Built by Grok 4.6 High

Reports Tuesday, October 27, 2026. Prepared Thursday, September 24, 2026. This is the Q3 2026 print (quarter ended September 30). It has not been reported. The date is the working calendar date, not an IR lock: Corning’s events page has no dated earnings event, and the last several prints were before the open with the call at 8:30 a.m. ET. Last reported quarter is Q2 2026, released the morning of July 28, 2026. Corning is a calendar-year company. The homepage calendar labels October 27 as 2026Q4. That label is one quarter ahead of the quarter being reported. This page is 2026Q3.

GLW | Earnings Preview

Corning reports Q3 on Tuesday, October 27. The July 28 framework is still the only filed guide: core sales of $4.9 billion to $5.0 billion and core EPS of $0.85 to $0.89. Nothing filed since that morning has replaced it. Q3 has not printed. Every sales and EPS figure management guides is core, not GAAP. Q2 core sales were $4,738 million. GAAP net sales that quarter were $4,505 million. FMP’s live Q3 estimate sits inside the core range, so the street number below is read as core.

Sales growth stepped up, then stalled at a high rate. EPS growth has not stalled. Core sales went from +14.4% in Q3 2025 and +13.9% in Q4 2025 to +18.1% in Q1 2026 and +17.1% in Q2 2026 ($4,738 million versus $4,045 million). That is a 100 bp cool-down inside the first half, not a break. Core EPS kept speeding up: +24.1%, +26.3%, +29.6%, +30.0%, the last of those on $0.78 versus $0.60. The Q3 low end gives the sales step-up back (+14.7% versus Q3 2025 core sales of $4,272 million). The high end holds Q2 (+17.0%). Ed Schlesinger told the July call the guide was not a deceleration and was meant to look like Q2. That sentence is the high end, not the $4.95 billion midpoint. The spoken “approximately 28%” on EPS is the low end (+26.9% versus $0.67). The midpoint, $0.87, is +29.9%, which matches the last two prints.

The street has already used the room. FMP on September 24 is Q3 sales of $4,999 million (5 analysts) and EPS of $0.878 (7 analysts). Sales are $1 million under the $5.0 billion ceiling and +1.0% versus the midpoint. EPS is +0.9% versus the $0.87 midpoint and $0.012 under the $0.89 high. On August 1 the same October 27 row was $0.85, the floor, and $4.983 billion. The range did not move. The estimate did. A normal Corning print — tag the sales high, land on or a penny through the EPS high — is a sales meet and a small EPS beat. A midpoint print would miss the street on sales and would be the first time in ten quarters core EPS did not reach the company’s own high end.

Margins are still expanding. The pace of that expansion is not. Core operating margin went 19.6% to 20.2% to 20.2% to 20.9%. The year-over-year delta peaked at +220 bps in Q1 and cooled to +190 bps in Q2. The standing line is “at or above 20%,” with no new target until later this year. That floor is 90 bps under the last print. It is not the hurdle. Core gross margin was negative year over year in both second-half 2025 quarters and is +120 bps in both 2026 quarters, to 39.6%. The delta has stopped widening. Schlesinger accepted that the Q3 guide implies a gross-margin step-up. The driver he named is solar profit improving after the wafer shutdown, not a new optical-margin regime. Optical net margin is already about 21% and was about 21% in Q1 as well.

The mix, not the range, is the argument. The consolidated band is only $100 million wide. Enterprise optical was $1,269 million, +65.0% versus $769 million, faster than Q1’s +36% and faster than Q3 2025’s +58.3%. Carrier, inside the same segment, collapsed from +36% in Q1 to +0.8%. Glass Innovations was +1.4%, and Q3 handheld was guided “significantly weaker than the first half” on memory, against a Q3 2025 glass comp of $1,560 million. Solar sales were $438 million, +89.6%, and solar net income was a −$7 million loss. Repeating Q2’s mix on the Q3 2025 base lands on about $5.0 billion. A return of enterprise toward the mid-30s, or a glass book that falls harder than content per device can cover, does not.

September aimed sales at the high end. It did not raise the range, and it did not pull 2027 into this quarter. Accounts of the September 9 Citi conference have Schlesinger putting Q3 at the high end of the July sales guide or slightly better, with the $20 billion annualized run-rate a quarter early and Q4 larger than Q3. That is not an 8-K, and Daloopa has no Q3 2026 documents. Verizon (September 8, fiber miles from 2027 through 2032), Zayo, the optional $2 billion at-the-market program, and Section 232 polysilicon protection (effective December 4) do not reset $4.9–$5.0 billion.

