Strip the Tariff Refund and the Beat Was Two Cents: Memory Takes 280bps of Gross Margin in Two Quarters; Downgrading to Hold

Results vs. Consensus

This is the first quarter of our coverage arc where the reported scorecard and the underlying scorecard tell materially different stories, and reading the difference correctly is the whole exercise. Apple beat on revenue, set June-quarter records across the top line, iPhone, Mac, Services and every geography, and printed the highest company gross margin in its history. It also disclosed, in the same breath, that roughly 2 percentage points of that margin and $0.11 of that EPS came from tariff refunds. Adjust for those and the quarter lands where the company guided it three months ago and roughly where the Street had it.

Year-over-Year Comparison

Sequential Comparison (vs. Q2 FY26, March quarter)

Quality of Beat/Miss

• Revenue: Real, demand-led, and understated by the constraint. The 16.4% growth is broad rather than concentrated: five of five geographies grew double digits, four of five product categories grew, and iPhone and Mac both set June-quarter records while management said supply, not demand, capped both. There is no acquisition contribution and no channel-fill argument available here. The one honest deduction is FX, which management characterized as a sequential headwind rather than a tailwind this quarter, so the constant-currency figure is if anything higher than reported. This is the strongest line in the print and the reason the downgrade is not a demand call.
• Margins: The reported 50.06% is the highest company gross margin in Apple's history and it is not repeatable. Roughly 2pp is tariff refunds. Products gross margin of 40.07% is up 555bps YoY, but over 2.5pp of that is the same refund, putting underlying expansion nearer 300bps. Services gross margin of 75.62% was essentially flat YoY and down 106bps sequentially on mix, which matters because Services is the mix ballast the company leans on when hardware margin compresses. Operating margin of 32.62% (+263bps) is genuine operating leverage, but it absorbs an R&D line growing 32.3%, more than twice the pace of SG&A, which is the deliberate AI investment ramp and is not going to slow.
• EPS: $2.02 (+28.7%) decomposes into operating strength plus three helpers of varying durability. Net income grew 27.1% and the share count shrank 1.6%. Below the line, other income swung to a positive $572M from a negative $171M a year ago, a roughly $743M pre-tax swing, and the effective tax rate rose 146bps to 17.9%, a partial offset. Strip the $0.11 refund and the underlying figure is roughly $1.91, up about 22% YoY. That is still a strong result. It is not the 29% the headline advertises.

Segment Performance

Apple reports a product-category cut and a geographic cut. This quarter the two cuts diverge in an informative way. The product cut shows concentration: iPhone and Mac supplied essentially all of the incremental growth while iPad shrank and Services decelerated. The geographic cut shows the opposite, an unusually even distribution in which the two largest non-US regions grew at virtually identical rates. Taken together, the picture is a hardware cycle running hot everywhere and a Services layer losing pace globally.

By Product Category

By Geography

iPhone

The engine, running its fourth consecutive quarter above 20%. iPhone grew 21.7% to $54.25B, a June-quarter record achieved while supply-constrained on the advanced-node SoCs the company has now flagged for three straight quarters. Apple set June-quarter records in every geographic segment and a June-quarter record for upgraders, gained share globally per IDC, and cited third-party survey data showing iPhone as the top-selling model in the US, urban China, the UK, France, Australia and Japan. US customer satisfaction on the 17 family was measured at 99%.

"iPhone revenue for the June quarter was $54.3 billion, up 22% from a year ago. We achieved June quarter records in every geographic segment and set a June quarter record for upgraders. According to IDC, we gained share globally during the quarter." — Tim Cook, CEO

Assessment: This remains the healthiest line in the model and it is the single reason the downgrade stops at Hold. Four quarters of 20%-plus growth with measured share gains and record upgraders is durable capture, not a launch-window effect. Two things now sit against it. The September guide steps iPhone to mid-teens, roughly a 700bp deceleration, which management attributes to FX and supply rather than demand. And the comp stiffens materially from here: the year-ago September quarter grew 6.1% and the December quarter that follows was itself a record, so the base Apple laps in FY27 is the hardest of the cycle. Management again declined to characterize post-cycle durability.

Mac

The upside surprise of the quarter and the clearest evidence that the constraint is a forecasting failure rather than a supply failure. Mac grew 28.7% to $10.35B against a Street model near $9.3B, its best June quarter ever, on MacBook Neo and MacBook Pro. Apple set an all-time Mac revenue record in Greater China, all-time records for upgraders and for customers new to Mac, and reported that roughly half of MacBook Neo large purchases by US education institutions displaced Windows and Chromebook devices. The CFO named specific enterprise and education deployments including a 25,000-student district transition in Florida and an 8,000-student Chromebook replacement in Washington.

"Mac delivered its best June quarter yet with $10.4 billion in revenue, growing an impressive 29% from a year ago despite significant supply constraints. This revenue growth was driven by the incredible strength of our latest lineup with MacBook Pro and the all-new MacBook Neo." — Tim Cook, CEO

Assessment: The most under-appreciated line in the print, and the one that best supports management's "demand forecast issue" framing. A 28.7% print against a 12% Street model, in a category that grew 12.7% in the year-ago September quarter, is a genuine share shift with an institutional displacement story behind it. That displacement is stickier than consumer upgrade demand because school districts and enterprises standardize for years. The offset is that Mac is now also a price-increase category and a constrained category, so FY27 conversion of this demand depends on both elasticity and node availability.

Services

The quarter's real disappointment, and the one that changes a thesis pillar. Services grew 12.1% to $30.74B, a June-quarter record in absolute terms but below the roughly $31.4B consensus and a clear step down from 16.3% in March. The CFO's walk from 16% to 12% named FX as the main driver, plus the absence of a theatrical release comparable to the prior year's F1 film, softness in mobile gaming within the App Store, App Store business-model changes in certain countries, and the continuing US court ruling on link-out transactions. Underneath, the structural markers stayed strong: paid subscriptions passed 1.5 billion, transacting and paid accounts hit all-time highs, and cloud services and payment services set all-time records.

