The Backlog Says 2027 and the Tape Says Cycle Top. We Side With the Backlog.

Results vs. Consensus

The June quarter closed fiscal 2026 with the cleanest print of the year. Revenue, gross margin and both EPS lines finished above the midpoint of guidance, and for the first time since the June 2025 quarter, non-GAAP earnings power broke to a new high rather than treading water.

Against the company's own guidance (issued April 29, restated for the split)

Year-over-year (June 2026 quarter vs. June 2025 quarter)

Sequential (vs. March 2026 quarter)

Fiscal year 2026 (year ended June 30, 2026)

Assessment by line

Revenue. The $3.658B print is a record and, more usefully, the moment the supply constraint stopped binding. Recall the shape of this fiscal year: December 2025 revenue of $3.30B, March of $3.415B, then a 7.1% step to $3.658B. Management has said since January that the first half of calendar 2026 was gated by long-lead-time materials rather than demand, and this is the first quarter where the shipment capacity arrived. The September guide of $4.0B, another 9.4% sequential step, is the confirmation.

Margins. Non-GAAP gross margin of 62.4% beat the 61.75% guide and, more importantly, the September guide of 62.5% is 75 bps above the guide the company issued three months ago for this quarter. That is the margin story unsticking after four quarters of being capped near 62%. It is happening despite, not because of, the memory input cost line, which management now sizes at slightly above 100 bps and expects to persist through calendar 2027. The offset is operating leverage, and at 59% incremental operating margin it is working.

EPS. Non-GAAP EPS of $1.05 is up 11.7% year-over-year and 11.7% sequentially. The symmetry is not a coincidence so much as a diagnosis: non-GAAP EPS was $0.94 in the June 2025 quarter, dipped to $0.885 in December 2025, recovered to $0.94 in March 2026, and only now exceeds the level it reached a year ago. Twelve months of flat earnings while revenue grew is the arithmetic cost of the supply constraint plus the DRAM cost drag. The June quarter is where that ends.

Cash and the balance sheet. Fiscal-year free cash flow of $3.767B on $13.579B of revenue is a 27.7% margin, still elite for a capital-equipment business, but the quarterly trend is down and the reason is a deliberate working-capital build. Receivables of $2.889B and inventories of $3.649B are both growing faster than revenue, which is what a company ramping into a 20%-plus second half and a bigger 2027 should look like. It is worth watching rather than worrying about, with one caveat: the payout target of more than 90% of free cash flow was not met this year, and if the working-capital build continues at this pace, either the payout or the balance sheet flexes.

Segment Performance

KLA reports three segments. The June quarter is the first in some time where the two non-core segments did something interesting, and one of them did something genuinely important.

Semiconductor Process Control

The core franchise grew 13.2% year-over-year in the quarter and 11.9% for the fiscal year, and it is where the argument about this stock is actually decided. Management guided the September semiconductor process-control systems mix to roughly 73% foundry/logic and 27% memory, a meaningful swing back toward memory from the roughly 82/18 guided for the June quarter, with DRAM at about 90% of the memory piece. Within calendar 2026 as a whole, foundry/logic settles in the low-to-mid 60s percent of that mix.

The important nuance sits in the growth math. Management now sizes the calendar-2026 wafer-equipment market in the low $150 billions, roughly mid-20s percent growth off approximately $120B in 2025. Its own company revenue growth for the calendar year is tracking to the low 20s. That means KLA is growing at, or slightly below, the market it serves this year, with the semiconductor process-control line running a few points faster than the total company because service dilutes the blend.

"If you look at our semi PC business, I would expect that to grow at least a few points... maybe a little more than that faster than the overall company as it's being diluted by service growth."
— Bren Higgins, CFO

Assessment: This is the one place where the story got harder to tell, not easier. In April the pitch was semiconductor process-control systems growing more than 20% against a mid-to-high-teens market, which is share gain you can see in the numbers. This quarter the denominator moved up faster than the numerator, and the CY2026 share gain is no longer visible in the reported growth rates. Management's explanation, that the first half was shipment-constrained rather than demand-constrained and the deficit is sitting in backlog, is consistent with everything else on the call and with a backlog that grew to $12.5B. But it is an explanation, not evidence. Calendar 2027 is where the share claim has to show up in the growth rate, and we will hold it to that.

PCB and Component Inspection

The quarter's genuine surprise. This segment, the core of what KLA bought with Orbotech in 2019, grew 56.5% year-over-year to $241.1M and 20.7% for the fiscal year. The driver is that high-performance computing packages are pulling value into the substrate and the board, which is exactly the thesis the company argued for the acquisition seven years ago and which has spent most of the intervening period not working.

"KLA entered many of these markets in 2019 as part of the Orbotech acquisition. Our investment thesis for the transaction was centered around the rising value of the chip package and the ability for the KLA operating model to drive product strategy, business execution to take advantage of this evolving trend. With these combined products expected to grow over 25% in calendar 2026, we are encouraged by the future opportunities in these markets."
— Rick Wallace, CEO

Management was candid that the inflection outran its own planning. Asked whether the long-term growth rate for the combined specialty and PCB businesses had changed from the mid-to-high single digits framed at the March Investor Day, the CFO nudged it to the high single digits over a multi-year period while flagging that something structural may be happening.