What to watch on October 27

• Core sales at or through $5.0 billion, about +17% versus $4,272 million. The midpoint is the print that breaks a nine-quarter pattern and misses a street already on the ceiling.
• Core EPS at or through $0.89. Versus $0.878 that is about a +1.4% beat. A penny through, the Q2 pattern, is about $0.90. The $0.87 midpoint would miss the street.
• Enterprise still well above Q1’s +36%, preferably still near +65%, versus the $831 million Q3 2025 base. Optical net margin holding about 21%.
• Solar net income no longer a loss, against −$7 million. That was the named source of the margin step-up. A profit at the company average is a 2027 statement, not a Q3 test.
• Glass within a couple of points of flat, not a mid-single-digit decline. Consensus has about $1 million of room under the sales high. Glass can use that room up.
• Do not score GAAP. Q2 GAAP EPS of $0.64 versus the $0.755 core estimate would have been a miss. FMP’s EPS actual matches core.

Optical pure-plays report after Corning. Lumentum, Coherent, Fabrinet, and Amphenol are not a pre-read. Ciena already beat on the September 3 FMP row. The names that can still change the setup are Nokia (October 22), Celestica (October 26), Ericsson (October 15, carrier capex), and Micron (September 30, the memory price that management tied to handheld).

Classification: conservative versus Corning’s own history, not conservative versus today’s street on sales. Nine closed quarters, results from Q2 2024 through Q2 2026, finished at or above the guided sales point or high and at or above the guided EPS high. The low end of the EPS range has not been an outcome. From Q1 2024, core EPS has been at or above the high end in 10 of 10 quarters. The published “about 16% / about 28%” language still sits under where the company has landed. FMP sales now sit on the ceiling, so repeating the recent pattern is a consensus meet on sales. EPS consensus is still inside the range.

The July 28 words, from the call and the same-day 8-K: core sales “approximately 16%” to $4.9–$5.0 billion and core EPS “approximately 28%” to $0.85–$0.89. The 16% matches the sales midpoint (+15.9% versus $4,272 million). The 28% matches the EPS low end (+26.9% versus $0.67), not the EPS midpoint (+29.9%). On the same call Schlesinger said the guide is “similar” year-over-year growth to Q2, when sales grew +17.1% and EPS grew +30.0%. A $5.0 billion quarter annualizes to $20 billion. He said the guide implies the end-2026 run-rate “a quarter or so early.” That sentence is the high end, not the $4.95 billion mid ($19.8 billion annualized).

FMP, not Visible Alpha and not Bloomberg. Averages as of September 24. The earnings-calendar row for October 27 rounds the same quarter to EPS $0.88 and revenue $4,999,016,000. Actuals are null. FMP does not publish Optical, Enterprise, Carrier, Glass, or Solar consensus. FMP Q3 EBIT of $1,149 million on $4,999 million is a 23% margin. That is not Corning’s core operating margin. Do not read FMP EBIT or EBITDA as the core-margin guide.

Q3 compares are versus core sales of $4,272 million and core EPS of $0.67. Q4 compares are versus $4,412 million and $0.72. FY2026 compares are versus $16,408 million and $2.52. The entire street EPS band ($0.858–$0.888) sits inside the company range. The street sales high ($5,039 million) sticks out above the company high. First-half 2026 core sales of $4,345 million plus $4,738 million are $9,083 million, +17.6% versus the first half of 2025. The $19.230 billion year leaves about $10.15 billion for the second half, +16.8% versus second-half 2025. The street’s year is a small second-half slowdown from the first-half rate, not a break.

Current quarter — Q3 2026

Quarter not reported. The only lines with a low and a high are core sales and core EPS. Segment dollars, gross margin, and EBITDA are not guided. Internal is n/a. Percent difference is (consensus minus guide mid) divided by the absolute mid, and only where both exist.

Eight quarters of the core P&L. Columns are periods. Year-over-year rates use the prior-year quarter, not the prior quarter. Q3 2026 is not a column.

Sales reaccelerated into 2026 and then gave back 100 bps. EPS growth is still rising, from +24.1% to +30.0% across the last four prints. Gross-margin expansion turned positive in 2026 and then stopped widening. Operating-margin level is still rising. The year-over-year operating-margin delta cooled 30 bps from Q1 to Q2. A Q3 sales print at the high end holds +17%. A print at the midpoint is another 120 bp slowdown, which is the reading Schlesinger rejected.