"I think when I step back, if I look at how we landed versus our expectations that we had outlined in the March quarter for the June quarter, I would say that we roughly met our expectations, but we did see a bit more softness on mobile gaming in the App Store." — Kevan Parekh, CFO

Assessment: Two of the five named causes are cyclical (FX, the film comp) and three are structural (mobile gaming, App Store business-model changes, the link-out ruling). The September guide compounds it: applying the disclosed 2.5pp sequential FX headwind to the 12.1% just reported implies roughly 9.5% reported growth, a figure an analyst put to management on the call as "below 10% year-over-year as reported" without contradiction. A Services line growing under 10% at a 75.6% gross margin no longer functions as the mix ballast that offsets hardware margin compression, which is precisely the job the thesis assigned it. This is the pillar that moved from confirmed to at risk.

Greater China and the Geographic Cut

Greater China grew 22.4% to $18.82B, a June-quarter record, with a China Mainland record and an all-time Mac revenue record in the region. What is more notable is the symmetry of the geographic cut: Europe grew 22.4% and Greater China grew 22.4%, effectively identically, while Rest of Asia Pacific grew 15.6% and Japan 13.4%. Americas, the largest segment at $45.78B, was the laggard at 11.1%.

"We achieved June quarter revenue records in every geographic segment. We were pleased to see strength across the board with June quarter records in the U.S., Latin America, Western Europe, India, China Mainland, Japan, and Southeast Asia." — Tim Cook, CEO

Assessment: The China bear pillar stays retired. Three consecutive quarters of strong double-digit growth (38% in December, 28% in March, 22% now) is a trend, though the sequence also shows the rate normalizing as the easy base laps, which is what we said we would not extrapolate. The more useful signal is the Americas line: 11.1% growth in Apple's largest and most saturated market, against 22.4% in Europe and China, tells you where the incremental cycle demand is actually coming from and where the September price increases will be tested first.

iPad and Wearables

iPad fell 5.9% to $6.19B, the only declining category, which the CFO attributed to the difficult compare against the prior-year A16-powered iPad launch. The installed base still reached an all-time high and over half of iPad buyers were new to the product. Wearables, Home and Accessories grew 6.5% to $7.88B with growth in every geography, a June-quarter record for Apple Watch upgraders, and over half of Watch buyers new to the category.

Assessment: Neither is a thesis driver, but iPad now carries a specific risk it did not carry last quarter: it is simultaneously a declining category, a price-increase category, and a newly supply-constrained category in the September guide. That combination makes it the most likely place for the elasticity question to show up first in reported numbers. The new-to-category rates on both lines remain the quiet strategic value, since they feed the installed base that Services monetizes.

Key Topics & Management Commentary

Overall Management Tone: Management was confident on the business and unusually plain-spoken about the cost problem, a combination that produced the sharpest risk disclosure of our coverage arc. Where prior calls hedged the memory question with framework language, this one quantified it twice, named it as more than the entire cause of two consecutive sequential margin declines, and reached for "100-year flood" to describe the environment. The countervailing shift is defensive: on forward visibility management contracted rather than expanded, refusing December color twice, declining to discuss long-term supply agreements, and leaving price elasticity, Siri economics and AI capital intensity unanswered. Tim Cook, in his final call, carried the disclosure himself rather than routing it through the CFO, which read as deliberate.

1. Memory: A "100-Year Flood" and 280 Basis Points of Gross Margin

This is the quarter's defining disclosure and the direct cause of the downgrade. Management laid out an escalating four-quarter cost chronology: Apple paid more for memory in the March quarter than in December, significantly more in June than in March, and expects to pay even higher costs in September. The CEO then extended the warning past the guide period, which is the part that matters most for FY27.

"For September, we expect to pay even higher memory costs, and we're able to offset partly by a few factors… If you look beyond September, we see the market pricing for memory continuing to increase, which could drive an increasing impact on our business, and we're continuing to evaluate this." — Tim Cook, CEO

The CFO then supplied the arithmetic, and it is unusually specific for Apple. Adjusted company gross margin walks from 49.3% in March to 48.1% in June to a 46.5% September midpoint, and on both legs he attributed more than the entire decline to memory.

"We look at our gross margin change from the 49.3% we had at the total company level for the March quarter, and as you mentioned, the 48.1% adjusted for the tariff refund in the June quarter, that 120 basis point change. You look at the drivers of that, more than 100% of that can be explained by the memory cost change that Tim outlined… That 160 basis point change is also more than that is explained by the change in memory cost." — Kevan Parekh, CFO

Assessment: Two hundred eighty basis points of adjusted gross margin in two quarters, attributed by management to a single input cost, with a stated expectation that the input keeps rising after the guide window and with the named offsets explicitly decaying. This is no longer the "live H2 variable" we carried last quarter; it is a realized, quantified compression with an open back end. The FQ2 recap said we would revisit conviction higher on "evidence the gross margin can out-run the memory ramp into H2." The evidence came in, and it went the other way.

2. The September Guide and the Refusal to Frame December

Apple guided September revenue to grow 9% to 11% year over year, against a Street bar of roughly 12% and a June quarter that just grew 16.4%. Management attributed the step-down to two factors: a 2.5pp sequential FX headwind and a significant sequential increase in supply constraints now spanning iPhone, Mac and iPad. The more consequential moment came when management was pressed on what the constraints do to the December quarter, where the Street was already modeling 8% to 9%.

"Beyond September, we're not providing any color at this point. I think you mentioned December, so we're not providing any kind of color or guidance beyond the September quarter." — Kevan Parekh, CFO

The refusal is consistent with Apple's long-standing one-quarter guidance practice, so it is not a change of policy. It is a change of usefulness. Both the memory cost curve and the supply constraint were described as worsening beyond the guide window, and the December quarter is Apple's largest.

Assessment: Declining to frame December is defensible in a normal quarter and expensive in this one. Management volunteered that both of the variables driving the September step-down get worse afterward, then declined to size either. The market is left extrapolating a deteriorating trend with no anchor, on the last call before a CEO transition. That combination is a legitimate reason to require a lower multiple, independent of any view on demand.

