Qualcomm Hit Every Line of Its Own Guide and Fell Anyway: the Squeeze Moves From Volume to Margin, and Apple Leaves Early

Results vs. Consensus

Three months ago Qualcomm guided the September quarter's predecessor to a handset trough of roughly $4.9 billion, put a date on the China bottom, promised ~50% automotive growth, and told investors the December quarter would carry the first data center revenue. Every one of those came in at or better than promised. The stock fell 4.4% on the day of the print before a single number was released, then another 2.6% on the reaction session and 2.6% again the next day.

That is not a market rejecting the quarter. It is a market repricing the two things the quarter revealed that were not in the guide: that the memory shock has stopped being a volume problem and become a cost-of-goods problem, and that Apple is leaving faster and larger than the number Qualcomm had been repeating for years.

Q3 FY26 scorecard

Year-over-year comparison

Sequential comparison (vs. Q2 FY26)

Revenue

Total revenue of $9,947M fell 4.0% YoY and 6.2% sequentially, landing at the high end of the $9.2-10.0B guide and roughly 3% above the ~$9.67-9.69B Street mark. The composition is the entire diversification thesis in one line: handsets contributed $5,086M, down 19.6% YoY, while automotive and IoT combined contributed $3,418M, up 28.3%. Automotive alone added $604M of YoY revenue; handsets subtracted $1,242M. Qualcomm is now growing everything except the segment that still supplies 60% of chipset revenue, and the arithmetic of that has not yet turned positive.

Margins

This is where the quarter turned. Gross margin of 53.1% fell 251bps YoY, and cost of revenues actually rose 1.4% on 4.0% lower revenue, which is the clearest single expression of what input-cost inflation does to a fabless model with fixed near-term pricing. At the segment level QCT EBT margin fell to 25.8% from 29.7%, and QTL slipped to 68.9% from 71.5%. Operating expense compounded it: R&D rose 17.1% and SG&A rose 26.6% against a shrinking revenue base, so GAAP operating margin compressed 1,030bps to 16.3%. Management attributes the gross-margin damage to two forces, a weaker mix within the premium tier as OEMs down-select to cheaper or prior-generation silicon to absorb memory costs, plus the broad-based input-cost increase itself. Both are real. Only the second is being addressed with pricing.

EPS

Non-GAAP EPS of $2.21 was down 20.2% YoY and 16.6% sequentially, and came in one to two cents below consensus while sitting slightly above the midpoint of the $2.10-2.30 guide. The penny miss is immaterial. What matters is that the September guide of $2.05-2.25 puts the earnings low point one quarter later than management framed it in April, when the Q3 trough was supposed to be the bottom in both revenue and earnings. Revenue does now trough in Q3; earnings do not. That distinction, between a volume trough that arrived on schedule and a margin trough that has been pushed out, is the honest summary of the quarter.

Segment Performance

QCT plus QTL of $9,782M reconciles to the $9,947M consolidated total with $165M in QSI and other non-reportable segments. Segment revenue figures above are the disaggregated QCT revenue streams as filed in the 8-K, and the QCT Total row is the filed segment figure rather than a sum of rounded call commentary.

QCT Handsets — the bottom held, and it cost more than it was supposed to

Handsets fell to $5,086M, down 19.6% YoY. That is $186M above the roughly $4.9B trough management guided in April, which means the specific, falsifiable commitment made last quarter was met. Management went further and confirmed the mechanism it described then had played out as described: Chinese-OEM handset revenue reached its bottom in this quarter and returns to double-digit sequential growth next quarter, because channel inventory has thinned to the point where OEMs can no longer draw it down and shipments must reconcile back to true end demand.

"consistent with our expectations, we estimate that QCT handset revenues from Chinese OEMs reached a bottom in the third fiscal quarter and will return to double-digit sequential growth in the fourth quarter." — Akash Palkhiwala, EVP, CFO & COO

The complication is that total handsets do not follow the China curve. Next quarter's ~$5.2B handset guide represents only ~2% sequential growth, because Android's double-digit recovery is being netted against an Apple decline that is now much steeper than the model carried. The recovery is real; it is simply arriving into an offsetting outflow.

Assessment: Management earned credibility here and should be given it. The bottom was called one quarter in advance, with a stated mechanism, and it landed slightly better than promised. But the value of a dated trough is that it lets you underwrite the recovery slope, and that slope is now controlled by Apple's exit rather than by China's return. The segment has stopped being a cyclical question and become a structural one.

QCT Automotive — the one part of the story that keeps outrunning its own guidance

Automotive delivered $1,588M, up 61.4% YoY against a guide of roughly 50%, its 23rd consecutive quarter of double-digit YoY growth, and grew 19.8% sequentially into a quarter when the rest of the chipset business shrank. Management raised the exit run-rate target for fiscal 2026 from the $6 billion figure given in April to approximately $7 billion, and signed what it describes as a landmark expansion with BMW covering both ADAS and digital cockpit.

"This quarter, we signed a landmark expanded agreement with BMW, winning a highly competitive selection process to become the lead compute silicon provider for their next-generation ADAS as well as digital cockpit. This agreement represents a material expansion of our automotive pipeline and establishes Qualcomm as the lead compute silicon partner for BMW extending across model programs well into the next decade." — Cristiano Amon, President & CEO

The structural point management keeps making, and which the numbers now support, is that automotive engagements have shifted from socket-by-socket awards to whole-platform, multi-generation commitments. The fifth-generation Snapdragon Digital Chassis begins ramping in September, and content per vehicle steps up sharply with it. Management also restated the ambition to become the largest automotive semiconductor supplier by revenue.