The drivers, year over year only. Enterprise is accelerating into a base that was already +58%. Carrier broke. Solar sales accelerated and solar profit did the opposite.

Optical was 44% of Q2 core sales ($2,072 / $4,738) and 39% of Q3 2025. Enterprise went from 50% of Optical a year ago to 61% last quarter. Optical profit grew faster than Optical sales (+77% versus +32%). Net margin in Optical was about 21% ($438 / $2,072), versus about 18% in Q3 2025 ($295 / $1,652). Q1 was already about 21% ($387 million on $1,846 million). The sequential leverage the street wanted in Q2 was not there. Holding 21% while sales rise is what the EPS guide needs.

Repeat Q2’s mix on the Q3 2025 base and the street’s $4,999 million falls out. Another +65% on enterprise of $831 million is about $1.37 billion. Another +1% on carrier of $821 million is about $8 million. Optical in that case is about +33%, in line with Q2. Non-optical then needs about +7%, which is what Q2 non-optical did. A carrier rebound is not in the number. A return of enterprise to the Q1 rate of +36% ($962 million versus $705 million) leaves a hole solar and carrier did not fill in Q2.

Holding the Q2 glass dollar of $1,463 million against the Q3 2025 comp of $1,560 million is −6.2%. Even a repeat of Q2 solar sales is +37% versus $319 million. A small recovery from the −$7 million loss can still be a profit decline versus the $30 million earned in Q3 2025. Q1 solar net income was a $7 million profit before the shutdown cost.

Bridge items on the July 28 slides, still unreopened: operating expense consistent with Q2, other income/expense −$100 million to −$90 million, non-controlling interest −$70 million to −$60 million, weighted-average shares about 875 million, annual core tax rate 18.5%, down from 19.0%.

FQ+1 — Q4 2026

No Q4 sales range and no Q4 EPS range were given. Schlesinger did not answer a “why does Q4 look limited” question with a dollar. He said growth continues after a possible early hit on the $20 billion run-rate, from Enterprise and Photonics, and that Solar should grow from current levels. Capex stays elevated. Consensus percent difference is versus Q4 2025, because there is no guide to diff against.

Street Q4 sales growth of +16.2% is a touch slower than the street’s Q3 +17.0% and in line with Q2’s +17.1%. The $5,125 million consensus annualizes to $20.5 billion, which is already through the end-2026 internal run-rate of $20 billion. That run-rate is a company plan, not a Q4 guide. Do not subtract FY revenue of $19.230 billion from $20 billion and call it a miss. One is a full-year total that includes a slower first half already booked. The other is a Q4 exit rate.

Full-year framework still in force

FY2026 has no revenue range and no EPS range. FY2027 is not guided. The $20 / $30 / $40 billion figures are exit run-rates, repeated on July 28, not fiscal-year revenue. The high-confidence plan, which haircuts scale-up timing, is $27 billion by the end of 2028 and $35 billion by the end of 2030. FY2028 FMP revenue of $27.692 billion sits next to the haircut exit, not next to the internal $30 billion. A full year is not an exit rate.

Five years. Sales growth accelerated from +6.5% to +13.4%. EPS growth accelerated faster, from +15.3% to +28.6%. Operating-margin expansion widened. Gross-margin expansion did the opposite: +190 bps in 2024 shrank to +20 bps in 2025. The quarterly gross-margin story has since turned back up. The annual gross-margin line is no longer the quarterly one.

The filed guide is unchanged. Confidence on the July 28 call was the highest of the last four, and it was the first time Schlesinger had to swat a deceleration narrative. January and April were open-throttle: growth accelerating, the plan being raised, the 20% margin in sight. July kept the long-range plan, raised capex, and wrapped Q3 in a range whose low end would have been a slowdown — then spent the Q&A saying that was not the message. September, in the Citi recaps, closes that argument on sales. It does not close it on EPS, on carrier’s quarterly rate, or on when solar profit looks like the rest of Corning.

The official series have not moved. Daloopa’s document list for calendar Q3 2026 is empty. The September 9 remarks are accounts of the conference (Investing.com, MarketBeat), not a new 8-K and not a new fundamental id. Those accounts put Q3 sales at the high end of the July range or slightly better, tie a $20 billion annualized rate in Q3 to that outcome, and have Q4 larger than Q3. Slightly better than $5.0 billion would be the first time this cycle they described a quarter as through the range before it printed. It is still not a filed number. He did not, in those accounts, walk the EPS range up. He repeated “above 20%,” with no new margin target.