3. The Constraint Is a Forecasting Miss, Not a Supply Failure

The most genuinely bullish disclosure on the call, and the reason this is a Hold rather than something worse. Pressed repeatedly on whether the supply problem implicated a partner or a fab, the CEO rejected the framing outright and located the cause in Apple's own demand planning.

"The root cause of it is not a regular supply issue. It's a demand forecast issue, to be candid, where the iPhone and the Mac are both doing remarkably better than we thought they would do. We had high expectations, so it wasn't that our expectations were low… The supply chain just has less flexibility in it than normal. We've been pulling supply ahead. At some point, there's a limit to that. We've got a quarter that we're going to be scrambling on the supply side, essentially." — Tim Cook, CEO

Asked separately whether broadening silicon sourcing could relieve it, the CEO again declined the premise and pointed to the Arizona fab ramp, where Apple expects to source over 100 million components this year.

Assessment: "Scrambling on the supply side" is not language Apple uses lightly, and the candor is worth something. If the bottleneck is a demand forecast miss, the September guide is a shipping cap set against unmet orders rather than a demand cap, and the deferred volume converts later rather than disappearing. We give this full weight in the revenue line and none of it in the margin line, because a constrained quarter with rising component costs is the worst combination for gross margin: you carry the cost inflation without the volume leverage that would normally absorb it.

4. Apple Takes Price, Reluctantly

Apple raised prices on iPad and Mac during the quarter, a notable break from a company that has historically absorbed component swings rather than passing them through. The CEO framed it as forced by the memory environment and declined to state whether the objective was protecting gross profit dollars or gross margin percentage.

"On the pricing front, we reluctantly raised prices, I would say. We did it because we're in what I would characterize as a 100-year flood on the memory pricing, with exponential increases in memory prices. That was the rationale for it. In terms of our philosophy on dollars or percentages, we look at units, revenue, and margin, then come to a business judgment as to how to handle that." — Tim Cook, CEO

On elasticity, management was explicit that it does not yet know the answer, citing channel inventory and consumer response lags.

"Obviously, we've now had to increase prices on iPad and Mac, the price elasticity there, it's just too early to come to a definitive conclusion of what happens there. Because it takes a little while for the channels to adjust since there's channel inventory, it takes a while for the consumer to respond. We'll understand that more in the weeks ahead." — Tim Cook, CEO

Assessment: The pricing action is the right response and it introduces a new unmodeled variable at the same time. It also lands on the two categories least able to absorb it: iPad is already declining and Mac's growth is partly an institutional displacement story where budget-constrained districts are the buyer. The September quarter is therefore the first clean test of whether Apple's pricing power extends to categories outside iPhone, and management has told us it will not have the answer until after the quarter is underway.

5. Services Decelerates from 16% to 12% to Roughly 9.5%

The CFO's walk from March's 16% to June's 12% named FX as the main driver, with the prior-year F1 theatrical release comp, mobile gaming softness, App Store business-model changes in certain countries, and the US link-out court ruling as contributors. He then quantified the FX drag across the full stretch.

"We expect foreign exchange to drive about a 5 percentage point headwind to the year-over-year growth rate from the March to September quarter. If we look at the sequential change from the June quarter, the 12% services we reported to what we're guiding for the September quarter, we are going to see another two and a half point sequential headwind." — Kevan Parekh, CFO

Management noted the Supreme Court has agreed to hear Apple's appeal of the link-out ruling, and separately that the App Store still set a June-quarter revenue record despite the headwinds.

Assessment: The FX component is transitory and should be looked through. The App Store components are not: mobile gaming softness, unilateral business-model changes in certain countries, and a link-out regime that persists at least until the Supreme Court rules are all structural pressure on the highest-margin line in the company. The thesis pillar said Services compounds mid-teens on roughly 77% margins. It is now compounding at roughly 9.5% on 75.6% margins. That pillar is at risk, and the timing is unfortunate: Services was supposed to be the mix offset to hardware margin compression.

6. Siri AI Ships, but Not in Europe or China

The AI overhang that gated our original Hold has largely resolved on timeline. Siri AI was unveiled at WWDC, developer and public betas are out, and management described the response as strongly positive and the fall launch as on track. The gating is geographic rather than temporal, and it lands on the two regions that matter most to the narrative.

"You look at the EU, we're working closely with the commission. Obviously, our complete desire is to launch everything everywhere at the same time. That's always the philosophy that we have. We have not been able to do that in the European Union, but we're working closely with them to try to get to something that would allow us to offer Siri AI there. It is offered or will be offered for the Mac there, because the Mac is not covered by the same regulations as the iPhone and the iPad." — Tim Cook, CEO

On China, management said Apple received approval last week to ship the original Apple Intelligence features, with Siri AI requiring further work.

"You look at China, last week we received approval to ship sort of the original features of Apple Intelligence, things like cleanup and so forth. We're working now through the rollout of those, and there will be more work required down the road for Siri AI. We're at the front end of that." — Tim Cook, CEO

Assessment: Genuine progress on the product, and a real commercial limitation. Europe and Greater China together produced $48.21B this quarter, 44% of company revenue, and both grew 22.4%. If Siri AI is the demand driver for the fall iPhone cycle that management implies, it is being withheld from the two fastest-growing regions in the business, on regulatory timelines Apple does not control. "We're at the front end of that" is not a schedule. We move the AI bear point from de-risking to neutral rather than to resolved.

7. Cook's Final Call and a 32-Day Handoff

Tim Cook confirmed at the top of Q&A that this was his last earnings call, with John Ternus taking over. Ternus joined the call and spoke only once, in response to a question about the competitive landscape, in roughly forty words.

"As you know, this will be my final earnings call, and John will lead these calls going forward. The transition is going seamlessly, and I am beyond excited for John to step into his new role and lead Apple into its next era." — Tim Cook, CEO

Assessment: Last quarter we called the handoff orderly and well-telegraphed, and nothing this quarter contradicts that. What changed is the operating environment the successor inherits. Ternus takes over on September 1 into a quarter management describes as one it will spend "scrambling on the supply side," with a component cost shock management calls a 100-year flood, a first-in-cycle pricing action of unknown elasticity, and no December guidance framework in place. Cook's two scarcest skills, supply-chain mastery and geopolitical navigation, are precisely the two the next four quarters will demand most. We keep this as a contained risk rather than an escalating one, but the timing has become adverse in a way it was not in May.