Assessment: This is the segment we would underwrite with the most confidence, and it is the one place where execution has consistently exceeded the guide rather than merely met it. A raise from $6B to $7B on the exit run-rate one quarter after setting it is a meaningful upward revision, and a decade-length lead-compute position at a premium German OEM is a durable annuity rather than a design win. At roughly 19% of chipset revenue it is now large enough to matter, and it is the single strongest argument against treating Qualcomm as a handset cyclical.

QCT IoT — bifurcating, with the industrial half doing the work

IoT grew 8.9% YoY to $1,830M and 6.0% sequentially, in line with the high-single-digit guide. The composition is increasingly split: industrial networking and robotics is growing double digits and drove the quarter, while consumer categories, tablets in particular, are being squeezed by the same memory constraint hitting handsets. Management guided IoT to approximately flat YoY next quarter for exactly that reason, with industrial double-digit growth offset by consumer weakness. The forward case rests on the design-win pipeline rather than current revenue.

"our industrial design win pipeline exceeds $7 billion with over $3.5 billion in design wins secured this fiscal year. This reflects strong customer demand and a meaningful increase in new businesses." — Cristiano Amon, President & CEO

Assessment: The step from high-single-digit growth to flat is a real deceleration and the consumer exposure is a genuine drag, but the industrial pipeline disclosure is the more informative number: $3.5B of multi-year design wins secured within a single fiscal year, inside a total IoT segment currently annualising near $7.3B of revenue. Design wins are bookings rather than revenue and convert over several years, but a pipeline above $7B against that revenue base is a credible forward signal. IoT remains the second, slower diversification lever behind automotive, and its value is still more about optionality in robotics and edge AI than about the revenue it prints today.

QTL — the annuity holds, but the royalty cap blunts the pricing offset

Licensing revenue of $1,278M fell 3.0% YoY and 7.5% sequentially with EBT margin at 68.9%, all inside guidance. The interesting disclosure was structural rather than numerical: asked whether the industry-wide device price increases would flow through to royalties, management explained that the royalty base is capped by device ASP, so increases below the cap accrue some benefit and increases above it do not.

"you should just think of QTL consistent with the model it has, right, which is the -- there is a cap on the total royalties based on the device ASP. And so to the extent that prices go up below the cap, there is some benefit that accrues to QTL. And then above the cap, it really doesn't make a difference in the revenue for QTL." — Akash Palkhiwala, EVP, CFO & COO

Assessment: QTL remains the high-margin annuity that funds the diversification investment, and its resilience through a low-teens decline in the handset market is genuinely impressive. But the cap mechanic means the segment is structurally unable to inflate with device prices, so investors should not model a pricing-driven QTL tailwind from the September increases. The larger unaddressed question, the Apple license renewal, went unmentioned for the second consecutive quarter.

Key Topics & Management Commentary

Overall Management Tone: Composed and unusually specific about bad news, quantifying the damage rather than deflecting it: the Android revenue hit put at more than $1.50 of EPS, the data center gross-margin drag at 150 to 200 basis points, the pricing response at double digits. The conviction on the long-term story was unchanged from Investor Day, but the framing had shifted from selling the vision to conceding that the market now wants proof, a more defensive register than April's. Where management was least convincing was timing: every question about when margins recover or when memory normalizes was answered in "over the next couple of quarters" and "over time" language rather than with dates, and the fiscal-2027 cost question went unanswered for the second consecutive quarter.

1. Input-cost inflation and the double-digit price increase

The defining topic of the quarter, and the one that turned a met-guidance print into a sell-off. Qualcomm is absorbing cost increases across essentially every stage of its supply chain simultaneously, and its near-term pricing is locked by existing contracts and product cycles. The gap between the two is landing directly in gross margin.

"In the short term, the entire industry continues to be impacted by unprecedented memory prices, higher manufacturing and input costs as well as supply chain shortages driven by overall data center demand. In addition to the resulting revenue decline in mobile and consumer electronics, this is creating short-term pressure on QCT gross margins which will be slightly below our historical range." — Cristiano Amon, President & CEO

The response is a broad price increase, and management put a magnitude on it for the first time on this call, framing it as an industry-standard pass-through rather than an opportunistic move.

"it's no different than what a lot of the peers in our industry have done, and you should expect something that the scale of the increase that we're looking at is double digit and consistent with some of the actions from other players." — Akash Palkhiwala, EVP, CFO & COO

Note that the existence of the increase was already public before the print; press reporting in the final week of July described notices to hardware partners taking effect September 1. What was new on the call was its magnitude, its slow phase-in, and the size of the margin hole it is meant to fill.

Assessment: Two of the three variables here are known and one is not. The cost increase is known, the pricing response is now quantified, and the timing of the offset is not: contracts and product cycles mean the benefit "will come up over time," with the baseline 48-50% QCT gross margin range restored only "when we get through it." A margin recovery with no date attached is exactly the shape of disclosure that cost this stock 8.5% in February, and management has reproduced it here on a different variable.

2. The Apple exit accelerates, and the number that was supposed to be bounded is not

The single most damaging disclosure of the call, delivered inside the prepared remarks rather than in response to a question. Qualcomm's modem share in the fall iPhone launch will now be materially below the 20% figure the company had held for years, and the revenue effect compounds through fiscal 2027.