"I wish I could ship more. It certainly turned on much faster than we thought... high-performance compute is changing this market and we'll have to see how it plays out."
— Bren Higgins, CFO

Assessment: At 6.6% of revenue this does not move the model much yet, and one quarter of 56% growth off a small base is not a trend. What it does provide is a second, independent read on the same underlying phenomenon: the value migrating into the package is showing up in two unrelated product lines at once. That is corroboration of the intensity thesis from an angle the bears are not watching. Treat it as confirmation, not as a new growth engine.

Specialty Semiconductor Process

The smallest segment grew 12.6% year-over-year in the quarter but was flat to slightly down for the fiscal year at $584.1M. It serves specialty devices and is exposed to the same high-performance-compute packaging pull as the PCB business, which is why the quarterly rate turned positive after a soft year.

Assessment: Immaterial to the thesis at 4.4% of revenue. Worth tracking only as a confirming indicator alongside PCB and Component Inspection, since the two are driven by the same substrate and packaging dynamic.

Advanced packaging (a cross-segment revenue pool)

Advanced packaging is not a reported segment; it is a pool of process-control revenue that KLA sizes separately, and it was raised again. Management now expects roughly $1.1B of advanced-packaging process-control systems revenue in calendar 2026, up more than 70% year-over-year, against a target of approximately $1B and high-50s percent growth set just 90 days ago and roughly $635M of actual revenue in 2025. The market itself is moving too: management now puts advanced-packaging market growth in the mid-to-high 30s percent, up from roughly 20% at the start of the year, with KLA growing close to twice that rate.

"We now expect our advanced packaging process control systems revenue to grow to approximately $1.1 billion in calendar 2026, up more than 70% year-over-year, above our prior expectation of high 50% growth and almost 2x faster than the advanced packaging market."
— Rick Wallace, CEO

On what is driving it, management pointed to front-end tools being pulled into packaging applications by customers rather than to a purpose-built product cycle, with hybrid bonding and stacked-die integration creating the need.

"The main drivers for that is the accelerated share and adoption of some of our systems that were designed for front end... a lot of it was leveraging the portfolio that we already had. And it was our customers pulling us into that."
— Rick Wallace, CEO

Assessment: This is now three consecutive quarters of upward revision to the same line, from in excess of 30% in January, to high-50s in April, to more than 70% today. Two things make it more than a good number. First, KLA is growing at roughly twice the market rate, so this is unambiguous share capture in a segment where the market itself is accelerating. Second, the incremental revenue is coming off an existing tool portfolio with modest incremental R&D, which means it carries better-than-average drop-through. Advanced packaging has moved from optionality to the most reliable positive surprise in the model.

Services

Service revenue reached $820.4M, up 16.8% year-over-year, and $3,125.9M for the fiscal year, up 16.5%. Roughly 80% of service revenue is contract-based. Management said total service growth is currently at the bottom of the 13% to 15% long-term range set at the Investor Day and should operate toward the high end of that range next year, given the higher shipment levels running through this year and next. A questioner put semiconductor service growth near 20% for the year; management neither confirmed nor repeated the figure.

"80% of revenue is contract-based, so it provides a nice anchor for the company in terms of visibility moving forward. And certainly, our customers in this environment are running their tools at very high utilization."
— Bren Higgins, CFO

Assessment: The annuity keeps compounding and its composition keeps improving, since high-utilization customers running high-value die will pay for uptime. The one honest observation is that service growth is what dilutes the company's blended growth rate below the semiconductor process-control line, which is the arithmetic behind the share-gain question above. That is a good problem: a lower-growth, higher-visibility, higher-margin annuity is worth having even when it makes the headline growth rate look less impressive than the core franchise.

Key KPIs

Key Topics & Management Commentary

Overall Management Tone: The most forward-leaning call in the four quarters we have covered, and notably more concrete than April's. Management raised numbers it had raised only 90 days earlier, volunteered a backlog figure it has spent years declining to give, and described sizing the supply chain against scenarios 10% to 20% above an already-raised 2027 view. The restraint was concentrated in two places: it would not put a KLA-specific growth number on 2027, and it would not touch the 2030 model it published in March even after conceding the assumptions underneath it have moved.

1. The Fourth Wafer-Equipment Raise of the Year

Management raised its calendar-2026 wafer-equipment market estimate, including advanced packaging, to approximately the low $150 billion range, up from the "$140 billion plus" it set in April and roughly mid-20s percent growth off an approximately $120B 2025 base. On the call this was characterized as the fourth raise of the year, a framing management did not dispute.

"We're raising our expectation for the wafer equipment market, including advanced packaging, to approximately the low $150 billion range in calendar 2026, up from our prior expectation of $140 billion plus and mid-20% growth above the approximate $120 billion level in calendar 2025."
— Bren Higgins, CFO

Assessment: Four upward revisions in seven months is a company that started the year badly wrong about its own market and has been catching up ever since. That cuts both ways. It vindicates the January conservative-guide read we made our upgrade on, and it demonstrates real demand momentum. It also means the forecasting process has almost no predictive value, which is worth remembering when the same process tells you 2027 is fine.

