus capital spending for the nine months was US$841 million, close to what was paid in dividends (US$757 million).

This report uses the October 27, 2026 earnings release and the quarterly report for September 30, filed on October 29: the GAAP figures come from the quarterly report and the "core" figures from the release, which the company defines and does not audit.

This report describes what was published and the current state of the figures. It does not say what the stock will do: that reading and the decision belong to the reader.

Corning makes materials: glass, ceramics and optical fiber. Since 2026 it reports four segments plus a residual group. Optical Communications sells fiber, cable and connectors to telecom carriers and to data centers (the network that links artificial-intelligence processors). Glass Innovations combines the display glass for televisions and the cover glass for phones and laptops with specialty materials for semiconductors, aerospace and defense. Automotive makes ceramic emission filters and technical glass. Solar makes polysilicon (Hemlock), wafers and modules in the U.S. The residual group includes life sciences (pharmaceutical and laboratory glass).

Segment profit tells the same story more forcefully: of US$892 million in Q3 2026, optical communications contributed US$532 million (60%) with 47% of sales. Glass Innovations earned US$328 million (margin 22%, against 23% a year ago). Solar lost US$7 million against a profit of US$30 million: the company says production is rising "at a slower pace than anticipated" and that profitability will come with productivity.

Corning published before the open. GAAP sales went from US$4,100 million to US$4,855 million; operating income from US$589 million to US$911 million (margin 18.8%, against 14.4%); earnings per share from US$0.50 to US$0.78. For Q4 2026 the company guides core sales of US$5,200 million to US$5,300 million (≈ +19%) and core earnings per share of US$0.96 to US$1.00 (≈ +36%). It also said it signed four long-term, multibillion-dollar supply agreements in the quarter, among them with Verizon and AT&T.

Corning presents alongside the GAAP figures a "core" set that excludes certain items. The largest is the "adjustment for hedged exposures": it values in dollars, at the rates of its hedges rather than market rates, the part of sales and costs in yen, pesos, yuan, won, euros and Taiwan dollars that the company has hedged. The most important adjustment is the yen in Glass Innovations. It is a legitimate method to see the business at the exchange rate the company locked in; it is also a figure the reader cannot reproduce from the audited accounts.

What the bridge shows: "core" adds US$226 million to sales and US$0.20 per share through the hedge method, removes US$0.12 of hedge gain that the GAAP figure does include, and adds US$0.05 for tax items. That is why the gap between US$0.91 and US$0.78 is not one single thing. And it is why the GAAP gross margin barely moved (37.3%, against 37.1%) while core rose 150 basis points, to 40.5%.

The company warns that, with the new method applied to 2025, annual "core" growth would have been higher: the comparison it shows today (+19% in sales, +36% in earnings per share) uses the previous method for 2025 and the new one for 2026. That is not a flaw, but it means the year-over-year change should be read carefully.

Inside the GAAP figure there are three elements that also change from one quarter to the next. First, the translated earnings contract gain: US$138 million before taxes (US$65 million realized and US$73 million unrealized), against US$33 million a year ago. Second, stock-based compensation fell from US$99 million to US$48 million; the report attributes the US$61 million drop in selling and administrative expense to compensation tied to the share price: when the stock falls, that expense falls. That ≈ US$51 million is about 16% of the US$322 million increase in operating income (≈). Third, the effective tax rate rose from 14.5% to 22.9%.

The company presents an "adjusted free cash flow" of US$661 million for the quarter and ≈ US$2,264 million for nine months. Its definition can be rebuilt from the cash flow statement: operating cash flow, minus capital expenditures, plus realized gains on translated earnings contracts, which the cash flow statement classifies as investing rather than operating. It is its own definition, not the operating cash flow minus capital expenditures that most analysts use.

Both readings are true. With the company's definition, nine-month free cash flow is ≈ US$2,264 million and ≈ US$1,264 million without the deposit (+28% y/y). With the simple definition, operating cash flow minus capital expenditures without the deposit is US$841 million, against US$793 million a year earlier (+6%): capital spending rose 53% (US$1,299 million) and absorbed almost all the increase in operating cash flow without the deposit (US$2,140 million, +30%). Cash also benefited from US$442 million of receivables sold or collected ahead of their due date, against zero at the end of 2025.

What it means for someone reading the figures: the company invests more than it burns in order to grow and shares part of the risk with customers who advance cash. That is neither good nor bad in itself; it is what makes "free cash flow" have at least three possible figures and why it is worth always saying which one is used.