Sales have printed above the guided point, or above the high end of a range, on every quarter in the window below. Core EPS has printed at or above the high end for the same quarters. The pedigree is “at or through the top,” not “the midpoint.” That is why the September sales comment matters. It lines the quarter up with the pattern. It does not replace the range.

Tone, same two speakers. The caveat rotated from “solar ramp is a cost” to “solar profit improves from here,” and a new one appeared in consumer hardware.

The assumption stack inside the Q3 number, in the order that can move the print:

1. Enterprise stays the growth engine. If that order book pauses, the high end is how they miss a “similar to Q2” outcome without leaving the range.
2. Solar stops being a sequential cost. The Q2 extra roughly $30 million versus Q1 is not repeated, and profit improves off −$7 million. Missing it pressures EPS more than sales.
3. Handheld is a headwind they believe they can outgrow. It is the one end market flagged as below normal seasonality. Apple’s July 30 outlook, two days after Corning guided, said September-quarter supply constraints would increase for iPhone, Mac, and iPad, while Apple still guided its own revenue up. That is consistent with “weaker than the first half, but we outperform.” Corning has not commented on it.
4. The $20 billion run-rate “a quarter early” only happens at the top of the sales range.
5. Margins are not guided. EPS growing faster than sales is the claim. There is no quarterly margin range to beat or miss.
6. Scale-up does not have to contribute in Q3. The gap between the $27 billion high-confidence plan and the $30 billion internal plan is a timing reserve. Weeks said outside arguments about when scale-up starts have not changed the May plan.
7. The bridge holds: opex flat with Q2, other expense about $90–$100 million, NCI about $60–$70 million, shares about 875 million, tax about 18.5%. None of these was restated at Citi. The $2 billion ATM is unused capacity as of the filing. If shares are sold before quarter-end it can nudge the share guide. No sale has been confirmed.

The number that confirms the tone is core sales at or through $5.0 billion with enterprise still carrying the growth. The number that breaks it is a sales print toward $4.9 billion after they told Citi the quarter was already at the high end.

Optical pure-plays do not set up this tape. Ciena (September 3) is the only direct optical print already in, and it beat on the FMP row (EPS 2.11 versus 1.73; revenue $1.671 billion versus $1.637 billion). Still ahead of October 27, on FMP dates that are not Daloopa facts: Micron on September 30 (memory, the handheld headwind), Ericsson on October 15 (carrier capex), TSMC on October 15 (AI silicon, not a fiber comp), Nokia on October 22 (optical networks), and Celestica on October 26 (AI networking assembly; the date can slip). Amphenol is October 28. Lumentum and Coherent are early November. Corning will set the optical tone rather than receive it.

Consensus is a repeat of Q2’s growth rate, taken at the top of the sales guide and in the upper half of the EPS guide. It is not a Springboard upgrade, not a photonics quarter, and not a Verizon quarter. FMP moved from $0.85 and $4.983 billion on August 1 to $0.878 and $4.999 billion on September 24. That is the tape marking “land on the high end.” It did not break the July framework. There is no street optical, enterprise, carrier, solar, or glass estimate. What consensus believes about each line is inferred from where company-level FMP sits versus the guide and versus the Q2 mix.

Confirms the print: enterprise still well above +36%, optical sales still up about 30% versus $1,652 million, optical net margin holding about 21%, solar net income no longer a loss, glass within a couple of points of flat, and core EPS inside $0.85–$0.89 without a lower share count doing the work. Q3 2025 core net income was $585 million. Q2 core net income of $680 million was +30% versus $523 million, the same pace as EPS, so the algorithm is earnings, not the share count.

Breaks the print: enterprise cooling back toward the mid-30s with carrier still at +1%. Or enterprise holds and glass absorbs the memory downturn harder than content per device can cover, which pulls $4,999 million back toward the middle of a range the street has stopped using. Or solar profit does not improve, and EPS lands on the $0.85 low end the August file was still using.

The comments that matter beyond the print are whether the Q4 guide finally moves the $20 billion plan, whether carrier is described as timing or as a slower book, and whether scale-up language changes after a quarter whose only new evidence was a lab demonstration and a stack of 2027–2032 fiber contracts.

