The Conservative Guide Was Sandbagging After All: Upgrading to Outperform
Key Takeaways
- KLA beat again (revenue $3.415B, non-GAAP EPS $9.40, 62.2% gross margin, all above guidance and Street), but the story was the guide: management flipped calendar-2026 revenue growth from the "mid-single digits" that broke the stock in January to high-teens, with the semiconductor process-control systems business now guided to grow over 20%.
- Our January thesis that the December-quarter guide was conservatism, not deceleration, played out almost exactly: management raised its 2026 WFE view to exceed $140B and, unusually for an April call, volunteered that 2027 growth should be even higher, citing "unprecedented demand visibility."
- Advanced packaging inflected hard: KLA claimed the #1 process-control position in advanced wafer-level packaging for 2025 (+14 points of share, ~70% revenue growth) and lifted its 2026 process-control packaging revenue target to ~$1B from ~$635M in 2025, well above the framing of just 90 days earlier.
- The blemishes were minor: operating expense ran hot ($670M vs. ~$645M guide) on timing and reserves, the EPS beat leaned partly on an OI&E mark-to-market gain and a higher tax rate, and the DRAM-cost gross-margin headwind firmed to ~100 bps; the stock faded 3.6% on profit-taking after a ~50% year-to-date run.
- Rating: Upgrading to Outperform from Hold. The single biggest overhang from January (a below-market growth guide) has reversed into a tailwind, share gains are now confirmed by third-party data, and the March 12 Investor Day raised the long-term model. Valuation is the residual risk, but a re-accelerating, share-gaining franchise with rising capital returns should beat the market over the next twelve months.
Results vs. Consensus
The March quarter was the fourth consecutive beat, and a higher-quality top line than the headline EPS beat implies (more on that below). Every guided line came in above the midpoint.
| Metric | Actual | Consensus / Guide Mid | Beat/Miss | Magnitude |
|---|---|---|---|---|
| Revenue | $3.415B | ~$3.37B / $3.35B | Beat | +1.2% vs. Street; +1.9% vs. guide mid |
| Non-GAAP Gross Margin | 62.2% | 61.75% (guide mid) | Beat | +45 bps |
| Non-GAAP Operating Margin | 42.6% | n/a | In line | ~43% model |
| Non-GAAP EPS | $9.40 | ~$9.17 / $9.08 | Beat | +2.5% vs. Street; +$0.32 vs. guide mid |
| GAAP EPS | $9.12 | n/a | Beat | above guide mid |
| Operating Expenses | $670M | ~$645M (guide) | Above | +$25M (timing/reserves) |
| Free Cash Flow | $622M | n/a | In line | 31% FCF margin |
Year-over-year
| Metric | Q3 FY26 (Mar-26) | Q3 FY25 (Mar-25) | YoY |
|---|---|---|---|
| Total revenue | $3.415B | ~$3.08B | +11% |
| Service revenue | $775M | ~$668M | +16% |
| Non-GAAP EPS | $9.40 | ~$8.16 | ~+15% |
Sequential (vs. December 2025 quarter)
| Metric | Q3 FY26 (Mar) | Q2 FY26 (Dec) | QoQ |
|---|---|---|---|
| Revenue | $3.415B | $3.30B | +3.5% |
| Non-GAAP EPS | $9.40 | $8.85 | +6.2% |
| GAAP EPS | $9.12 | $8.68 | +5.1% |
| Non-GAAP gross margin | 62.2% | 62.6% | -40 bps |
| Operating expenses | $670M | $653M | +$17M |
| Free cash flow | $622M | $1.26B | -51% (seasonal) |
| Capital returned | $875M | $797M | +$78M |
Assessment by line
Revenue. The $3.415B print grew 11% year-over-year and 3.5% sequentially, driven by leading-edge foundry/logic and high-bandwidth memory. Importantly, it sits at the front of a year management now expects to accelerate sequentially every quarter. The June guide of $3.575B (+4.7% at the midpoint) confirms the ramp is underway.
Margins. Gross margin of 62.2% beat the 61.75% guide on better service mix and manufacturing scale, but is down 40 bps sequentially and management held the June guide at 61.75%, reiterating that the DRAM component-cost headwind is now firmly ~100 bps. The margin story remains range-bound near 62% for the year, the one part of the model still on the back foot.