8. Capital Return Normalizes and the Balance Sheet Absorbs the Shock

The buyback returned to scale after last quarter's transition-blackout dip. Apple returned $33B to shareholders in the quarter, including $4B in dividends and equivalents and $25.8B in share repurchases, against $11B of repurchases in March. The board declared a $0.27 dividend payable August 13. Cash and marketable securities ended at $147B against $84B of total debt.

"We ended the quarter with $147 billion in cash and marketable securities and $84 billion in total debt. During the quarter, we returned $33 billion to shareholders. This included $4 billion in dividends and equivalents and $25.8 billion in share repurchases." — Kevan Parekh, CFO

Assessment: The pillar holds and the blackout read from last quarter was correct. Nine-month repurchases of $62.09B against $11.78B of dividends, with shares outstanding down 1.1% since fiscal year-end and diluted shares down 1.6% YoY, keep the buyback a dependable 1.5 to 2 point EPS tailwind. This is the one thesis pillar that strengthened, and it is doing real work: it is roughly a quarter of the EPS growth in a quarter where gross margin is under pressure. It does not offset a 280bp margin slide, but it cushions it.

9. The Balance Sheet Tells the Memory Story Before the P&L Does

The most informative disclosure of the quarter was not spoken on the call. Inventories nearly doubled since fiscal year-end, rising from $5,718M at September 27, 2025 to $11,092M at June 27, 2026, a $5,374M build. The cash flow statement shows inventory consuming $5,461M of operating cash over nine months against a $1,223M source in the prior-year period. This is the "carry-in inventory" the CEO repeatedly cited as the offset to memory inflation, sitting visibly on the balance sheet.

"The first is, as you would expect, we have a benefit from some carry-in inventory in the September quarter. However, we believe we'll see decreasing benefit from this over time beyond the September quarter." — Tim Cook, CEO

Assessment: Apple pre-bought roughly $5.4B of components ahead of the price curve, which is exactly the right operating response and is why June and September margins are cushioned at all. It is also a depleting asset with a disclosed expiry. Once the low-cost inventory clears, the reported cost of sales converges on spot memory pricing that management says is still rising. The balance sheet is therefore a leading indicator here: the inventory line tells you the size of the buffer, and management has told you it runs out after September. That is the mechanism by which the FY27 margin question becomes concrete.

10. AI Spend Runs Through the P&L, Not Capex

Asked directly whether Siri AI changes Apple's capital intensity, the CEO described a hybrid model and declined to commit to a direction, noting that AI spending appears in cost of sales as well as operating expenses.

"We use a hybrid model, as I know Kevan has reviewed with you earlier. We use some third-party cloud, and we do our own data centers. There will be a mix. Generally speaking, as you know, we have been growing our OpEx and spending more in AI in general and quite a bit more. There are other locations on the P&L other than OpEx, like COGS etc., that also have AI expenditures." — Tim Cook, CEO

The filings answer what the call did not. Over the first nine months of FY26, R&D grew to $34,035M from $25,684M, up 32.5%, while payments for acquisition of property, plant and equipment fell to $6,799M from $9,473M, down 28.2%. Nine-month operating cash flow of $116,996M grew 43.1%, and free cash flow after capex reached $110,197M, up 52.5%.

Assessment: Apple is running the most capital-light AI strategy of any hyperscale-adjacent company, and it is doing so while its capex line actually shrinks. That is a real structural advantage against peers spending tens of billions on training infrastructure, and it is the main reason free cash flow is compounding faster than earnings. The unresolved question is whether it is a choice or a constraint. If Siri AI usage scales the way management hopes, the hybrid model implies rising third-party cloud costs landing in cost of sales, at the same time memory is already there. Management would not size it, and the iCloud+ upsell it floated as the offset has no disclosed economics.

11. Broadcom, Houston and the $600B US Commitment

Apple announced a new multi-year agreement with Broadcom for custom silicon components and wireless connectivity technologies, described as exceeding $30B and the company's largest-ever American manufacturing program commitment. The Apple Advanced Manufacturing Center in Houston opens shortly, in a facility already assembling AI servers, with Mac mini production moving there later this year.

"The new multi-year agreement with Broadcom, which is part of Apple's American Manufacturing Program, is expected to exceed $30 billion. This marks our largest-ever American manufacturing program commitment. It's also an important step forward in our work to build an end-to-end silicon supply chain here in the U.S." — Tim Cook, CEO

Assessment: Strategically sound and near-term immaterial to the model. The domestic silicon build-out is a multi-year hedge against exactly the node-availability constraint capping shipments today, and it gives political cover for the tariff-refund reinvestment commitment. But none of it addresses memory, which is the actual cost problem: DRAM and NAND supply is not a US manufacturing question on any timeline that matters to FY27 gross margin. The announcement answers a 2029 risk while the 2027 risk goes unaddressed.

Guidance & Outlook

The guide is weaker than the Street expected on revenue and weaker than it looks on margin. On revenue, +9% to +11% against a September-quarter base of $102,466M implies roughly $111.7B to $113.7B, and a midpoint near $112.7B. Measured against the 16.4% just delivered, that is a deceleration of 540 to 740 basis points, roughly 640bp at the midpoint. Management attributes it to a 2.5pp FX headwind and a significant sequential increase in supply constraints, which together account for most but not all of the step-down.

On margin, the 47% to 48% range embeds approximately 1 percentage point of tariff refund against roughly 2 points this quarter. The comparable underlying figure is a 46.5% midpoint, which the CFO used himself, and it sits 160bp below the 48.1% underlying result just reported and 280bp below the 49.3% printed in March.