"as a result of our supply constraints, we now expect an acceleration in the step-down of Apple product revenues starting in the fourth fiscal quarter as our share for upcoming iPhone launch is expected to be materially lower than our prior estimate of 20%." — Akash Palkhiwala, EVP, CFO & COO

The quantification followed in Q&A: Apple product revenue declines roughly 50% from the September quarter to the December quarter, and fiscal 2027 Apple product revenue will be below the "little over $2 billion" previously indicated. Management attributes the change to its own supply constraints and to where negotiations landed, not to a competitive loss.

Assessment: The attribution matters less than the fact. Last quarter we moved structural handset erosion to contained specifically because the Apple decline was bounded at 20% share and roughly $2B of FY27 revenue, described by management as an assumption held consistent for years. One quarter later both bounds moved the wrong way. Whatever the cause, the practical consequence is that the company's most-repeated fixed input is no longer fixed, and the Street has to widen its confidence interval around every handset line accordingly. This is the deterioration that justifies not upgrading on the price move.

3. The China handset bottom held exactly as promised

The commitment we flagged as the primary thing to grade this quarter, and management delivered it. The mechanism it described in April, channel inventory drawdown exhausting itself so that shipments reconcile up to true end demand, played out on schedule.

"earlier in the year, the OEMs were buying based on the size of the market, but they were also drawing down on channel inventory. And where we are at now is the channel inventory has thinned down so that they can't draw down anymore. And so we are reconciling -- the revenue is reconciling to the size of the total market. And so that kind of change in profile of purchases is what gives us confidence." — Akash Palkhiwala, EVP, CFO & COO

Management also reaffirmed the bottom under direct challenge on pricing elasticity, tying it to the observation that Qualcomm's own price increase is small relative to the memory increases already in the bill of materials.

Assessment: Full credit. A specific, dated, falsifiable call was made one quarter in advance and it proved correct, with the mechanism management named as the reason. That is the strongest evidence this year that the licensing-derived demand visibility management claims is real rather than rhetorical, and it should raise the weight investors place on the company's forward statements. It is also, unfortunately, the only forward-looking commitment from April that could be graded this cleanly.

4. Two hyperscaler custom-silicon programs, with purchase orders and wafers in motion

The data center story advanced from one disclosed hyperscaler engagement to two, and from a stated intention to ship in December to contracted orders and started production. This is the most concrete de-risking step available short of actual revenue.

"we have 2 custom chip engagements, and both of these are global scale hyperscalers and we're going to expect to start seeing revenue from both of them starting in the December quarter. As you know, the December quarter is right there. And so we do have POs from these engagements, and so we've already started wafers. So we're very confident about the engagement with both the customers." — Akash Palkhiwala, EVP, CFO & COO

Management framed both as first phases of multi-year relationships expected to expand, and sized the total data center opportunity at $5 billion in fiscal 2027 rising to more than $15 billion in fiscal 2029.

Assessment: Purchase orders and wafer starts are a materially higher evidentiary standard than the "multi-generation engagement" language of April, and the count doubling from one customer to two is a real proof point. The disclosure gap that remains is the same one as last quarter: no named customers, no per-customer dollar sizing, and no split between the two. The FY27 $5B target is now the number the entire growth case rests on, and it is being carried on management's word plus two unnamed purchase orders.

5. The margin cost of entering the data center

The disclosure that reframed the data center story from unambiguous positive to a trade-off, and which management volunteered with more precision than the question required. The first data center revenue is custom ASIC work, structurally the lowest-margin flavor of the business.

"our baseline business has a certain gross margin range, which has been in the 48% to 50% range. And so with the price increases coming online over time, we expect to be at that range. And really, the data center revenue coming in since the first revenue is mostly from custom chip engagements. We do expect that to be significantly lower than our ongoing, kind of, baseline gross margin percentage. And so that will be a drag of 1.5% to 2% on the weighted average gross margin for QCT." — Akash Palkhiwala, EVP, CFO & COO

Assessment: This is the most important number on the call for anyone modelling fiscal 2027, and it changes the sequencing of the thesis. The pricing action is supposed to restore QCT gross margin to 48-50%; the data center ramp then removes 150-200bps of it. The two largest gross-margin variables therefore work against each other through the entire recovery window, which means gross margin does not simply return to trend as memory normalizes. Note also the asymmetry management has flagged over two quarters: in April the custom-silicon win was described as accretive at the operating-margin level, and it is now confirmed as dilutive at the gross-margin level. Both can be true, and investors modelling the gross line need the second one.

6. HBC Gen 1 taped out, and the merchant accelerator moves from slideware to silicon

Beyond the custom ASIC work sits the differentiated part of the data center roadmap, a high-bandwidth compute architecture that integrates compute directly with high-density memory. Management reported a concrete engineering milestone.

"I'm pleased to report that we have completed the tape-out of HBC Gen 1, an engineering milestone that moves us into the next phase of customer engagements. We expect to demonstrate HBC performance on silicon in the coming quarters ahead of the launch of our first HBC solution in mid-2027." — Cristiano Amon, President & CEO

Management was candid that the merchant conversations, which span accelerators, SerDes connectivity, and server CPUs, are gated on customers physically seeing working silicon rather than on further commercial persuasion.

"I know everybody will live in an era of everybody wants instant gratification. But the reality is there's a lot of customers that want to see silicon." — Cristiano Amon, President & CEO

Assessment: A tape-out is a real, dated milestone and the honesty about what gates the pipeline is welcome. But it also sets the correct expectation: the merchant business, which is where the higher-margin dollars and the path from $5B to $15B live, cannot convert until silicon demonstrations land, and the first HBC product does not launch until mid-2027. Everything between now and then is custom ASIC at depressed margin. That is a long carry.

7. Modular closes, and Qualcomm bids for the open-software layer

The Modular acquisition, announced at Investor Day in June, closed on the day of the print. The stated ambition goes well beyond adding software headcount.