EPS. The $9.40 result is up ~15% year-over-year, the strongest of the trailing four quarters. Adjusting for the OI&E gain and tax, the underlying earnings power is a touch lower but the trajectory (mid-single-digit sequential growth compounding through the year) is what matters, and it is intact.
Segment Performance
| Segment / metric | Revenue | Growth | Read |
|---|---|---|---|
| Total revenue | $3.415B | +11% YoY, +3.5% QoQ | Above guide midpoint |
| Service (across segments) | $775M | +16% YoY | -1% QoQ on rev-rec timing |
| Advanced packaging (process-control products) | ~$1B (2026E) | high-50s% growth | Up from ~$635M in 2025 |
| June-quarter semi PC mix: Foundry/Logic | ~82% | up from ~60% | Leading-edge ramp |
| June-quarter semi PC mix: Memory | ~18% | DRAM ~84% / NAND ~16% | Timing of memory pull |
Semiconductor Process Control
The core engine is guided to grow over 20% in calendar 2026, faster than the total company (high-teens) and well ahead of a WFE market management now sizes above $140B. The June-quarter mix swings sharply toward foundry/logic (~82% of semi process-control systems revenue), reflecting the leading-edge ramp; the memory-heavy mix will return as DRAM/HBM capacity comes online later in the year and into 2027.
"We expect our semiconductor process-control systems business to grow over 20% [in 2026]... we begin the year well positioned to continue to increase our share of the overall market in 2026 and 2027."
— Bren Higgins, CFO
Assessment: This is the number that resolves the January debate. A >20% core-systems growth rate against a mid-to-high-teens market is unambiguous share gain, exactly what management could not spell out precisely three months ago and can now.
Advanced Packaging
The standout. KLA claimed the #1 process-control position in advanced wafer-level packaging for 2025, with share up 14 percentage points and revenue up ~70%, and raised its 2026 process-control packaging revenue target to ~$1B from ~$635M in 2025, a high-50s% growth rate and a sharp upward revision from the >30% framed a quarter earlier.
"We now expect semiconductor process-control product portfolio revenue for advanced packaging will grow from approximately $635 million in 2025 to approximately $1 billion in 2026, well above our prior estimates."
— Rick Wallace, CEO
Assessment: Advanced packaging has gone from optionality to a bona fide growth driver in two quarters. As inspection value migrates into the package (hybrid bonding, SOIC), KLA's front-end toolset is winning share, and the shorter-lead-time nature of packaging means this can flex up faster than core WFE.
Service
Service was $775M, up 16% year-over-year but down 1% sequentially on revenue-recognition timing. Management reaffirmed the 13–15% long-term services CAGR set at the Investor Day and expects the wave of 2026 shipments to feed service growth toward the high end of the range in 2027 and beyond.
Assessment: The annuity keeps compounding. The sequential dip is optical (timing); the trajectory and its role anchoring capital returns are unchanged.
Key KPIs
| KPI | This Q (Mar) | Last Q (Dec) | Read |
|---|---|---|---|
| Non-GAAP gross margin | 62.2% | 62.6% | Range-bound near 62% |
| Non-GAAP operating margin | 42.6% | 42.8% | Best-in-class |
| R&D | $389M | $384M | Investing through the ramp |
| SG&A | $281M | $270M | Above guide (reserves) |
| Free cash flow | $622M | $1.26B | ~$4.0B LTM, 31% margin |
| Capital returned | $875M | $797M | $3.2B LTM |
| Diluted shares | ~131.7M | 132.0M | Buyback grinding lower |
Key Topics & Management Commentary
Overall Management Tone: Markedly more confident than the defensive posture of the January call. Management moved from justifying a cautious number to describing "unprecedented" visibility, volunteering 2027 commentary it would normally withhold in April, and leaning into the share-gain data that third parties had since confirmed. The one area of restraint was margin, where the tone stayed careful and the DRAM headwind was, if anything, nudged worse.
1. The Guide Reversal: Mid-Single Digits Becomes High-Teens
The defining development. Three months after guiding calendar-2026 revenue growth to mid-single digits, management now expects high-teens company growth and >20% growth in the core semiconductor process-control systems business, with sequential growth every quarter and a second half up 15–20% versus the first.