Street at. Consensus entering the print had September at roughly 12% revenue growth and the December quarter at roughly 8% to 9%, figures put to management on the call and left uncorrected. The 9% to 11% guide therefore sits below the Street at every point in the range, including at the top. The implication is downward September and FY26 revenue revisions, and larger downward EPS revisions than revenue revisions given the margin guide. The December bar is the more dangerous one: it was already modeling deceleration, and management declined to say whether that deceleration is enough.

Guidance style. Apple guides to ranges it expects to meet or beat, and it has cleared its own guide in each of the last four quarters. Two features of this guide argue it is beatable on revenue: it is set against a stated shipping cap rather than a demand estimate, and unfulfilled demand carries forward. One feature argues the margin guide is not conservative: the CFO gave a specific, mechanical attribution to a component price he does not control and said the offsets shrink. We read revenue as credible-to-beatable and margin as the honest number.

Analyst Q&A Highlights

Decomposing the September Deceleration

The opening question of the call went straight to the guide, asking how much of the roughly 500 basis point step-down was supply constraint versus FX, and what the constraints actually were. Management split the answer, with the CFO handling the arithmetic and the CEO the cause, and made a claim worth marking: that FX and supply together substantially bridge the gap back to the June growth rate.

Q: "If I think about the 9%-11% sort of growth that's guided for September, it's about a 500 basis point or so deceleration versus what we've seen really in June quarter, even through this year, I would say. Can you just talk about how much of this decel is really supply constraint versus other factors like FX? If you could just flesh out what these supply constraints are that are broadening beyond these advanced SOCs you talked about last quarter as well."
— Amit Daryanani, Evercore ISI

A: "We expect foreign exchange to be a sequential headwind of around two and a half percentage points to the year-over-year total company growth rate. The second impact is that the impact from supply constraints is we expect that to increase significantly when we go sequentially from June to September. That projected supply constraint in the September quarter will affect the iPhone, Mac, and the iPad. Really, when you combine those two factors, we get pretty close to the June overall total company growth rate."
— Kevan Parekh, CFO

Assessment: The claim that the two factors bridge "pretty close to" the June rate is the most important assertion in the guide, and it is unfalsifiable as stated because the supply-constraint magnitude was never quantified. FX is disclosed at 2.5pp of the roughly 640bp midpoint deceleration, leaving about 390bp attributed to a constraint whose size management has declined to size for two consecutive quarters. If the attribution is right, underlying demand is running near 16% and the guide is deeply conservative. If it is partly right, the December quarter is where the difference shows up. Management gave the market no way to check.

Memory Sourcing and a Three-Supplier DRAM Market

The follow-up moved from the guide to the structural question underneath it: whether Apple's reported push for greater memory sourcing flexibility is about securing supply or about mitigating price. The answer contained the clearest statement of the constraint's nature, and an unusually direct admission of limited leverage.

Q: "There are reports that suggest that you're seeking greater sourcing flexibility for memory. Can you just talk about, is this sourcing options really about ensuring that you have supply and it's a way to mitigate memory inflation? Or is it more to preserve the value proposition for your customers?"
— Amit Daryanani, Evercore ISI

A: "In terms of the sources of supply, primarily the DRAM market has three suppliers. Obviously if there were more suppliers, that would be good, and it would help us on the supply side and perhaps the pricing side. It's unclear on the pricing side, it could help on the supply side. We're evaluating all options."
— Tim Cook, CEO

Assessment: Apple rarely concedes limited negotiating leverage, and this is close to a concession. In a three-supplier market with demand pulled by AI infrastructure buyers, the world's largest consumer electronics purchaser is saying more suppliers "would be good" and that even then the pricing benefit is "unclear." That reframes the memory problem from a procurement execution issue, which Apple would historically solve, to an industry structure issue, which it cannot. It is the strongest single argument that the margin pressure persists into FY27 rather than normalizing.

December Visibility, Requested Twice and Declined

A recurring line of questioning tried to extract any framing for the December quarter, noting that the Street was already modeling a step down to 8% to 9% and that the after-hours reaction reflected uncertainty about whether the supply constraints extend into it. Management addressed the constraint qualitatively and then closed the door on the quarter explicitly.

Q: "The Street had you at 12% growth for the quarter, September. You could argue that to get to the 10% you guided, that's just the FX. In the 1st quarter, which is December, Street has you decelerating quite a bit, to like 8%-9%. I think what we're struggling with after hours here is how much do these supply constraints really hit you in December and impact that because the Street did a pretty good job of taking that number into the higher single digits."
— Benjamin Reitzes, Melius Research

A: "The dynamics Tim mentioned, combined with the foreign exchange impact I mentioned earlier sequentially, is really what's built into the 9%-11% guidance we're giving for the September total company revenue year-over-year growth rate. Beyond September, we're not providing any color at this point. I think you mentioned December, so we're not providing any kind of color or guidance beyond the September quarter."
— Kevan Parekh, CFO

Assessment: The question was well-constructed and it went unanswered. It correctly identified that FX alone bridges the guide to the Street number, which implicitly asks whether the supply constraint is incremental to the Street's existing December deceleration or already in it. Management neither confirmed nor denied. Combined with the disclosure that both memory costs and supply constraints worsen past September, the silence functions as information: the two variables the company says are deteriorating are the two it declines to project.

Pricing Philosophy and Long-Term Supply Agreements

One exchange asked two questions that go to the heart of the FY27 margin path: whether Apple will lock in memory pricing through multi-year supply agreements, and whether its pricing actions aim to defend gross profit dollars or gross margin percentage. Management answered the second and did not engage the first at all.