"We have the ambition to change the current industry approach to AI software from closed to open systems to promote enhanced competition, innovation and resilience." — Cristiano Amon, President & CEO

The consideration was 17,826,566 shares issued in a private placement, roughly 1.7% of shares outstanding, registered for resale two days after close. That is the second acquisition-driven expansion of the cost base this year alongside Alphawave, and both now sit inside the ~$2.7B quarterly non-GAAP operating expense guide.

Assessment: Strategically this is the right instinct: a hardware-agnostic software layer is the only credible answer to the incumbent's software moat, and buying it is faster than building it. Financially it lands at an awkward moment, adding operating expense and equity dilution in the same quarter that gross margin broke and the buyback halved. The ambition is a multi-year bet; the cost is immediate and in this quarter's guide.

8. The fiscal 2027 arithmetic, in management's own numbers

The most consequential framing of the call, and it is a subtraction problem. Non-handset revenue growth accelerates from 24% in fiscal 2026 to more than 60% in fiscal 2027, and management states plainly what that growth is for.

"The way you should think about '27 is really the growth that we are targeting in the non-handset areas, which we said is going to be greater than 60% on a year-over-year basis will replace the entire Apple product revenue within the year." — Akash Palkhiwala, EVP, CFO & COO

Against that, management still expects the company to grow: "Despite these headwinds, we expect top line growth for QUALCOMM in fiscal '27, driven by an inflection in non-handset revenues throughout the fiscal year." Both statements can hold only if Android handset recovery supplies the incremental growth, since by management's own construction the non-handset inflection is fully consumed offsetting Apple.

Assessment: This is the sentence that should govern how anyone models fiscal 2027, and it is considerably less exciting than the $40B fiscal 2029 headline it sits next to. The greater-than-60% non-handset acceleration, the single most impressive number Qualcomm has produced this cycle, does not add a dollar of growth in the year it happens. Fiscal 2027 growth therefore depends entirely on an Android recovery whose pace is hostage to memory pricing that management has now twice declined to forecast. The growth story is real and it is a fiscal-2028 event.

9. The FY29 targets were nearly doubled, and the stock did not care

At the June 24 Investor Day Qualcomm raised its fiscal 2029 non-handset revenue outlook to more than $40 billion from the $22 billion target set in November 2024, comprising more than $24 billion of automotive and IoT plus more than $15 billion of data center, with non-handset rising past 50% of chipset revenue in fiscal 2027 and roughly two-thirds by fiscal 2029.

"We also updated our fiscal 2029 financial targets, which now include more than $24 billion in revenue across automotive and IoT plus more than $15 billion in data center, bringing our total non-handset revenue outlook to $40 billion by fiscal 2029, up from our previous target of $22 billion." — Cristiano Amon, President & CEO

The market's verdict on that disclosure is already in the tape. The stock closed at $204.13 the day before Investor Day and $189.39 two sessions after it, and has fallen every month since. Management acknowledged the credibility gap directly rather than arguing with it.

"Our data center business is just at the beginning of its journey, and we recognize that investors want to see more proof points that we can successfully execute on our plans as a new entrant." — Cristiano Amon, President & CEO

Assessment: Doubling a four-year-out target is the kind of disclosure that re-rates a stock when the base business is stable and does nothing when it is not. Qualcomm delivered the largest upward revision to its long-term outlook in years and lost 41% of its value in the two months that followed, which tells you the market has repriced the discount rate on Qualcomm's forward promises rather than the promises themselves. Management's acknowledgment of that is the right posture; the only thing that changes it is shipping.

10. Supply chain: comfortable on volume, exposed on cost

Asked directly about wafer availability, management drew the analogy to the pandemic and argued that Qualcomm's scale and multi-node footprint is a competitive advantage in a shortage, as is the inventory it has built.

"the industry now is probably operating very similar to what was in the pandemic. Everything is at 100% utilization, everything. So we're seeing a shortage in price increases across wafers, across assembly, across testing -- testers, everything." — Cristiano Amon, President & CEO

The balance sheet corroborates the inventory point. Inventories stood at $8,379M at June 28, up 28.4% from $6,526M at fiscal year-end, built while revenue was falling. Management explicitly frames that as strategic positioning rather than a demand miss.

Assessment: The scale argument is credible and the volume risk does appear contained, which matters because supply is what constrained the Apple share this quarter. But a 28% inventory build into a declining revenue line is a two-sided fact: it is either the strategic hedge management describes or a working-capital drag that becomes a margin problem if the Android recovery underwhelms. It consumed $1,798M of operating cash flow in the first nine months. We accept the strategic framing for now and flag it as a line to watch next quarter.

11. Capital return decelerated in the quarter that needed it least

Qualcomm returned $2.3 billion in the quarter: $973 million of dividends at $0.92 per share, and $1.4 billion repurchasing 8 million shares, implying an average repurchase price around $175. Against a nine-month repurchase total of $6,806 million, the first half averaged roughly $2.7 billion of buyback per quarter, so the pace roughly halved in the June quarter.

Across the first nine months the company returned $9,674 million against free cash flow of $6,827 million (operating cash flow $8,405M less capital expenditures $1,578M), or 142% of free cash flow. That gap shows on the balance sheet: cash plus marketable securities fell from $10,155M at fiscal year-end to $8,304M, while total debt rose from $14,811M to $15,270M, taking net debt from $4,656M to $6,966M.