Since July 28 the company has kept adding evidence that optical demand is being contracted forward. None of that evidence rewrites the quarter being reported. Enterprise was already the fast piece. The new agreements are the slow piece — carrier and long-haul — and they start contributing in 2027. Against that, Corning opened a $2 billion equity tap it does not have to use. No guidance cut, no CEO or CFO change, and no new Daloopa series.

Sell-side color in the window focused on optical spending and funding, not a filed Q3 demand change. Plaintiff-firm “investigation” notices after the July release were not matched to a filed complaint and are not treated as a legal development.

The October 27 debate is whether 2027 carrier fiber and a protected solar chain are enough to keep the Springboard slope, and whether that slope still funds itself without the equity facility. It is not whether Verizon miles land in this quarter. They do not.

Consistent beater on core EPS and core sales, with the surprise compressed to cents until Q2 2026 went back through the top of the guide. Corning beat FMP core-EPS consensus in 10 of 12 quarters and in 4 of 4 since Q3 2025. That is a 10-quarter streak, Q1 2024 through Q2 2026. Core sales beat in 11 of 12 and in 4 of 4 recently, an 11-quarter streak from Q4 2023. The only misses are Q3 2023 (both lines: one cent, and 1.2% of sales) and Q4 2023 (EPS, one cent). Q3 2026 has not printed.

Scored actuals are core sales and core diluted EPS. On every quarter in the window, FMP epsActual equals core diluted EPS and does not equal GAAP diluted EPS. FMP revenueActual is not one series: it equals GAAP net sales from Q3 2023 through Q3 2025, then switches to core sales from Q4 2025 through Q2 2026. Q3 2025 FMP revenue actual is the GAAP $4,100 million, not core $4,272 million. The FMP revenue estimate tracks core sales in every quarter, including the quarters where the stored actual is still GAAP. Scoring GAAP, or scoring the stored FMP revenue actual, against that estimate would manufacture a sales-miss streak the company and the street are not discussing. Q2 GAAP EPS of $0.64 versus $0.755 is about a 15% miss. That is the wrong line.

The regime improved, then narrowed, then widened once. The first four quarters averaged about +1.2% on EPS. The middle four averaged +4.0% on EPS and +2.3% on sales, led by Q1 2025 (+6.5%) and Q2 2025 (+5.1% EPS, +4.9% sales). The last-four EPS average of +1.8% is less than half the middle four. Three of those four dollar surprises are a penny or less (+$0.005, +$0.013, +$0.008). Q2 2026 is +$0.025, because the print went through the range: core EPS $0.78 versus a $0.755 estimate, and core sales $4,738 million versus $4,630 million. It is not a return to the +5% to +8% surprises. Those showed up when FMP was still below the guide. It is not below the guide now.

The cap is the guide, not a weaker company. Core EPS growth is still accelerating at the recent prints. From Q1 2024 through Q2 2026, core EPS has been at or above the company’s own high end in 10 of 10 quarters: four pennies through the high, six prints exactly on it. The FMP “beat” is what is left after the street climbs toward that high end. Versus the company’s own sales anchor, the beat did not fade: Q2 2026 was +3.0% above the approximate $4.6 billion point, wider than the prior three quarters.

Green is a beat, red is a miss, yellow is a beat that has compressed under about 1.5%. Solid fill is a surprise of about 3% or more, or any miss. Smaller surprises are the brand color on white. Q3 2026 is not on the grid.

Read left to right. Both rows are red only in Q3 2023, and the EPS row is red once more in Q4 2023. From Q1 2024 both rows stay green. The dark cells are the quarters when the street was still below the guide. From Q3 2025 through Q1 2026 the cells fade to about +1%. Q2 2026 darkens again because the print cleared a range the estimate had not fully climbed.

Into October 27 the base rate is another core beat on both lines, small on EPS, and a sales beat only if they clear $5.0 billion. Street sales at $4,999 million are the ceiling. Printing $5.0 billion is in line, not a beat. Printing the EPS high of $0.89, which is what core EPS has done for ten straight quarters, is about +1.4% versus $0.878. Printing the midpoint, $0.87, would be a small miss versus $0.878. They have not printed below their own EPS high since Q4 2023. It is not a base rate for another +5% to +8% surprise.

Q2 was a beat on both core lines and cleared the guide. The Q3 outlook initially looked soft against that result, but by September 24 FMP estimates had moved close to the July sales high. The next print tests whether that estimate move was justified.

Data sourced from Daloopa.