"For 2026, we expect sequential revenue growth for the company to accelerate, leading to high-teen revenue growth year-over-year, and the semiconductor process-control systems business to grow over 20%."
— Bren Higgins, CFO
Assessment: This is a rare, clean vindication of a "sandbagging" read. The demand that was supply-gated in the first half is converting, and the company that would not guide the year in January is now comfortable putting a high-teens number on it. The magnitude of the swing (from mid-single to high-teens) is the whole reason for our upgrade.
2. Unprecedented 2027 Visibility
Management broke its own convention by discussing 2027 in April, saying the 2027 WFE growth rate should exceed 2026's, driven by a wave of greenfield fab projects across logic, memory, and flash.
"While normally we would not comment on 2027 growth rates in April of 2026, this demand environment gives us confidence... we expect the 2027 year-over-year growth rate to be higher than our growth-rate expectations for 2026."
— Bren Higgins, CFO
Assessment: Volunteering 2027 upside a year early is a strong signal from a management team that has earned a reputation for conservatism. It extends the runway well beyond the current print and materially de-risks the multi-year model.
3. The March 12 Investor Day: A Bigger 2030 Model
Between the two prints, KLA held its Investor Day and raised the long-term framework: a 13–17% revenue CAGR through 2030 (revised up), a 13–15% services CAGR, a baseline semiconductor-industry CAGR of 11% with WFE growing ~1% faster to ~$215B (±$20B) by 2030, and a capital-allocation target lifted to over 90% of free cash flow. It also announced the 17th consecutive dividend increase (to $2.30 per quarter, beginning May) and an incremental $7B buyback authorization.
"We introduced new long-term revenue growth targets along with a 2030 financial model and increased our capital allocation to target over 90% of free cash flow."
— Rick Wallace, CEO
Assessment: The Investor Day did exactly what we hoped in January: it reframed the story around durable multi-year intensity and share, and it stepped up the capital-return commitment. The stock's ~27% recovery between the two prints reflects the market re-embracing that framing.
4. Share Gains Now Confirmed by Third-Party Data
Management cited independent industry research showing KLA increased its share of both the overall wafer-equipment and process-control markets in 2025, with process-control share up ~360 bps since 2021 and roughly 7x the nearest competitor.
"Since 2021, KLA's share of process control has grown by 360 basis points and is approximately 7x greater than the nearest competitor."
— Rick Wallace, CEO
Assessment: This is the external validation the January bears wanted. The "where are the share gains" question is answered not by management assertion but by third-party data, which is far harder to argue with.
5. Gross Margin: The DRAM Headwind Firms, OpEx Runs Hot
The lone soft spot. The DRAM component-cost headwind, sized at 75–100 bps in January, is now put firmly at ~100 bps, with management flagging that DDR4-versus-DDR5 pricing has worsened at the margin. Tariffs remain a 50–100 bps drag, trending toward the low end. Operating expense of $670M ran ~$25M above guide on materials timing and reserves.
"I thought the headwind was 75 to 100 basis points. I think it's 100 basis points now... you still have tariff dynamics. Overall, we said 62% plus or minus 50 basis points, and we still feel that's appropriate."
— Rick Wallace, CEO
Assessment: The one place the story did not improve. Margin remains capped near 62% for the year, and OpEx creep bears watching. We view it as contained and self-correcting (memory supply normalizes, tariff mitigation continues), but it is why the upgrade rests on growth and capital returns, not margin expansion.
6. Advanced Packaging Inflection (What Changed in 90 Days)
Asked directly what drove the upward revision, management pointed to short-lead-time packaging demand that was not visible at the start of the year and to competitive share wins in more advanced systems.
"A quarter ago we thought the overall growth rate in process-control advanced packaging was somewhere in excess of 30%. If you do the math on the numbers now, we're in the upper 50% range."
— Rick Wallace, CEO
Assessment: Packaging is the fastest-moving positive in the model and the clearest example of demand surprising to the upside in real time. It supports the broader "supply-gated, not demand-gated" read from January.
7. AI CapEx Durability
With a couple of hyperscalers printing lighter quarterly capex the same evening, an analyst asked management to bridge its >$140B WFE view to underlying AI infrastructure spend. Management's answer was that the industry is structurally under-supplying AI demand, so quarterly hyperscaler capex noise does not map cleanly to WFE.