2. Calendar 2027: A Second Consecutive Mid-20s Year

Management framed 2027 by pointing at an external consensus rather than issuing its own number, putting the industry view near $190B of wafer equipment, roughly mid-20s percent growth again, and saying it plans for "significant growth."

"There's a consensus view out there, I would say, that's somewhere in and around the $190 billion range. There is some upside or more bullish views of that. There's not too many bearish views of different levels... more or less, you're in and around $190 billion or so translates into a mid-20s type growth rate, which is similar to the growth rate in 2026."
— Bren Higgins, CFO

The drivers named were a broadening of leading-edge logic investment beyond the single leader, high-bandwidth memory at process-control intensity rivalling advanced logic, greenfield DRAM and flash fabs rather than technology upgrades, and continued packaging inflection.

Assessment: The construction of this answer matters as much as the number. Management deliberately anchored 2027 on third-party consensus rather than its own forecast, which is either appropriate humility from a team that has been wrong four times this year or a way to avoid owning a number in a nervous tape. The bull point is the composition: greenfield fabs and node transitions carry higher process-control intensity than capacity adds to existing lines, and both feature prominently.

3. Backlog Comes Back Out of the Drawer

Asked to talk about visibility directionally, given the company stopped reporting remaining performance obligations, the CFO simply gave the number a week before the annual filing would.

"...would expect that the RPO would be or backlog would be at about $12.5 billion, and we've seen that grow pretty consistently over the last couple of quarters and would expect that given the order funnel that we'll see that continue to grow."
— Bren Higgins, CFO

For scale, $12.5B is roughly 0.9 times fiscal 2026 revenue and just over three quarters of revenue at the September guided run-rate.

Assessment: This is the single most important disclosure on the call and the strongest counter to the cycle-top narrative. In January we listed backlog magnitude as something management would not quantify; in April we listed it again. Volunteering it now, unprompted by a specific request for a figure, is a deliberate choice by a management team that wants the number in the market while the stock is being sold. It is not non-cancellable and it does not make 2027 safe, but it converts "unprecedented visibility" from an adjective into a measurable quantity that can be tracked quarter to quarter. We will track it.

4. Advanced Packaging Raised for a Third Straight Quarter

Covered in the segment section above: the calendar-2026 advanced-packaging process-control revenue target went to approximately $1.1B and more than 70% growth, from high-50s growth in April and more than 30% in January, with the underlying market itself re-rated from roughly 20% growth to mid-to-high 30s.

Assessment: The cleanest evidence on the call that intensity is rising rather than merely capacity being added. Growing at twice the market rate off an existing tool portfolio is share gain with above-average drop-through, and it has now surprised to the upside three quarters running.

5. Gross Margin: Leverage Arrives, the Memory Bill Runs Through 2027

The margin story finally moved. The September guide of 62.5% is 75 bps above the guide issued three months ago for the June quarter, and management pointed at operating leverage on revenue growth as the reason. It also confirmed the memory input-cost headwind is slightly worse than the 100 bps sized in April and persists into next year.

"We've been able to secure supply. As we've seen demand strengthen, we've been having to procure a new memory at higher prices. And so over the last couple of quarters or so, we thought the impact was somewhere around 100 basis points. It's probably a little bit more than that. I think that likely continues as we move forward through 2027."
— Bren Higgins, CFO

Against that, the CFO pointed to the Investor Day framework of 60% to 65% incremental gross margin trending toward a 63.5% target model, and said new product introductions are the mechanism for resetting both cost structure and price.

Assessment: Margin moves from the soft spot of this thesis to a neutral-to-positive. The headwind got marginally worse and lengthened by a year, and the margin outcome improved anyway. That is the definition of the leverage working. A company absorbing a 100-plus basis point structural cost increase while guiding margin up 75 bps is demonstrating pricing and mix power in the only way that counts.

6. The Pricing-Power Gap

The most uncomfortable exchange of the call. Pressed on why KLA cannot raise prices to offset memory inflation when peers are visibly repricing, management gave a mechanical answer about order timing rather than a strategic one.

"It's pretty hard to go back to your customers after you've taken orders and start to change prices on those orders."
— Bren Higgins, CFO

The CEO framed it as an ongoing conversation about value capture rather than a repricing.

"The input prices have gone up. Customers recognize that. And so we have had conversations about different discussions about how KLA will help to capture some of that value, so we can continue to make the R&D investments that the industry needs."
— Rick Wallace, CEO

Assessment: This is the structural cost of a long backlog, and it is a real limitation. A twelve-month order book priced at last year's terms means input inflation lands on KLA before it lands on the customer, and the correction only comes with the next product generation. It is the mirror image of the backlog's virtue: the same visibility that protects revenue in a downturn locks in price in an inflation. Not a thesis-breaker, but it caps how quickly margin can respond, and it is the reason we do not model margin expansion aggressively.

7. Supply, Lead Times, and Planning for 2029

Management put company-wide lead times at roughly twelve months, with certain products at 18 to 24 months, and said current customer conversations are about second-half 2027 deliveries. The more striking disclosure was the planning horizon on the supply side.