Q: "Is it your intention to pursue some of these multi-year LTAs with your suppliers just to ensure access to supply at pre-agreed prices? When you approach product pricing in this environment, is it your intention to protect product gross profit dollars? Is it your intention to protect product gross margins?"
— Erik Woodring, Morgan Stanley

A: "On the pricing front, we reluctantly raised prices, I would say. We did it because we're in what I would characterize as a 100-year flood on the memory pricing, with exponential increases in memory prices… In terms of our philosophy on dollars or percentages, we look at units, revenue, and margin, then come to a business judgment as to how to handle that. It's not a mathematical formula that gets us to a specific result or just looking at one dimension of that. We look at all three dimensions and think about it over the long term, instead of a 90-day clock."
— Tim Cook, CEO

Assessment: The long-term-agreement half of the question disappeared entirely, and it was the more valuable half. Pre-agreed multi-year memory pricing is the one mechanism that would convert an open-ended margin risk into a bounded one, and Apple has the scale to pursue it. Not answering is consistent with active negotiation and equally consistent with not having the leverage the prior exchange suggested it lacks. On pricing philosophy, refusing to commit to either dollars or percentage is the correct posture for a company mid-negotiation and useless for modeling: it means the FY27 margin depends on a business judgment not yet made.

The Services Deceleration Walk

Questioning on Services was pointed, noting the quarter came in below guidance, calculating that the September framework implies below 10% reported growth, and asking directly whether AI is diverting user time away from the App Store. The response walked the sequential bridge in detail and addressed every named factor except the AI one.

Q: "12% services growth was just a bit below your guidance. I imagine FX maybe played a role there. I think the fiscal 4Q guidance assumes another deceleration. I think the math would imply below 10% year-over-year as reported. Can you maybe just help us understand the kind of function factors underlying that deceleration? If that's App Store, which I think some third-party data sources would suggest, is that a function of AI maybe reprioritizing time away from parts of the App Store?"
— Erik Woodring, Morgan Stanley

A: "Foreign exchange was the main driver for the change in the year-over-year growth rate sequentially… We also had some factors that impacted the performance of the App Store. We did see some headwinds in mobile gaming. Keep in mind, we also made some changes to the App Store business model in certain countries. In the U.S., we do continue to operate under a court ruling impacting the link-out transactions."
— Kevan Parekh, CFO

Assessment: The sub-10% calculation went uncorrected, which effectively confirms it. More telling is the omission: the question offered a specific hypothesis, that AI assistants are absorbing engagement that previously flowed through the App Store, and the answer routed around it to FX, a film comp, gaming, regulation and litigation. That may simply be because management does not have the data. But it is the single most consequential structural question facing the Services line, and Apple did not engage with it on a call where it decelerated 420 basis points.

Whether Siri AI Changes Apple's Capital Intensity

A question on whether Siri AI shifts Apple's capital intensity, given the on-device and distributed-compute architecture, drew a hybrid-model description with no directional commitment and an acknowledgement that AI costs appear in cost of sales as well as operating expenses.

Q: "Do you expect that Siri AI would change the capital intensity of Apple, despite the fact that you have the ability to do so much differentiated workloads on device, and you have this distributed compute. You do have some requests that go into the back end, both in your own first-party cloud as well as third party. Is it right to think that the capital intensity of Apple will change in the future because of Siri AI?"
— Wamsi Mohan, Bank of America

A: "We use a hybrid model… We use some third-party cloud, and we do our own data centers. There will be a mix… There are other locations on the P&L other than OpEx, like COGS etc., that also have AI expenditures. We'll see what Siri AI does from the cost side of it. There's also the ability when people use it a lot for them to move up on an iCloud Plan as well. What the balance of that is a bit uncertain at the moment."
— Tim Cook, CEO

Assessment: "A bit uncertain at the moment" is the honest answer and an uncomfortable one four weeks before a fall launch. The filings show nine-month capex down 28.2% while R&D rose 32.5%, so today the strategy is unambiguously capital-light and free-cash-flow accretive, which is a real competitive advantage. The disclosure that AI costs also sit in cost of sales is the part worth carrying forward: if third-party cloud consumption scales with Siri AI usage, it lands in the same line as memory, at the same time, and the iCloud+ upsell floated as the offset has no disclosed pricing, attach rate or economics.

The Gross-Margin Bridge, Quantified

The most useful exchange of the call asked management to walk the sequential margin decline explicitly and to separate FX from commodity costs, noting the pattern looked atypical against Apple's normal seasonality. The response produced the two-step, memory-attributed bridge that anchors this report's central judgment.

Q: "If I look at the gross margin, you delivered 48% in the quarter without the tariff refund benefit. You're guiding to like 46.5%. Any way to walk us through the sequential driver there and how much of that is FX impacting it versus maybe increased commodity costs, et cetera, that's driving that sequential moderation? It does seem a bit more atypical than your normal sort of years."
— Samik Chatterjee, JPMorgan

A: "You look at the drivers of that, more than 100% of that can be explained by the memory cost change that Tim outlined. While FX was a factor, really the main driver was really the memory cost impact… That 160 basis point change is also more than that is explained by the change in memory cost… Generally, the sequential change from the March quarter to the 46.5% you referenced is really driven by memory."
— Kevan Parekh, CFO

Assessment: Management answered fully and specifically, which is to its credit and is also why the stock fell. Attributing more than 100% of two consecutive sequential declines to a single purchased input removes every alternative explanation an investor might reach for: not mix, not FX, not seasonality, not tariffs. It also removes the argument that the pressure is transitory, because management separately said memory pricing keeps rising past the guide window and the offsets decay. This exchange, more than the guide itself, is what changes our rating.

The Incoming CEO's Forty Words

The one question directed at the incoming chief executive asked for a high-level view of whether the competitive landscape is changing, referencing speculation about AI-native devices from other technology companies. The response was brief and deferred entirely to the outgoing CEO's framing.

Q: "I'd love to get maybe a just a high-level take from you if you think that the competitive landscape here is changing, especially as you hear about potentially companies like OpenAI building an AI-enabled competitive device or SpaceX AI potentially having a phone that could bypass some of the typical carrier attacks. Just would love your high-level thoughts on how you see the competitive landscape evolving and Apple's position there."
— Wamsi Mohan, Bank of America

A: "Well, thank you for asking. I guess I would just reiterate what Tim said, there is so much opportunity for us with everything that's happening in this space. We're just really focused on our plans and very excited about it."
— John Ternus, CEO-designate

Assessment: Declining to make news four weeks before taking over is defensible and probably correct protocol. It is also the second consecutive quarter in which the incoming chief executive has appeared on the call and said nothing substantive about strategy, competition, AI or capital allocation. Last quarter we named this as something we would want to see resolved and flagged it as a condition for higher conviction. It has not been resolved, and the first opportunity to resolve it is now the December-quarter call, by which point the memory question will already have been answered by the numbers.