"There's not enough silicon to be able to support the plans that people have. So WFE is literally just as fast as we can go as an industry... the reason '26 isn't bigger is because they can't build enough fast enough."
— Rick Wallace, CEO
Assessment: A confident, supply-side framing of the AI debate. It will not satisfy investors worried about a capex air pocket, but it is consistent with the extended lead times and greenfield-fab visibility management describes, and it reframes the bear case from "demand" to "digestion timing."
8. Foundry/Logic Mix Surges Toward ~82%
The June-quarter semi process-control mix swings to ~82% foundry/logic and ~18% memory, a sharp move from the ~60/40 split in December, reflecting the timing of leading-edge shipments. Management framed the memory piece as returning later in the year and into 2027 as DRAM/HBM greenfield capacity ramps.
Assessment: The near-term mix is logic-heavy, which is favorable for intensity and margin (logic carries the highest process-control intensity). The memory wave is deferred, not diminished, consistent with the 2027 acceleration commentary.
9. Supply and Capacity: Positioning for 2027
Management acknowledged the speed and slope of the first-half ramp caught it somewhat off guard on supply, but said it is now much better positioned to support the 2027 buildout, and is hiring installation and service resources in parallel with tool capacity.
Assessment: The candid admission that demand outran supply in the first half is, counterintuitively, bullish: it confirms the January "supply-gated" framing and shows management scaling ahead of an even bigger 2027.
10. Memory Intensity and the Metrology Technology Debate
On an analyst question about x-ray versus optical metrology, management gave a nuanced answer: KLA maintains capability across modalities (its X-ray/Axion tool holds ~60% of a small ~$75–100M market), but volume production gravitates to the highest-throughput, lowest-cost-of-ownership tool that can do the job. New technologies win in development and characterization first, then must clear a cost/throughput bar for high-volume manufacturing.
Assessment: This is classic KLA portfolio discipline, and a reminder that its moat is as much about cost-of-ownership and breadth as any single modality. It also signals management is not banking on speculative technology shifts to hit its numbers.
11. Capital Returns Step-Up
Capital returned in the quarter was $875M ($626M buyback, $249M dividend), $3.2B over the trailing twelve months, and the Investor Day lifted the payout target to over 90% of free cash flow, with the dividend rising to $2.30 per quarter (17th consecutive annual increase) and a fresh $7B buyback authorization.
Assessment: The capital-return commitment is now more aggressive and more explicit. Against a re-accelerating revenue base, a >90%-of-FCF payout is a meaningful support for per-share value and a differentiator versus more cyclical peers.
Guidance & Outlook
| Metric | Q3 FY26 Actual (Mar) | Q4 FY26 Guide (Jun) | Change |
|---|---|---|---|
| Revenue | $3.415B | $3.575B ± $200M | Up ~4.7% at mid |
| Non-GAAP EPS | $9.40 | $9.87 ± $1.00 | Up ~5% at mid |
| GAAP EPS | $9.12 | $9.66 ± $1.00 | Up |
| Non-GAAP gross margin | 62.2% | 61.75% ± 1.0 pt | Down ~45 bps (mix + DRAM) |
| OpEx | $670M | ~$665M | Roughly flat |
| Foundry/logic mix (semi PC) | — | ~82% | Up from ~60% |
The June guide continues the sequential ramp management promised, with revenue up ~4.7% at the midpoint and EPS pushing toward $9.87. The full-year framing is the headline: high-teens company revenue growth, >20% for the core process-control systems business, WFE above $140B, and gross margin held at ~62% (±50 bps) with the ~100 bps DRAM drag baked in. Management explicitly expects sequential growth in every quarter of 2026 and a second half up 15–20% versus the first, consistent with a full-year revenue framework around $15B.
Implied ramp: With the June quarter guided near $3.575B and the second half stronger still, the company is tracking to a mid-teens-plus full-year growth rate, comfortably above management's own $140B+ WFE market growth, which is share gain in plain sight.
Street at: Consensus had modeled the March quarter near $3.37B and has been chasing estimates higher since the March Investor Day; the guide raise pulls numbers up again across 2026 and now 2027.
Guidance style: Still conservative by DNA, but the company is now guiding up rather than managing expectations down. The January-to-April reversal is a reminder that KLA's initial guides tend to mark the floor, not the ceiling.