On May 6, in connection with a long-term customer partnership to help strengthen U.S. manufacturing for AI infrastructure (the report does not name the customer), Corning issued two warrants for 18 million shares (2.1% of the 863 million outstanding): a traditional one for 15 million at US$180 and a pre-funded one for 3 million at a nominal exercise price, sold for US$500 million (≈ US$167 per share). The traditional warrant was recorded at US$296 million of fair value as consideration paid to a customer: it is subtracted from sales as sales are made under the agreement, which also brought a US$1,000 million deposit with supply through December 31, 2029.

With the stock at US$123.30, the traditional warrant is 46% out of the money; the shares of the pre-funded one already count as outstanding. There are other routes open: on September 11 the company signed an "at the market" share sale program for up to US$2,000 million, of which it sold nothing in the quarter; it did not repurchase shares in 2026 (in the first nine months of 2025 it repurchased 3.1 million for US$158 million, and ≈ US$3,000 million remains authorized); and it keeps an agreement with Samsung Display under which it may repurchase up to 22 million shares through 2027. Average basic shares rose from 856 to 863 million (+0.8%).

Against that, the balance sheet improved on what the company measures: total debt of US$8,200 million (it was US$8,434 million at the end of 2025), debt to capital of 36% (it was 41%) and US$2,537 million in cash, with an unused credit line of US$1,500 million. Inventories rose 18% in nine months to US$3,646 million; days sales outstanding fell to 57 and days payable rose to 71. Net debt (US$5,663 million) equals 1.3× trailing twelve-month earnings before interest, taxes and depreciation (≈).

• Corning
• Amphenol
• Coherent
• Lumentum
• Arista Networks

Corning pays 6.4× its trailing twelve-month sales, in the same range as Amphenol, Coherent and Ciena (5.9× to 6.8×) and far from Lumentum and Arista (20.4× and 19.4×), which grow faster or have a higher margin. It has the lowest growth rate in the group (+19%) and an intermediate operating margin (15%). The comparison is of multiples; it does not say which company is well or badly priced, because each sells different things (Corning also sells display glass and polysilicon, which grow less).

The inverse method does not ask what the stock is worth: it asks what future today's price assumes. We fix the annual return the buyer would demand, start from current free cash flow without the deposit (8.2% of sales) and assume it rises to a target margin in five years; sales growth falls linearly to 4% by 2036. The table calculates the growth that fits US$123.30. It is neither a forecast nor a valuation by Quantika Research: it is the inverse reading of the price.

For scale: 2026 GAAP sales will be ≈ US$18.5 billion. With a required return equal to the cost of capital (10.9%) and a free-cash-flow margin of 16%, today's price assumes growing 16% a year for ten years, to ≈ US$81 billion in 2036 (4.4 times 2026). The company's own reference is Springboard: doubling its annualized core sales run rate from US$20,000 million to US$40,000 million by the end of 2030 (≈ +19% a year for four years). With a 20% margin and 10.9%, the model asks for ≈ US$43 billion in 2031: today's price is of the order of delivering that plan and also reaching free cash flow of 20% of sales; with a 12% margin it asks for well above the plan. A reader with another rate or margin gets another answer.

The stock closed on November 2 at US$123.30: 52% below its closing high (US$255.72, June 29; the intraday high was US$271.39), 59% above its 52-week low (US$77.41), and below its 50-day (US$147.38) and 200-day (US$155.83) averages. Between June 15 and June 29 it rose from US$187.71 to US$255.73; on July 28 it closed at US$126.02. Its 30-day volatility is 57% annualized, and its two-year beta, 1.3. A year earlier (October 20, 2025) it traded at US$85.33: it is now 44% higher.

After all five reports the stock fell the same day, between 1.0% and 12.0% (median 5.0%, average 6.0%), even in quarters with sales and guidance rising. It is a price fact, not an explanation: it does not say why it fell. In Q3 2026 the fall was the smallest of the five, and the stock rose 3.9% the next day. Among options expiring in the next 45 days, the largest open interest is in calls at the US$130 strike (9,868 contracts) and in puts at US$110 (14,006). A wall shows where positioning is concentrated; it is not a forecast.

Between May 6 and September 4, 11 executives sold 305,256 shares for ≈ US$57.5 million in 14 transactions, at prices from US$151 to US$208; the chief executive sold 100,000 of them (≈ US$18.6 million, 33% of the total), after exercising US$27 options. Together they equal 0.04% of shares. There were no open-market purchases between April 15 and September 30, the period consulted. The data do not say whether the sales followed pre-arranged plans.

What was published leaves a list of figures worth watching in the next reports. They are not forecasts: they are the questions the accounts leave open.

What these data do not answer: what the stock price will do; why it fell after each of the last five reports and 52% since June; how much of the demand for optical fiber is durable and how much is an advance purchase of capacity; who the customer behind the warrants and the deposit is; or whether Solar will become profitable. Nor do they say whether the current price is high or low: only what it assumes.