"One of the things that we've spent a lot of time on over the last few months is meeting with some of our critical suppliers and planning and doing capacity agreements for our needs, we believe, in the 2029 and beyond time frame... it can take 12 to 24 months to put new capacity in place for those types of components."
— Bren Higgins, CFO

On how much capacity it is building, the CFO described scenario-planning at 10% and 20% above the base 2027 view and said it would carry the cost of that flexibility.

Assessment: Signing multi-year optical-component capacity agreements out to 2029 is a real commitment of capital and credibility, not a talking point. It is also the reason the working-capital line is expanding. This is the clearest expression on the call of a management team that believes its own demand picture, and it is the kind of evidence that is hard to fake because it shows up in the balance sheet.

8. China: The New Bear Input, and the Process-Control Moat

The sector's freshest fear is domestic Chinese progress in lithography, and the question on the call was whether that read-across reaches process control. Management's answer leaned on structure rather than on export controls.

"Process control is tricky to get into because it's a very high mix, low volume market. So unlike litho, frankly, is much higher kind of similar tool, higher volume, process control, there's a lot of nuance. There's also a lot of work with algorithms and development in terms of being able to help customers determine what's valuable. Add to that the 1,600, 1,700 applications engineers that KLA has worldwide, and it's a pretty good competitive moat that we've established over time."
— Rick Wallace, CEO

The CFO added an inversion of the usual complaint about export restrictions.

"We've gained share at a time where a number of our competitors have been able to ship into fabs in China that we haven't been able to ship into. So to Rick's earlier point, where we can compete, we generally win."
— Bren Higgins, CFO

Assessment: The structural argument is sound and the inversion is a genuinely good point: KLA has been gaining share while handicapped in the one market where domestic substitution is furthest along. But the answer addresses today's competitive gap, not the trajectory, and management offered no view on how fast Chinese process-control capability is closing. The read-across from lithography to process control is weaker than the tape is pricing, and this is the risk we can least handicap from public information.

9. The 2030 Model Is Already Stale, and Management Will Not Touch It

The March Investor Day set a 2030 wafer-equipment figure of approximately $215B. With 2027 now framed near $190B, that model implies almost no growth across the following three years. Asked about it directly, management conceded the point and declined to move.

"We also couldn't have imagined some of the pricing increases that have driven up the revenue for semi this year... if you envision that the market is going to be much higher than that, once we normalize for some of this pricing a little bit, then the process control intensity, we feel pretty confident about... then we would exceed the model that we laid out for 2030... But we're still a long way. I mean, we only had our Investor Day in March and here we are in July. So we're in no position to reset anything as it pertains to 2030."
— Rick Wallace, CEO

He then delivered the most quotable line of the call, and arguably the most useful.

"We're really good at creating systems that are valuable to our customers for process control and engaging with them. We are not very good at forecasting."
— Rick Wallace, CEO

Assessment: We read the refusal to reset as discipline rather than evasion, and the self-deprecation as accurate. This is the same team that guided calendar-2026 growth to mid-single digits in January and is now describing low 20s. Investors should extract exactly one thing from the 2030 model: the process-control intensity and share assumptions inside it, which management says it feels confident about. The wafer-equipment level on top is a placeholder, and management has now said so out loud.

10. The Share-Gain Question Gets Harder to Answer

The sharpest challenge of the call came on the arithmetic. If process-control intensity is rising, why is KLA's revenue growing roughly in line with the wafer-equipment market rather than faster? Management attributed it to first-half supply constraints and to service diluting the company blend.

"Certainly, the first half of '26 was slower from a sequential point of view given some of the lead time challenges we had and how that translated into the constraints around shipments... if you look at our semi PC business, I would expect that to grow at least a few points... maybe a little more than that faster than the overall company as it's being diluted by service growth."
— Bren Higgins, CFO

Assessment: The answer is coherent and consistent with a backlog that grew to $12.5B, but it is the weakest link in an otherwise strengthening story. In April, the framing of core systems growing over 20% against a mid-to-high-teens market made share gain arithmetically obvious. This quarter the market grew faster than KLA did, and the share claim now rests on an explanation rather than on the growth rate. If 2027 wafer equipment lands near $190B and KLA does not visibly outgrow it, this bear point stops being about timing.

11. Services Cross $800 Million

Service revenue of $820.4M grew 16.8% year-over-year, with roughly 80% under contract. Management placed total service growth at the bottom of its 13% to 15% long-term target this year and expects to operate toward the high end of that range next year, given the higher shipment levels running through 2026 and into 2027.

Assessment: The most underappreciated line in the model. A contracted, high-margin annuity growing in the mid-teens off a rapidly expanding installed base is what makes KLA's downside different from a pure equipment vendor's, and the 2026 shipment surge mechanically pre-funds 2028 service revenue. It is also, as noted above, the reason blended growth understates core growth.

12. The Cost of the Ramp: Working Capital and the Payout

Free cash flow of $817.1M represented 22.3% of quarterly revenue against 27.7% for the fiscal year. Receivables rose 27.6% year-over-year and inventories 13.6%, both against 15.2% revenue growth, and the CFO framed the build as deliberate.