Assessment: The buyback slowed in the quarter when the stock was falling from $251 toward $150, and the repurchases that did happen were executed at an implied ~$175, above where the stock now trades. The company is also returning materially more than it generates, funded by drawing down securities and adding short-term debt, while simultaneously issuing 17.8 million shares for Modular. None of this is alarming at Qualcomm's balance-sheet quality, but the $20B authorization is doing less work at exactly the moment it should be doing more, and the "structural bid under the stock" argument we credited last quarter needs to be marked down accordingly.

Guidance & Outlook

The shape of this guide is the analytical centre of the quarter. Revenue is guided up sequentially at the midpoint and slightly above the Street. Earnings are guided down and roughly 9% below the Street. There is no demand problem in this outlook. There is a 180 basis point further compression in QCT segment margin, gross margin still running below its 48-50% baseline, and operating expense stepping up to ~$2.7 billion as Alphawave and Modular consolidate alongside continued data center investment ahead of revenue.

Implied trajectory: Revenue troughed in the June quarter and grows from here; earnings trough in the September quarter. Beyond that, the December quarter is guided qualitatively to be slightly up sequentially, with a 50% decline in Apple product revenue offset by Android growth and the first data center revenue. Management explicitly retired the company's historical seasonality: the December quarter is no longer the annual peak, and revenue is now expected to grow from December into March.

Street at: Consensus entering the print sat at roughly $10.02B and $2.36 for the September quarter. The revenue number was broadly right; the earnings number was not, and the gap is entirely margin. Fiscal 2027 estimates are the ones now in motion, since the Apple step-down, the pricing phase-in, and the custom-ASIC margin drag all land in that year and none were in the pre-print model.

Guidance style: Qualcomm has now guided conservatively and delivered at or above the high end for two consecutive quarters, including calling a handset bottom one quarter in advance and hitting it. The pattern argues for treating the $2.05-2.25 range as achievable rather than as a stretch. The risk in this guide is not the range; it is that the margin recovery beyond it has no date.

Analyst Q&A Highlights

When gross margins return to their prior range

The first question of the call went straight to the issue that defined the print, and asked management to bridge the timing gap between input costs rising now and average selling prices rising later. The answer named two causes and no date.

Q: "It's pretty clear. You explained what was going on with the rising input costs and now you're raising prices. Can you walk us through how we should think about QCT gross margins returning to their prior levels, how long between the ASP increases kind of match the input costs rising?"
— Joshua Buchalter, TD Cowen

A: "we're taking action to increase the prices and reflect it in our customer product pricing. And we expect this benefit to show up in our gross margins over the next couple of quarters. These changes, as you would expect, come in gradually as we have some contracts in place, and so we need to get past those contracts. There are also product cycles that happen. So this will come up over time. But I think when we get through it, we expect to be consistent with the historical gross margin range we have."
— Akash Palkhiwala, EVP, CFO & COO

Assessment: Management answered the mechanism fully and the timing not at all. "Over the next couple of quarters," "gradually," "over time," and "when we get through it" are four hedges in one answer to a question that asked for a duration. The second cause disclosed here, a weaker mix within the premium tier as OEMs down-select to cheaper silicon, is arguably more concerning than the input costs, because pricing actions do not fix mix. That half of the compression persists as long as memory keeps inflating handset bills of materials.

Magnitude of the price increase and whether the market can absorb it

A follow-on line of questioning pressed on how large the increases would be and how a company raises prices into a market already suppressed by component cost inflation. Management sized the action for the first time, then made an elasticity argument in its defence.

Q: "can you just flesh out the pricing actions, like any commentary in terms of the magnitude of price increase that you're looking to take? ... And just as we think about maybe pricing actions in some of these consumer markets like handsets, how do you navigate rising prices in a market that has already seen demand affected by cost inflation and other components?"
— Joseph Cardoso, JPMorgan

A: "the market is actually down because of the magnitude of increases in the bill materials with memory. So even a double-digit price increase, which is just a pass-through of the input cost increase and wafer price increases, it's actually small when you compare it to the order of magnitude of the memory bill of materials. So we actually don't expect that fundamental changes in the premium tier and the higher tier volume. And we maintain the position that China handset this Q3 is the bottom."
— Cristiano Amon, President & CEO

Assessment: The elasticity argument is sound and the framing is honest: Qualcomm's increase is genuinely small relative to memory's contribution to the bill of materials. But the argument concedes the more important point, which is that the handset market is down because of costs Qualcomm neither controls nor benefits from, and the company's pricing response does not change that trajectory. It defends the volume assumption while confirming there is no volume upside from here.

Whether Qualcomm is walking away from Apple rather than losing it

The sharpest exchange on the call, and the one that most directly tested the credibility of the supply-constraint explanation. The question put the uncomfortable interpretation on the record explicitly: that the company is choosing to redirect constrained silicon to better-margin customers and accelerating the exit on its own terms.

Q: "The wording on the paragraph in the slide, it sounded like you guys were making the conscious choice not to sell to them nearly as much as you were before. I mean, you sort of blame the supply -- is that true? I mean I don't want to be too dramatic, but are you basically starving Apple, getting them out quicker than you could have and using that silicon to send it elsewhere?"
— Stacy Rasgon, Bernstein Research

A: "Yes. I wouldn't characterize it as such, Stacy. You should think of it as supply -- our supply constraints were a part of it and then where discussions ended up is that we'd have a share less than -- materially less than 20%, and our revenue as a result in '27 from Apple product would be less than the previous guidance we had given, which was a little over $2 billion."
— Akash Palkhiwala, EVP, CFO & COO

Assessment: A denial that concedes most of the substance. "Supply constraints were a part of it" and "where discussions ended up" together describe a negotiated outcome in which Qualcomm had limited allocation and limited incentive to fight for a departing customer's volume. The commercial logic is defensible and probably even correct. The thesis consequence is unchanged either way: the FY27 Apple revenue floor moved down, and the company's willingness to let it move down means investors should stop treating it as a floor at all.