Analyst Q&A Highlights
Lead-time visibility into 2027 and the CY2026 revenue framework
The opening exchange probed how far the extended lead times and backlog now see, and whether ~$15B is the right calendar-2026 revenue frame.
Q: "As you think about the sequential going to the high teens in the second half, should we be thinking about $15 billion as the right framework for calendar 2026 revenues?"
— C.J. Muse, Cantor Fitzgerald
A: "Yes, I think so. If you just take the commentary, getting to the high teens gets you into the 15-ish range... 15% to 20% type second-half growth over the first half. It puts you up into that ballpark. You're thinking about it the right way."
— Bren Higgins, CFO
Assessment: Management endorsing a ~$15B frame is a notable step up in specificity from the January refusal to guide the year. That willingness to anchor a number is itself a confidence signal.
Why isn't the 2030 model even higher?
A recurring line of questioning noted that KLA looks likely to hit its 2030 targets years early and pressed on the conservatism embedded in the long-term model.
Q: "It's looking increasingly likely we might get to those 2030 levels this year or next year. Why isn't it higher, given where we're sitting right now?"
— Stacy Rasgon, Bernstein Research
A: "The number you're referring to is the semi revenue number, not the equipment number... If we had to redo that today, there are a lot of reasons why you'd push that up. That was six weeks ago, so things have changed. But the numbers around equipment haven't moved nearly as fast as the numbers around semi revenue associated with pricing."
— Bren Higgins, CFO
Assessment: Management all but conceded the six-week-old 2030 model is already stale to the upside. That is a strikingly constructive admission and reinforces our view that the estimate revisions run higher from here.
Outgrowing WFE and the right baseline
An analyst pushed on the gap between KLA's high-teens growth and a broader ~mid-20s WFE consensus, testing whether the baseline is the source of the difference.
Q: "The general consensus among other companies is that WFE is growing mid-20s. Is it that you think that WFE growth is too high, maybe your baseline for last year is more like $120B?"
— Timothy Arcuri, UBS
A: "You're right. The baseline is about $120B, and that aligns with the third parties... growing to about $140B plus, which translates into mid- to high-teen growth. The semi PC business we expect to grow in excess of 20%. If you're going to measure yourself on share of market, you've got to get the numerator right, but you also get the denominator right."
— Bren Higgins, CFO
Assessment: The definitional dispute from January is now resolved in KLA's favor, with the baseline reconciled and the >20% core-systems number doing the work. The "denominator" point is a pointed rebuttal to peers using a bigger WFE number to flatter their own share.
The advanced-packaging upward revision
A recurring topic was the sharp lift in the advanced-packaging outlook and how it squares with peers now citing 50%+ growth.
Q: "You gave that advanced-packaging revenue outlook from $635 million to $1 billion. One quarter ago you were calling much lower growth. What was the big upward revision about? What's changed in 60 to 90 days?"
— Multiple analysts incl. Charles Shi (Needham), Shane Brett (Morgan Stanley)
A: "A quarter ago we thought the growth rate was in excess of 30%. Now we're in the upper 50% range. Packaging is shorter-lead-time business, and there's clearly been momentum from a number of customers for additional capacity this year that we didn't have visibility to going into the year."
— Rick Wallace, CEO
Assessment: A near-doubling of the packaging growth rate in one quarter is the clearest evidence that first-half caution was about visibility, not demand. This is the segment we would watch most closely for continued upside.
AI CapEx durability
With hyperscaler capex prints mixed that evening, an analyst asked management to reconcile its raised WFE view with the underlying AI infrastructure spend.
Q: "Given sporadic market concerns around data-center capex durability, can you bridge your greater-than-$140B WFE view and 2027 growth to the underlying AI infrastructure assumptions? What level of hyperscaler capex does your WFE forecast effectively require?"
— Edward Yang, Oppenheimer
A: "You can't draw a straight line to WFE, because we're way under-serving those demands. There's not enough silicon to support the plans people have. So WFE is literally just as fast as we can go as an industry."
— Rick Wallace, CEO
Assessment: The strongest articulation of KLA's structural-shortage thesis. It reframes the AI-capex bear case as a timing/digestion question rather than a demand question, which is the right frame given the lead times, though it does leave the company exposed if hyperscaler budgets genuinely reset lower.