What They're NOT Saying

1. Any December-quarter framing: Asked twice, declined twice, on a call where management volunteered that both memory costs and supply constraints get worse after September. The December quarter is Apple's largest and the Street was already modeling 8% to 9%. Declining to say whether that is enough leaves the market extrapolating a deteriorating trend without an anchor.
2. A gross-margin floor beyond September: For the second consecutive quarter management gave a detailed sequential bridge and refused to bracket the period after it. Last quarter that refusal covered one unknown quarter. This quarter it covers the period in which the carry-in inventory offset expires, which management itself identified as the mechanism holding margin up.
3. Whether Apple will sign long-term memory supply agreements: Asked directly, and the half of the question addressing LTAs went entirely unanswered while the pricing half received a full response. This is the one lever that would convert open-ended memory exposure into a bounded, modelable cost, and Apple would not confirm it is even pursuing it.
4. The size of the supply constraint: Three quarters of citing constraints as a material revenue cap, and still no quantification of what they cost in dollars or units. This matters more now than before, because management's bridge back to the June growth rate depends on a constraint magnitude it will not disclose. The claim is unverifiable by construction.
5. Price elasticity on the iPad and Mac increases: Management said it is "too early to come to a definitive conclusion" and pointed to channel-inventory lags. The increases landed on the only declining category and on a category whose growth is partly institutional and budget-sensitive. The answer will exist inside the September quarter and management will not see it before the quarter is well underway.
6. Siri AI monetization economics: On compute cost, "I don't want to say that we have a complete plan for that." On the iCloud+ upsell offered as the offset, no pricing, no attach-rate expectation, no revenue framing. A flagship feature ships this fall with an undisclosed cost curve and an undisclosed revenue model.
7. A Siri AI timeline for Europe or China: "We're working closely with them" for the EU and "we're at the front end of that" for China. Those two regions produced 44% of revenue this quarter and both grew 22.4%. If Siri AI is the fall cycle's demand driver, no date exists for nearly half the business.
8. The $9.2B increase in net intangible assets: Intangible assets, net rose from $11,093M at fiscal year-end to $20,342M at June 27, an 83% increase, alongside a $13.5B increase in other non-current liabilities. Neither movement was mentioned in the prepared remarks or raised in Q&A. For a balance sheet this heavily scrutinized, a $9.2B swing in a single asset line passing without comment is an omission worth noting.

Market Reaction

• Pre-print setup: AAPL closed at $333.43 on July 30, 2026 entering the print, near the top of a 52-week closing range of $202.38 to $340.08. The setup was hot: +22.6% YTD (from $271.86 at 2025 year-end), +15.2% over the trailing 30 days (from $289.36), and +59.5% over the trailing twelve months (from $209.05). The S&P 500 was +8.7% YTD entering the print, so AAPL had outrun the index by roughly 14 percentage points on the year and by most of that margin in the final month alone.
• Reaction session (July 31, 2026): The stock gapped down 8.6% to open at $304.81, traded a range of $300.00 to $310.49, and closed at $308.91, down 7.4% (a decline of $24.52). Volume was 127.4M against a 60.4M 30-day average, 2.1x normal. The S&P 500 rose 0.7% on the session, so AAPL underperformed the tape by roughly 8.1 percentage points.
• Post-reaction position: Even after the decline, AAPL closed +13.6% YTD against the index at roughly +9.5%. The drop unwinds about half of the trailing-30-day run and leaves the stock above where it traded through most of the June quarter.

The magnitude is explained by the setup as much as by the print. A stock that has risen 15.2% in a month into a quarter carries no margin for a guide that lands below the Street on every line, and the reaction was consistent all day: the open at $304.81 was the low of the range, the stock recovered modestly through the session, and it never traded back toward the prior close. That shape, a gap that partially fills but does not reverse on 2.1x volume, reads as genuine repositioning rather than a liquidity air pocket.

Three specific elements did the damage, in descending order. First, the margin bridge: management's own arithmetic put adjusted gross margin on a 280 basis point slide across two quarters and attributed more than all of it to a purchased component whose price it said keeps rising. Second, the guide: 9% to 11% against a roughly 12% consensus, with December declined. Third, the composition of the beat: once the tariff refund is stripped, a quarter that looked like a thirteen-cent beat with a 220 basis point margin surprise was a two-cent beat with a 20 basis point margin surprise, and Services actually missed.

The reaction is proportionate rather than an overreaction, which is the reason we are not treating the decline as an entry point. At $308.91 the stock trades near 35x our FY26 EPS estimate of roughly $8.86, above the roughly 31.6x it carried at the close following last quarter's print. The multiple expanded during a quarter in which the forward margin outlook deteriorated. The sell-off corrects the price move, not the risk.

Street Perspective

Debate: Is the memory squeeze a two-quarter dip or a structural reset of Apple's cost base?

Bull view: The bull case holds that memory is the most reliably cyclical input in electronics, that current pricing reflects an AI-infrastructure demand spike that will provoke a supply response, and that Apple has more tools than anyone to bridge it: pre-bought inventory, the largest purchasing scale in consumer hardware, product-level pricing power it has now demonstrated it will use, and a 75%-margin Services mix. On this reading the trough is a two-to-three-quarter margin dip and buying a franchise like this into a component cycle is a gift.

Bear view: The bear camp points to the structure management itself described: three DRAM suppliers, a demand pull from AI buyers with far less price sensitivity than a consumer hardware company, and a chief executive conceding that more suppliers would help on supply but that the pricing benefit is "unclear." Add that the disclosed offsets are stated to decay after September and that management would not confirm it is even pursuing long-term supply agreements, and the case is that this resets Apple's cost base rather than dipping it.