Whether the data center ramp offsets the gross-margin recovery

A follow-up question connected the two margin stories that had been discussed separately, asking whether the incoming data center revenue eats the pricing-driven recovery. Management not only confirmed it does, but quantified the drag.

Q: "I know the gross margins, hopefully, get better over the next several quarters, but you also have data center ramping as well. And you sort of talked to it at the Analyst Day that data center stuff should be dilutive to gross margin. So I mean does it offset some of that gross margin recovery? Or how do I think about that?"
— Stacy Rasgon, Bernstein Research

A: "our baseline business has a certain gross margin range, which has been in the 48% to 50% range. ... And really, the data center revenue coming in since the first revenue is mostly from custom chip engagements. We do expect that to be significantly lower than our ongoing, kind of, baseline gross margin percentage. And so that will be a drag of 1.5% to 2% on the weighted average gross margin for QCT."
— Akash Palkhiwala, EVP, CFO & COO

Assessment: The most useful disclosure of the call and the one that should most change models. Management deserves credit for volunteering a specific number to a question that invited vagueness. The implication is structural: the recovery to 48-50% and the data center ramp are simultaneous and opposing, so blended QCT gross margin does not return to the historical band on the timeline the pricing action alone would suggest. Anyone modelling a clean margin snapback into fiscal 2027 is modelling only half the equation.

Whether the data center pipeline extends beyond the two committed customers

A question aimed at the gap between two contracted programs and a $15 billion fiscal 2029 target asked what the reception has been outside the disclosed engagements. The answer separated the committed revenue from the pipeline, and was unusually direct about what gates the latter.

Q: "In June, you had enough visibility to double your target. And just wondering if how is the reception outside of the 2 customers? And when you talk about the pipeline and some of the commentary around that, are you implying that there could be more than 2 customers? And how do you see that expanding beyond the current reach?"
— Benjamin Reitzes, Melius Research

A: "there's a lot of customers that want to see silicon. And that's both true on the accelerator on our HBC plus the accelerator as well on the CPU. ... In the coming quarters, we're going to have silicon, and we are going to be able to do silicon demonstration evaluation. I think that has the potential to unlock new opportunities. We have additional conversation with customers. But the next milestone is Qualcomm, we needed to see silicon and see how it performs before we make a decision that we're marching forward."
— Cristiano Amon, President & CEO

Assessment: Honest and appropriately unpromotional, and it establishes the correct milestone for tracking the story: silicon demonstrations in the coming quarters, not further commercial announcements. It also confirms that the path from $5 billion to $15 billion runs through merchant products that have no customers committed yet. The two contracted programs are real; everything beyond them is currently gated on a chip that has taped out but not been demonstrated.

The fiscal 2027 handset market and the size of the memory drag

A question about whether the handset market can recover to the growth rate laid out at Investor Day if memory prices stay elevated produced the largest single quantification of the call, and a notable non-answer on the year it asked about.

Q: "how are you thinking about handset market TAM for Qualcomm in 2027 if memory prices continue to be a headwind. Is it going to continue to kind of remain suppressed? Or can it rise up to the kind of 5% CAGR that you outlined at the Investor Day?"
— Liam Pharr (for Vivek Arya), Bank of America

A: "at this point, where we are looking at for the handset market is we think the market will be down low teens relative to '25 -- to '26. And a lot of the impact is in the lower tiers. ... The net impact on QCT Android revenue is -- the revenue is down 20% year-over-year and the EPS impact of it is greater than $1.50. So the way we think about it is longer term as conditions stabilize. And with our Snapdragon leadership and the agentic AI transformation of phones, we have the opportunity to reinstate the scale of the handset business and the growth that comes with it."
— Akash Palkhiwala, EVP, CFO & COO

Assessment: More than $1.50 of earnings per share is a very large number against a fiscal 2026 base tracking near $10.51, and management is right to frame it as a recoverable tailwind rather than a permanent impairment. But note what the answer does: it quantifies fiscal 2026's damage in detail and then declines to say anything about fiscal 2027, which is what was asked. That is the second consecutive quarter the fiscal 2027 memory question has been deflected, and it is now the single largest unquantified variable in the model.

The December quarter and the end of Qualcomm's historical seasonality

Asked for early colour on December-quarter handset seasonality given the offsetting Apple and China dynamics, management declined to guide but disclosed something more structurally important: the company's annual revenue shape has permanently changed.

Q: "can you provide any early thoughts on how you're thinking about handset seasonality into the December quarter, just given kind of the moving pieces here around, obviously, Apple and then the recovery on the China OEM side of things?"
— Joseph Cardoso, JPMorgan

A: "we are forecasting approximately 50% decline from September to December quarter [in Apple product revenue]. This obviously accelerates, kind of, the exit of Apple revenue out of our model. The second is Android. We expect Android to grow and significantly offset the reduction in Apple and then data center revenue that starts to ramp in the first quarter. ... you should not think of previous seasonality where first quarter was the high quarter for us. The profile obviously changes with Apple having a very strong December quarter, not in our model anymore. So we actually expect revenues to grow when we go from the December to the March quarter now."
— Akash Palkhiwala, EVP, CFO & COO

Assessment: A genuinely useful modelling disclosure buried in a routine seasonality question, and the clearest single confirmation that the Apple relationship is now a rounding error rather than a swing factor. The December quarter ceasing to be the annual peak is what an Apple exit looks like in the shape of the P&L. It also flatters fiscal 2027's optics, since a business that grows December-to-March will look better sequentially than one that always fell off a December peak, without anything improving underneath.