Is the demand real, or double-ordering?
An analyst asked how much of the unprecedented visibility is genuine end-demand versus customers reserving slots against feared shortages.
Q: "How much of that visibility is actually true demand versus customers making sure there's enough capacity and service personnel?"
— Krish Sankar, TD Cowen
A: "These are significant investments; they're going to open these fabs. Part of the discussions are not only around tools and delivery timing, but also our support, installation, and applications resources... it's really across the company that we're in position to support what they expect to be a pretty significant ramp."
— Rick Wallace, CEO
Assessment: Management tied the visibility to physical fab construction and staffing, not just paper orders, which is the more durable form of demand. We give the answer reasonable credence given the greenfield-project specificity.
Gross-margin puts and takes for the year
A closing question sought the moving pieces behind the unchanged ~62% full-year margin.
Q: "You're sticking with 62% for the year. Can you talk about the pluses and minuses? You had mix headwinds earlier and some cost increases."
— Chris Caso, Wolfe Research
A: "The memory pricing environment is, on the margin, worse; I thought the headwind was 75 to 100 basis points, I think it's 100 now... The tariff headwind should become less as we move through the year. Overall mix is pretty consistent. So 62% plus or minus 50 basis points is still an appropriate way to think about it."
— Rick Wallace, CEO
Assessment: The candid "on the margin, worse" on DRAM is the one place the story did not improve quarter-on-quarter. It keeps margin the watch item and is the reason our upgrade leans on growth and capital returns rather than margin leverage.
What They're NOT Saying
- A precise 2027 growth number: Management volunteered that 2027 will grow faster than 2026 but declined to quantify it, so the runway is directionally strong but not yet sized.
- The path back above 62% gross margin: The company reiterated ~62% for the year and a longer-term 63%+ goal but would not commit to when the DRAM headwind fully rolls off, leaving margin timing vague.
- OpEx discipline specifics: With OpEx above guide on "timing and reserves," management did not detail how much is one-time versus a higher run-rate, a small but worth-watching ambiguity.
- China quantification: Beyond "grows slower than WFE," there was no hard China number, and the new export-control letter was waved off as immaterial without much detail.
- Backlog magnitude: "Unprecedented visibility" and a "growing backlog" are again qualitative; management still will not put a coverage figure on the order book.
Market Reaction
- Pre-print setup: KLAC closed at $181.62 entering the print, up 49.5% YTD, up 31.4% over the trailing 30 days, and up 164.0% over the trailing twelve months, near the top of its 52-week closing range of $67.63–$193.50. The stock had already recovered all of the January drop (and more), helped by the March 12 Investor Day. The S&P 500 was up 4.2% YTD.
- Reaction: Shares opened the next session down 7.1% at $168.80, traded as low as $164.60 (-9.4%), then recovered most of the drop to close at $175.04, down 3.6% ($6.58), on 18.9M shares versus a 9.2M 30-day average (2.0x). The S&P rose 1.0% that session.
The move tells you the raise was largely priced in: after a ~50% year-to-date run and the Investor Day re-rating, a strong beat-and-raise merited profit-taking rather than a fresh leg up. The intraday recovery from -9.4% to -3.6% is the more telling detail: dip-buyers stepped in once the market digested the high-teens/>20% growth framing and the 2027 commentary. This is the opposite pattern from January, when a beat faded to a 15% loss; here a beat faded only modestly and found support. The stock is expensive, and expensive stocks fade good news, but the fundamental direction of travel has clearly reversed.
Street Perspective
Debate: Was January's caution vindicated, and is the raise sustainable?
Bull view: The mid-single-to-high-teens reversal proves the January guide was sandbagging; with 2027 now flagged higher than 2026 and lead times extended, estimate revisions run up from here.
Bear view: A guide that swings this much in one quarter cuts both ways: if demand can surprise up this fast, it can normalize down just as fast, and the company is now guiding into peak enthusiasm.
Our take: The bull side has the better of it. The raise is corroborated by third-party share data, greenfield-fab specifics, and the packaging inflection, all of which are harder to reverse than a single quarter's order flow. Vindication, not luck.
Debate: Can the multiple hold after +164% over twelve months?