Our take: The bears have the better of it on the next four quarters and the bulls have the better of it on the next four years. Memory cycles do turn, but the specific things management said on this call, that pricing keeps rising past the guide window, that the inventory buffer depletes, and that supplier concentration limits leverage, all point to pressure persisting through FY27 rather than resolving inside it. We are not modeling a permanent reset. We are modeling a longer trough than the bull case assumes, and at 35x that difference matters more than it would at 25x.

Debate: Does a shipping cap make the guide conservative, or is the demand claim unverifiable?

Bull view: The optimistic reading takes management at its word that the constraint is a demand-forecast miss rather than a supply failure. If iPhone and Mac demand genuinely exceeds plan, the September guide is set against what Apple can ship rather than what customers want, the unmet demand carries into December and March, and the deceleration is an accounting artifact of timing. On this view the guide is beatable and the December quarter benefits from a back-order book.

Bear view: The skeptical reading notes that Apple has cited material supply constraints for three consecutive quarters and has never once quantified them, and that the bridge back to the June growth rate depends entirely on a magnitude the company will not disclose. Constraints are also a convenient explanation: they attribute a deceleration to strength rather than weakness while being impossible to check from outside. And the constraint now spans three categories rather than one, which is a widening, not a narrowing.

Our take: We lean bull here, with a caveat about where the benefit lands. The corroborating evidence is unusually good: Mac beat the Street by roughly 12% while constrained, iPhone printed 21.7% while constrained, every geography set a record, and the balance sheet shows a $5.4B inventory build consistent with a company scrambling for components. That is not the pattern of a demand shortfall dressed up. But a constrained quarter is the worst setting for gross margin, because the cost inflation arrives without the volume leverage to absorb it. We take the demand claim at close to face value in the revenue line and give it no credit in the margin line.

Debate: Does 35x survive a decelerating growth rate and a leadership handoff?

Bull view: Bulls argue the multiple is defensible because the quality of the franchise improved this quarter, not deteriorated: the installed base passed 2.5 billion devices, paid subscriptions passed 1.5 billion, nine-month free cash flow grew 52.5% while capex fell 28.2%, and Apple is running an AI strategy that requires almost no incremental capital while peers commit tens of billions. Post-print price targets clustering well above the current price reflect a view that a component cycle should not de-rate a compounding franchise.

Bear view: The bear case is arithmetic. Revenue growth is guided from 16.4% to roughly 10%, Services from 12.1% to roughly 9.5%, and adjusted gross margin from 49.3% to 46.5%, all in the span of two quarters, while the stock trades at a higher forward multiple than it did before the quarter. A leadership transition executes on September 1 into the hardest operating environment in years, and management has withdrawn forward visibility precisely where it is most needed. At 35x, none of that is priced.

Our take: The bear arithmetic is the stronger argument at this price, and the dispersion in post-print Street targets, which spans a range wide enough to imply both meaningful upside and meaningful downside from here, tells you the market agrees the outcome is genuinely uncertain. We do not think the franchise deserves a de-rating to a market multiple; the free-cash-flow generation and the capital-light AI position are real and undervalued in the bear case. We do think 35x is the wrong price for a business whose growth rate and gross margin are both compressing on management's own guidance, with the December quarter unframed. That gap between a great business and a full price is what a Hold is for.

Model Update Needed

Valuation impact. At the reaction-day close of $308.91, AAPL trades at roughly 35x our FY26 EPS estimate of about $8.86, against roughly 31.6x at the equivalent point after last quarter's print. The multiple expanded through a quarter in which forward gross margin, forward revenue growth and forward Services growth all deteriorated. Rolling forward, FY27 carries the memory headwind into a full year, laps the hardest iPhone comp of the cycle, and absorbs an operating expense base growing faster than revenue; we frame FY27 EPS in the $9.25 to $9.75 range, which puts the stock near 32x to 33x on next year. Applying a 30x to 32x multiple to that path frames a 12-month target of roughly $285 to $305, or about 8% below to 1% below the current price. That is an in-line outcome, not a negative one, and it is why the rating moves to Hold rather than further. We would revisit conviction higher on any of three developments: a long-term memory supply agreement that bounds the cost exposure, evidence in the December print that Services has stabilized above 10%, or a December guide that shows the supply constraint converting to revenue rather than deferring further.

Thesis Scorecard Post-Earnings

Overall: The thesis weakened, and it weakened on the two pillars that were carrying the rating rather than on the one everyone watches. Our Outperform, held since the October upgrade, rested on three conditions: the iPhone catalyst confirming, margins holding, and Services compounding mid-teens. Only the first survived this quarter, and it survived well. Margins did not hold: management's own bridge slides adjusted gross margin 280 basis points across two quarters and assigns more than the whole decline to a component whose price it says keeps rising after the guide window, with the inventory buffer that is cushioning today explicitly running out. Services did not compound mid-teens: it decelerated 420 basis points to 12.1%, missed consensus, and is guided to roughly 9.5%, with three of the five causes management named being structural rather than cyclical. What did not break is the demand story, and that is not a small thing: iPhone ran 21.7% and Mac 28.7% while supply-capped, every geography set a record, the installed base passed 2.5 billion devices, and free cash flow grew 52.5% on nine-month capex that fell 28%. This is a franchise operating extremely well with a cost problem and a price problem.

Action: Downgrade to Hold from Outperform. The rating change is a margin-and-multiple call, not a demand call, and the distinction matters for how to hold the position. We would not sell this franchise on a component cycle: the demand signal is the strongest in years, the capital-light AI posture is genuinely differentiated against peers committing tens of billions to infrastructure, and the buyback plus a $147B cash position underwrite the downside. But at roughly 35x FY26 earnings, a higher multiple than the stock carried before this quarter, an investor is paying a premium price for a business whose revenue growth is guided from 16.4% to roughly 10%, whose highest-margin segment is decelerating below 10%, and whose gross margin is compressing with no floor disclosed, four weeks before a leadership handoff. We move to the sidelines and would look to re-engage on a long-term memory supply agreement that bounds the cost exposure, evidence that Services has stabilized above 10%, or a December guide showing the constrained demand converting rather than deferring again. We expect AAPL to perform roughly in line with the S&P 500 over the next twelve months.