What They're NOT Saying

1. No date for the gross-margin recovery: the pricing action is quantified at double digits and the baseline range at 48-50%, but the return to that range is described only as "over the next couple of quarters," "gradually," and "when we get through it." The first question of the call asked for a duration and did not get one.
2. No fiscal 2027 memory call, for the second consecutive quarter: asked directly whether the handset market can recover to the Investor Day growth rate if memory stays elevated, management answered with fiscal 2026 figures and said nothing about fiscal 2027. That was also the shape of the answer in April.
3. No fiscal 2027 margin or earnings framing at all: management supplied a detailed fiscal 2027 revenue mix (non-handset above 50% of chipset revenue, greater-than-60% non-handset growth, $5B of data center) and not one line on what it earns. Given that the year contains a pricing phase-in, a 150-200bp custom-ASIC drag, and a step-up in operating expense, the omission of the margin side of a year whose revenue side was disclosed in detail is conspicuous.
4. Still no named data center customers or per-customer sizing: two "global scale hyperscalers" with purchase orders is a stronger disclosure than April's single unnamed engagement, but the $5B fiscal 2027 target remains unattributable to any customer, product, or contract the market can verify.
5. Nothing on the Apple licensing renewal or Huawei: QTL's largest structural questions went unmentioned for the second consecutive quarter. That silence is more conspicuous now than last quarter, because the Apple product relationship just deteriorated materially and management's standing position is that the royalty is an independent business unaffected by the chip decline. That claim was not retested on this call.
6. The Samsung framing softened without comment: in April management described the relationship as reset to "greater than 70%" share and stated that framework held for that year and the next. This quarter the language was Snapdragon powering "approximately 70% of their flagship devices." The two statements are not strictly comparable, one being a planning framework and the other an observation about launched devices, but the direction of the change went unremarked.
7. No explanation for the buyback deceleration: repurchases roughly halved sequentially, to $1.4B from a first-half average near $2.7B per quarter, in the quarter the stock fell hardest. Neither the prepared remarks nor Q&A addressed the pace, and no analyst asked.

Market Reaction

• Pre-print setup: QCOM closed at $155.68 on July 29, already down 38.0% from its 52-week closing high of $251.02 set on May 29, down 9.0% year to date, down 17.5% over the trailing 30 days, and down 3.9% over the trailing twelve months. The stock fell 4.42% on the day of the print itself, before the release. The S&P 500 was up 6.9% year to date entering the print.
• After-hours move: shares fell roughly 7% to 8% in the extended session, trading near $144.50, on the Q4 earnings guide and the Apple disclosure.
• Reaction session (July 30): the stock gapped down 5.9% to open at $146.55, traded a $146.00 to $154.16 range, and closed at $151.60, down 2.6% or $4.08, on 24.3 million shares against a 17.6 million 30-day average (1.4x). It recovered roughly half the after-hours decline intraday.
• Follow-through session (July 31): the stock opened at $152.95 and closed at $147.61, down another 2.6%. The two-session post-print move is -5.18%, leaving the stock 41.2% below its May 29 high, down 13.7% year to date, and 19.0% above its 52-week closing low of $124.07.
• Relative and peer context: the S&P 500 rose 1.66% on July 30 and 0.70% on July 31, so QCOM underperformed by roughly 428bps and 333bps respectively, about 755bps cumulative. The semiconductor tape was strongly positive on the reaction session, with the memory complex in particular rallying hard.

The reaction is not a verdict on the quarter, which met or beat every line of the company's own guidance. It is a verdict on two disclosures that changed the fiscal 2027 model: a margin trough with no stated end, and an Apple exit that broke a bound management had held for years. The market has been repricing this stock for two months on exactly those two questions, and the print resolved neither in its favour.

The peer tape sharpens the point. On the same session Qualcomm fell 2.6% on margin compression caused by memory and input-cost inflation, the memory complex posted one of its largest single-day gains of the year. Qualcomm sits on the paying side of the AI infrastructure buildout: the same demand that is inflating its bill of materials is being monetised elsewhere in the supply chain, and its own data center revenue, the mechanism by which it eventually sits on the receiving side, is still one quarter away and margin-dilutive when it arrives.

Worth separating from that is what the pre-print setup already contained. This stock had fallen 38% before the release, through an Investor Day that nearly doubled its four-year non-handset target, and through public reporting of the September price increase. Very little of the sell-off is the print. Most of it is a two-month de-rating of the multiple the market will pay for Qualcomm's forward promises, and the quarter neither arrested nor accelerated that in any decisive way.

Street Perspective

Debate: Is this a cyclical margin trough or a structural reset in earnings power?

Bull view: the compression is cyclical and self-correcting. Input costs are being passed through with double-digit price increases already notified, the baseline 48-50% gross margin range is management's stated destination, and the memory-driven Android damage has been quantified at more than $1.50 of earnings per share that reverses as the cycle normalises. Buying a fabless franchise at 14x trough-cycle earnings with a quantified normalisation tailwind is what cyclical investing looks like.

Bear view: two of the three compressive forces are not cyclical. The premium-tier mix shift, where OEMs down-select to cheaper silicon to absorb memory costs, is not fixed by raising prices, and the custom-ASIC gross-margin drag is a permanent feature of the business mix Qualcomm is deliberately building toward. Add an operating expense base stepping to ~$2.7 billion per quarter from two acquisitions, and the earnings power that emerges on the other side of the cycle is structurally lower than the one that went in.