"KLA has made meaningful investments in our working capital and facilities to support the current growth environment. Given the expected aggregate investment in wafer equipment over the next several years, our expectation is that these investments will continue."
— Bren Higgins, CFO

Meanwhile capital returns of $876.3M exceeded quarterly free cash flow, and the fiscal-year payout of $3.348B was 88.9% of free cash flow, marginally short of the more-than-90% target set at the Investor Day.

Assessment: Building working capital ahead of a 20%-plus second half is the right call and is not a quality-of-earnings problem. But it is the first quarter in our coverage where cash conversion visibly lagged earnings, and the payout target was quietly missed without comment. Neither is alarming in isolation. Both belong on the watch list, because a company funding a multi-year capacity build while committing more than 90% of free cash flow to shareholders has less room than the headline balance sheet suggests.

Guidance & Outlook

The September guide is above consensus on both lines, at roughly $4.0B versus a Street near $3.92B to $3.95B and $1.16 versus $1.13 to $1.14. It also carries two headwinds that make the underlying operating beat larger than it looks: other income and expense swings from a $68.7M GAAP contribution in June to a guided $25M expense, and the planning tax rate steps from a 12.0% effective GAAP rate to 14.5%. Growing EPS 10.5% sequentially while absorbing both is a stronger operating guide than the headline conveys.

Implied ramp. Management said second-half calendar 2026 revenue should be approximately 20% above the first half. The first half was $7.073B (March $3.415B plus June $3.658B), which puts the second half near $8.49B. Against the September guide of $4.0B, that implies a December quarter of roughly $4.49B, another step of about 12% sequentially, and a calendar-2026 total near $15.6B. Management did not guide the December quarter and did not dispute an analyst's characterization of it as "up another 10%."

Street at. Consensus modelled $3.61B and $1.00 for the June quarter and $3.92B to $3.95B and $1.13 to $1.14 for September. The guide pulls September numbers up modestly, and the second-half framing pulls the December quarter and calendar 2026 up more. Calendar-2026 company revenue growth is now tracking to the low 20s percent, against the high-teens framed in April and the mid-single digits framed in January.

Guidance style. Still structurally conservative, and now demonstrably so. Three consecutive quarters of raising the calendar-year growth rate, from mid-single digits to high-teens to low 20s, plus four raises to the wafer-equipment market view, establish that KLA's initial framing marks a floor. The corollary is uncomfortable: a forecasting process that has been wrong by that much in one direction is not a reliable instrument for detecting a turn in the other.

Analyst Q&A Highlights

Gross-margin trajectory into 2027 and 2028

The opening question sought the shape of gross margin as new supply comes online, mix shifts toward systems, and memory-component inflation eventually passes through. The answer was the first clear signal that the margin cap is lifting: leverage is now visible in the model, the September guide sits above the 62% full-year framing, and new product introductions are the mechanism for resetting cost and price together.

Q: "Hoping to kind of hear how you're thinking about the additional supply you're bringing on. Any impact there? And as you think about a greater mix shift to tools over other, coupled with, I imagine, at some point, you'll be able to pass on the inflation related to memory and other components, how do you see kind of the trajectory for your gross margins into '27 and '28?"
— C.J. Muse, Cantor Fitzgerald

A: "We're certainly starting to see some benefit from leverage in the overall model. We're still dealing with some headwinds as related to memory pricing... We talked about 2026 overall gross margins likely being in the 62%, plus or minus. Clearly, if you look at the guidance we provided expectations for next quarter, we're going to be above that. So we feel pretty good from a trajectory point of view."
— Bren Higgins, CFO

Assessment: The most consequential change in the model this quarter, and the least discussed. Four quarters of margin capped near 62% ends here, and it ends while the input-cost headwind is getting slightly worse, which means the improvement is leverage and mix rather than relief.

Squaring the wafer-equipment raise with process-control growth

A line of questioning worked the arithmetic backwards: if the market view has been raised four times this year to the low $150 billions, the implied company growth should be mid-20s and process control closer to 30%, well above the "greater than 20%" framed in April. Management neither confirmed the 30% figure nor rejected the direction, redirecting instead to second-half acceleration and to the supply constraint that shaped the first half.

Q: "The team has raised their WFE outlook, I think, 4 times this year including today to now sort of that low $150 billion range, so up kind of mid- to high 20% range. I would think that this implies your overall business growing sort of mid-20% range, and I assume that this implies that your process control business grows closer to 30% versus your prior view of greater than 20%."
— Harlan Sur, JPMorgan

A: "Given how quickly the industry turned and started to ramp, we did, as we talked about in the past, deal with some supply chain shortages as it relates to some of our longer lead time materials and how that's impacted in the first half. Certainly, in the second half, we're seeing that supply come on, and we're seeing the second half acceleration as we talked about, 20% over the first half."
— Bren Higgins, CFO

Assessment: A question about share answered with a question about timing. The evasion is understandable given the first half genuinely was shipment-constrained, but it leaves the CY2026 share-gain claim resting on backlog rather than on revenue. This is the exchange to re-read a year from now.

Whether backlog corroborates the visibility claim

After several quarters of describing visibility only in adjectives, management was asked whether it could characterize the order book directionally now that it no longer reports remaining performance obligations. It responded with the actual figure, a week ahead of the fiscal-year filing.