Bull view: A process-control near-monopoly with >20% core growth, a >90%-of-FCF payout, and unusual 2027 visibility deserves a premium; the earnings ramp grows into the multiple.
Bear view: Near ~50x trailing earnings and up 164% in a year, the stock prices in a lot of the good news; any wobble in AI capex or a margin miss would hit a richly valued name hard.
Our take: Valuation is the real risk now, and we do not dismiss it. But with the growth rate re-accelerating and the payout stepping up, we think the earnings trajectory supports market-beating returns over twelve months even from a premium multiple. Risk/reward has shifted from "is growth too slow" to "is the price too high," a better problem to have.
Debate: Is the AI-driven WFE ramp real demand or a coming air pocket?
Bull view: The industry is structurally under-supplying AI silicon; lead times and greenfield fabs give multi-year visibility largely independent of quarterly hyperscaler capex noise.
Bear view: Hyperscaler capex is lumpy and a couple of prints came in light; WFE ultimately depends on that spend, and today's shortage narrative resembles prior cycle tops.
Our take: We side with the structural-shortage view for the next 12–18 months given the lead-time evidence, while acknowledging that a genuine reset in hyperscaler budgets is the tail risk that would break the thesis. It is the single most important variable to monitor.
Model Update Needed
| Item | Prior Assumption (post-Q2) | Suggested Change | Reason |
|---|---|---|---|
| CY2026 revenue growth | Mid-single digits | High-teens (~$15B) | Guide raise; 2H up 15–20% over 1H |
| CY2026 semi PC systems growth | ~market | >20% | Confirmed share gain |
| CY2026 gross margin | ~62% | ~62% (±50 bps) | DRAM headwind firmed to ~100 bps |
| Advanced packaging (PC products) | mid-to-high teens growth | ~$1B, high-50s% growth | Upward revision from ~$635M |
| 2027 revenue growth | not modeled | Above 2026 | Management flagged early |
| Capital-return payout | ~$3B/yr | >90% of FCF; +$7B buyback | Investor Day step-up |
Valuation impact: The estimate revisions are meaningfully higher across 2026 and now 2027, which lifts fair value even before any multiple change. Against the FMP reaction close, we would frame a constructive twelve-month target roughly 12–18% above the post-print level, driven by the high-teens revenue ramp, >20% core-systems growth, operating leverage, and a >90%-of-FCF payout, with the caveat that the premium multiple leaves less margin for error than in January. We would fade the position into any AI-capex-driven de-rating and add on pullbacks that do not impair the multi-year demand thesis.
Thesis Scorecard Post-Earnings
| Thesis Point | Status | Notes |
|---|---|---|
| Bull #1: Process-control share leadership (~7x nearest competitor) | Confirmed | Third-party data: +360 bps share since 2021; #1 in advanced WLP |
| Bull #2: Rising process-control intensity (AI, HBM/DRAM, packaging) | Confirmed | >20% core-systems growth vs. mid-to-high-teens market |
| Bull #3: Best-in-class model + capital returns | Mixed (improving) | Payout raised to >90% FCF, +$7B buyback, 17th dividend hike; but margin still capped ~62% |
| Bull #4: Advanced packaging growth vector | Confirmed | ~$1B 2026E (from ~$635M); +14 pts share, ~70% growth |
| Bear #1: Near-term growth below market | Challenged (defused) | Guide reversed to high-teens / >20% core; largely neutralized |
| Bear #2: Gross-margin headwinds (DRAM, tariffs) | Confirmed | DRAM drag firmed to ~100 bps; June GM held at 61.75% |
| Bear #3: Valuation + China/regulatory risk | Escalating | +164% TTM, ~50x trailing; now the primary risk |
Overall: The thesis strengthened materially. The central concern from January (a below-market growth guide) has flipped into a tailwind confirmed by third-party data, the Investor Day raised the long-term model, and 2027 visibility extends the runway. The residual risks migrated: margin remains capped and, more importantly, valuation is now the dominant risk rather than growth.
Action: Upgrade to Outperform from Hold. The overhang that kept us on the sidelines has cleared, and a re-accelerating, share-gaining franchise with a >90%-of-FCF payout should beat the market over the next twelve months. We size the risk around the premium multiple and the AI-capex debate, and would use any capex-scare pullback to add.