Our take: the bear case is more right than the bull case on composition, and the bull case is more right on magnitude. Blended gross margin will not return to trend on the pricing action alone, because the data center mix removes 150 to 200 basis points of whatever pricing restores, and management has now confirmed both halves. But more than $1.50 of recoverable earnings per share is a very large offset against a base near $10.51, and it is real. The honest position is that fiscal 2027 earnings power is lower than the pre-print consensus and higher than the post-print panic, which is precisely a Hold.

Debate: Does the data center story justify owning the stock through the transition?

Bull view: the evidentiary standard has risen every quarter. One unnamed hyperscaler in April became two with purchase orders and started wafers in July, the HBC accelerator has taped out for a mid-2027 launch, the fiscal 2029 non-handset target nearly doubled to $40 billion, and an open-software layer was acquired to attack the incumbent's real moat. Buying that pipeline while the stock trades 41% off its high, before the first revenue prints, is the whole point.

Bear view: the first data center dollars are the lowest-margin flavour of the business, they dilute gross margin by 150 to 200 basis points, and by management's own arithmetic the greater-than-60% non-handset growth in fiscal 2027 does nothing but replace departing Apple revenue. The path from $5 billion to $15 billion runs through merchant products with no committed customers, gated on a silicon demonstration that has not happened. Investors are being asked to fund a dilutive ramp for two years for growth that shows up in fiscal 2028.

Our take: the direction is no longer in serious doubt and the execution evidence keeps improving, which is a genuine change from twelve months ago. What is in doubt is the price of the wait. The transition year is dilutive to margin, neutral to growth, and dependent on an Android recovery hostage to memory pricing management will not forecast. That is a reasonable thing to own and an unreasonable thing to pay a premium for, and at $147.61 the market has stopped paying a premium without yet offering a discount.

Debate: Is the Apple step-down finally in the numbers, or is there more to come?

Bull view: this is the clearing event. The 20% share assumption that hung over every model is gone, fiscal 2027 Apple product revenue is now guided below $2 billion, the December quarter loses roughly half of it sequentially, and the annual seasonality has already been reset to reflect a business without Apple. Everything from here is a cleaner, Android-and-non-handset company, and the overhang that suppressed the multiple for three years is being removed rather than deferred.

Bear view: the number moved once and can move again. Management held 20% share and roughly $2 billion of fiscal 2027 revenue as a stable assumption "consistent for years," and revised both within a single quarter, attributing it to supply constraints and where negotiations landed. An input that behaves that way is not a floor, and the licensing royalty, which is far larger and whose renewal has now gone undiscussed for two consecutive calls, is the same relationship with more dollars attached.

Our take: the product revenue probably is close to cleared, and the bulls are right that a known zero is worth more than an uncertain decline. But the bears identify the correct residual risk, and it is not the chip revenue. It is that the same counterparty whose product relationship just deteriorated faster than modelled also holds a licensing agreement worth multiples of it, and management has stopped talking about the renewal. We treat the product step-down as largely in the numbers and the licensing renewal as the larger unpriced risk in the name.

Model Update Needed

Valuation impact: Fiscal 2026 non-GAAP earnings per share now track to approximately $10.51, computed as the $8.36 delivered across the first nine months plus the $2.15 guide midpoint for the fourth quarter. At the July 31 close of $147.61 that is 14.0x fiscal 2026 earnings, against 17.1x on the same earnings base when we last wrote at $179.58. Fiscal 2027 is the harder year to underwrite, because it combines a partial and undated gross-margin recovery, a 150 to 200 basis point mix drag from the custom-ASIC ramp, an operating expense base stepping higher, and revenue growth that management's own framing puts at modest since the non-handset inflection is consumed offsetting Apple. We therefore assume fiscal 2027 earnings roughly flat with fiscal 2026 rather than growing, and apply a 15-17x multiple appropriate to a transition year with a credible but unproven growth vector. We cut our 12-month fair-value range to $150-$180 from $165-$195. Against the $147.61 close that range implies roughly +2% at the low end and +22% at the high end, with a midpoint of $165 implying about +12%.

Thesis Scorecard Post-Earnings

We grade the standing thesis carried since initiation, against this quarter's print and call.

Overall: The thesis is more polarised than it was, not uniformly weaker. Both bull pillars that describe the future of the company strengthened on real evidence, and management's single most important forward commitment from April, the dated China handset bottom, was met precisely. But both bear pillars deteriorated on the axis that determines the next two years of earnings: the memory shock stopped being a volume problem and became a margin problem with no stated end, and the Apple bound that justified calling structural erosion contained moved materially against the company within one quarter. The valuation leg that broke in April has been restored by a 41% drawdown, which is the reason this is not a downgrade.

Action: Maintain Hold, with fair value cut to $150-$180 from $165-$195. We are deliberately not executing the upgrade we telegraphed last quarter at "a pullback toward the low-$160s." That trigger presumed a pullback in price against an intact earnings base; what arrived was a pullback with fundamental deterioration in the two variables that set fiscal 2027 earnings. We would upgrade to Outperform on either of two specific developments: a dated commitment to the 48-50% gross-margin range with the data center drag disclosed inside it, or a successful HBC silicon demonstration that converts a named merchant customer beyond the two contracted programs. We would move to a more cautious stance if the December data center revenue slips, if the Android sequential recovery fails to deliver double-digit growth in the September quarter as guided, or if the Apple licensing renewal surfaces on terms below the current royalty scale.