Q: "Just talking about visibility, I know you don't give RPO anymore. I'm just kind of curious if there's a way you could talk about it directionally. Is it improving? Or maybe you can talk about your outlook visibility. How far is that extending? Is it changing?"
— Blayne Curtis, Jefferies

A: "...would expect that the RPO would be or backlog would be at about $12.5 billion, and we've seen that grow pretty consistently over the last couple of quarters and would expect that given the order funnel that we'll see that continue to grow."
— Bren Higgins, CFO

Assessment: Volunteering a number this company has avoided for years, in a week when its stock is down more than 35% from its high, is a considered decision. It converts the central claim of the last three calls from rhetoric into something falsifiable. That is a net positive for the thesis and a new metric for our quarterly scorecard.

Why revenue is not outpacing the wafer-equipment market

The most direct challenge to the intensity thesis: revenue as a share of wafer equipment has held roughly flat through the year, which is not what rising process-control intensity should produce. The response separated a structural claim from a timing one, and conceded that the first half was constrained rather than representative.

Q: "When I look at your revenue over the quarter, it seems like your revenue as a percentage of WFE is pretty consistent through the year. I'm just wondering why... if process control intensity is increasing, shouldn't your revenue be outpacing WFE?"
— Krish Sankar, TD Cowen

A: "Certainly, the first half of '26 was slower from a sequential point of view given some of the lead time challenges we had and how that translated into the constraints around shipments... Our story at Investor Day was our view that because of the dynamics of high performance compute... that the next 4 to 5 years would be as strong as the last 4 to 5 years in terms of KLA's overall market relevance."
— Bren Higgins, CFO

Assessment: The right question, asked precisely. Management's answer is defensible and matches the backlog build, but the honest read is that the CY2026 numbers do not demonstrate share gain and the argument has moved to a multi-year horizon. We accept the timing explanation for one year. We would not accept it for two.

Pricing power against memory-component inflation

A pointed follow-up asked why the company cannot offset input inflation with price when peers are visibly repricing, and whether there was a strategic reason for the restraint. The answer was mechanical rather than strategic, and it exposed the cost of a long order book.

Q: "I'm a little confused as to why you can't raise prices to at least offset memory prices. I know that you're coming from a different place than others are. But I mean, even ASML is talking about raising prices on EUV. So I would think you'd be able to at least offset the compute headwind, and it seems like you're struggling to do that. So is there something unique?... Or is there some reason why you're not able to at least offset that and maybe even get proactive and move margins even higher than what was in your model?"
— Timothy Arcuri, UBS

A: "We get price increases as it relates to increases in our cost structure. We deliver new capability to customers that meets their cost of ownership targets, and in that time, we get a chance to assess our costs relative to the pricing and make the adjustments we need to make... It's pretty hard to go back to your customers after you've taken orders and start to change prices on those orders."
— Bren Higgins, CFO

Assessment: Management did not dispute the premise, which is itself the answer. A twelve-month backlog priced at prior-period terms is the structural reason margin cannot respond quickly to input inflation, and it is the same feature that protects revenue on the way down. Investors should treat the backlog's virtues and this limitation as one thing, not two.

Whether the 2030 model is already obsolete

With 2027 wafer equipment being framed near $190B and the March Investor Day's 2030 figure at approximately $215B, a recurring line of questioning pressed on whether the long-term model still means anything. Management conceded the assumptions have moved and declined to update it four months after publishing it.

Q: "You're effectively suggesting WFE next year somewhere in the range of... around $190 billion, give or take. And your 2030 model had $215 billion, which isn't that much higher than that. I think it also had a semi market of $1.4 trillion that we seem likely to hit this year. Just any thoughts either qualitative or quantitative on that 2030 model..."
— Stacy Rasgon, Bernstein Research

A: "What we really were trying to do with 2030 was give... what would the industry look like if it's growing at a high single digit for the semi industry and capital intensity continues to go along at its level and process control strengthens and we gained share... if you envision that the market is going to be much higher than that, once we normalize for some of this pricing a little bit, then the process control intensity, we feel pretty confident about... we only had our Investor Day in March and here we are in July. So we're in no position to reset anything as it pertains to 2030."
— Rick Wallace, CEO

Assessment: The useful content of a target model is its structural assumptions, not its endpoint, and management said as much. The endpoint is stale to the upside; the intensity and share assumptions are the part to hold them to. Refusing to re-cut a model four months old is the correct instinct, particularly for a team that has been publicly wrong about its own market four times in seven months.

Chinese competition in process control

With domestic Chinese lithography progress driving a sector-wide de-rating, a question tested whether the read-across reaches process control, and what happens to share if import restrictions were ever lifted on the memory side. The answer was structural, resting on the shape of the market rather than on policy protection.

Q: "Recently, there have been more noise about more domestic competition, more and more in litho than in process control. But what makes the more for process control more difficult to replicate versus other tools? And let's say, if theoretically, the Chinese memory companies, right, who spend the bulk of their process control WFE with local suppliers... their import ban is lifted, does it mean that more of the WFE share will go to local suppliers?"
— Vivek Arya, Bank of America

A: "Process control is tricky to get into because it's a very high mix, low volume market. So unlike litho, frankly, is much higher kind of similar tool, higher volume, process control, there's a lot of nuance... Add to that the 1,600, 1,700 applications engineers that KLA has worldwide, and it's a pretty good competitive moat that we've established over time... And today, we continue to see whenever there is fair competition and we're allowed to compete, we do quite well."
— Rick Wallace, CEO

Assessment: The high-mix, low-volume, applications-heavy structure of process control is a genuine reason the lithography read-across is weaker than the market is pricing. What management did not offer, and was not pressed for, is any sense of the rate at which domestic capability is closing. The gap today is not the question; the derivative is.

What They're NOT Saying

1. A KLA-specific 2027 growth number: Management gave an industry consensus figure of roughly $190B for wafer equipment and said it plans for "significant growth," but declined to attach a company revenue growth rate to it, three months after volunteering unprompted 2027 commentary in April. Anchoring on someone else's forecast is a conspicuous change in posture.
2. Any refresh of the 2030 model: The March Investor Day's approximately $215B 2030 wafer-equipment figure is barely above the 2027 number management now cites, which implies near-zero growth for three years. Management acknowledged the tension and explicitly declined to reset anything.
3. That the capital-return target was missed: Fiscal-2026 capital returns of $3.348B were 88.9% of free cash flow against an Investor Day commitment of more than 90%. Neither the release nor the call addressed the shortfall.
4. How much worse the memory cost headwind actually is: "Probably a little bit more" than 100 bps, extended through 2027, is the third consecutive quarter this number has been revised the wrong way without ever being sized precisely.
5. Any China revenue figure or view on domestic capability: Management defended the process-control moat structurally but gave no China revenue disclosure, no assessment of how quickly domestic process-control capability is advancing, and no scenario for what happens if restrictions change.
6. What the working-capital build normalizes to: Receivables grew nearly twice as fast as revenue and free-cash-flow margin fell to 22.3% in the quarter, framed as a deliberate investment but with no target, no timeline, and no unwind assumption.
7. Anything about new-entrant fab projects: Asked directly whether KLA is engaged on a specific customer's fab-efficiency program, management declined to discuss any customer engagement, which is standard policy but leaves a visible, and increasingly relevant, gap.

Market Reaction

• Pre-print setup: KLAC closed at $190.80 entering the print, up 57.0% year to date and up 106.7% over the trailing twelve months, but down 23.3% over the trailing 30 days. The stock had closed at $301.71 on June 30, its 52-week closing high, and fell 36.8% in the 19 sessions to the print, including 11.8% and 11.5% single-session declines on July 1 and July 2 and a further 6.2% on print day itself, before the release. The S&P 500 was up 8.5% year to date.
• Reaction session: Shares opened the following session down 6.9% at $177.63, briefly touched $191.37 (+0.3%), then sold off all day to close at $170.19, down 10.8% ($20.61) and within $0.38 of the session low of $169.81 (-11.0%). Volume was 20.4M shares against a 14.1M 30-day average (1.5x). The S&P 500 fell 1.5% that session. The cumulative drawdown from the June 30 closing high reached 43.6%.
• Sector context: The decline was not idiosyncratic. Samsung Electronics fell 13.4% and SK Hynix 14.7% in the session before KLA reported, on a combination of Chinese deep-ultraviolet lithography development and concerns about the financing costs of AI infrastructure. A leading lithography peer had earlier declined roughly 10% after withdrawing its assurance of 2026 growth on macro and tariff uncertainty.

The mechanics of this sell-off are worth separating from its meaning. The June-quarter print beat consensus on revenue and earnings; the September guide midpoint is above the Street on both lines; the market forecast was raised for the fourth time this year; the advanced-packaging target was raised for the third consecutive quarter; and management pre-released a backlog figure it has declined to give for years. There is no line in the release or the call that a bear can point to as the trigger.

What happened instead is that the bar moved. Entering the print the stock was still up more than 100% over twelve months despite a month of violent de-rating, which meant a beat-and-raise of ordinary magnitude was priced. A 1.3% revenue beat and a roughly 2% guidance raise are good numbers in most tapes and insufficient in this one, because the market is not currently debating whether KLA executes. It is debating whether the wafer-equipment cycle peaks in 2027, and no quarterly print can settle that.

The shape of the session deserves its own note, because it argues against reading the move as a shrug. The stock printed its high in the first minutes, at $191.37, and then fell for the rest of the day to close $0.38 off the low on 1.5 times average volume. That is distribution, not a gap-and-stabilize. Buyers who wanted the raise had all day to take it and did not, and the S&P fell 1.5% alongside. The honest reading is that this tape is discounting the 2027 revenue KLA has already booked, and is willing to keep doing so until either the backlog stops growing or it becomes clear it will convert.

Note one distortion circulating in the coverage. Several summaries characterized the September outlook as a guidance miss by comparing the low end of KLA's guidance range ($3.8B revenue, $1.06 EPS) against consensus point estimates. On a midpoint basis the guide is above the Street on both lines. The stock did not fall because guidance disappointed; it fell because the sector is being repriced and this print was not large enough to interrupt it.

Street Perspective

Debate: Is 2027 wafer equipment near $190B a forecast or a cycle top?

Bull view: The visibility is unusually physical. Lead times of twelve months and up to 24 on some products, customer conversations already scheduling second-half 2027 deliveries, greenfield fabs under construction rather than tool adds to existing lines, and a backlog of roughly $12.5B growing consistently. Supplier capacity agreements now extend to 2029.

Bear view: A third consecutive year of mid-20s wafer-equipment growth would be without modern precedent, and every prior cycle has ended with participants describing structurally different demand. Backlog is not non-cancellable, and the company that raised its market view four times in seven months is not a credible detector of a turn.

Our take: The bull evidence is better than usual because it is physical rather than sentiment-driven, and fab construction is hard to reverse mid-build. But the bear point about forecast reliability is fair and management effectively conceded it. Our resolution is that 2027 is well protected by backlog and construction schedules, and 2028 is genuinely unknowable. That is a narrower window of certainty than the bulls claim and a wider one than the current multiple implies.

Debate: Does Chinese progress in lithography threaten process control?

Bull view: Process control is high-mix and low-volume, defended by algorithms, applications engineering and a large installed-base relationship rather than by a single manufacturable machine. KLA has been gaining overall share while excluded from the Chinese fabs where domestic substitution is most advanced, which is close to a controlled experiment.

Bear view: Every prior claim that a segment of semiconductor capital equipment was too complex to replicate has eventually been proven optimistic on a long enough horizon, and a domestic push that produces credible lithography will not stop there. Process control is also the segment where an adequate-but-cheaper tool at a large discount does the most damage.

Our take: The structural argument is real and the read-across from lithography is weaker than the tape is pricing, so we side with the bulls on the two-to-three-year view. But this is the risk we can least handicap from public disclosure, and management notably declined to characterize the rate of domestic improvement rather than the current gap. We would treat any specific evidence of Chinese process-control tool qualification at a leading-edge customer as a thesis-level event.

Debate: After a drawdown of more than 40%, is the multiple the opportunity or still the problem?

Bull view: At the post-print close the stock trades near 45 times trailing fiscal-2026 non-GAAP earnings and roughly 37 times the September guide annualized, against an implied December quarter that is materially higher again. The multiple has compressed by more than 40% from the June peak while forward earnings estimates went up.

Bear view: A high-30s multiple on peak-cycle earnings is not cheap, and semiconductor capital equipment has historically traded at its lowest multiple precisely when earnings are highest. The de-rating has been going on for a month and the market is usually early rather than wrong on cyclical peaks.

Our take: This is the debate that decides the rating, and we come down on the bull side with a caveat. The specific thing we identified in April as the primary risk was a stretched multiple, and roughly 44% of it has now been removed while the operating story improved three quarters running. The bear point about peak-cycle multiples is legitimate, but it applies to a business with a $12.5B backlog, a contracted service annuity approaching $3.5B annually, and share gains in packaging running at twice the market rate. That combination should compress less in a downturn than the multiple currently assumes.

Model Update Needed

Valuation impact: Estimates move up across calendar 2026 and 2027 while the share price fell, so fair value and price have moved in opposite directions for a second consecutive month. Anchored on the post-print close of $170.19, the stock trades near 45 times trailing fiscal-2026 non-GAAP earnings of $3.76 and roughly 37 times the September guide annualized, with the December quarter implied materially higher again by management's own second-half framing. We would frame a constructive twelve-month target roughly 20% to 30% above the post-print close, a range of about $204 to $221, which on a trailing-earnings basis twelve months forward implies a multiple in the high 30s to low 40s, below where the stock traded entering this print and far below the June peak. The framework rests on earnings delivery rather than multiple expansion. We would add on further sector-driven weakness that does not come with evidence of order cancellation, and we would reduce on any of three specific signals: backlog declining sequentially, a 2027 wafer-equipment view revised below $190B, or credible evidence of Chinese process-control qualification at a leading-edge customer.

Thesis Scorecard Post-Earnings

Overall: The thesis strengthened again on fundamentals, for the third consecutive quarter, and the residual risk moved almost entirely outside the company. Margin, which was the model's soft spot for four quarters, is now a positive. Advanced packaging, capital-equipment leverage and the newly disclosed backlog all improved. The one honest deterioration is that the share-gain claim is less visible in calendar-2026 revenue than April's framing implied, and rests on a supply explanation that 2027 will have to validate. Against that, the single risk we named as primary in April, a stretched multiple, has been substantially resolved by a drawdown the company did not cause.

Action: Maintain Outperform. A business raising its outlook for a third straight quarter, disclosing a growing $12.5B backlog, guiding margin up against a worsening input cost, and trading 44% below a high set a month earlier is the risk/reward this rating is designed for. Conviction is unchanged rather than higher, because the improvement in price is offset by a genuine increase in sector-level uncertainty we cannot handicap from public information. We would add on further sector-driven weakness absent order cancellations, and we would revisit on a sequential backlog decline, a 2027 market view cut below $190B, or credible evidence of Chinese process-control qualification at the leading edge.