Apple Inc.: Strip Out the Tariff Refunds and the 50.1% Margin Becomes 48.1%, While 308.91 Dollars Earns Less Than the Ten-Year Treasury

Services illustrates why “record” and “miss” can both be true. Revenue reached a June-quarter record of USD 30.739 billion and grew 12.1%, but consensus was approximately USD 31.22 billion, leaving a shortfall of roughly USD 481 million, or 1.5%. Apple’s filing says advertising and cloud services were the main growth contributors for the quarter; management also reported records in cloud and payment services. Services retained a 75.6% gross margin against 40.1% for products, which is why a modest Services miss matters disproportionately to group valuation.

The prior report identified three uncertainties: control of the AI entry point, European revenue-sharing regulation, and Greater China. None has disappeared, but their directions have diverged.

Apple’s AI position has improved from promise to testable product. WWDC26 introduced a rebuilt Siri with personal context, on-screen awareness, application actions, and broader knowledge functions. Developer testing began in June; broader user beta availability was scheduled for later in 2026. As of August 1, then, the rebuilt Siri had not shipped broadly to ordinary users. Apple retains the operating-system interface, device context, permissions, application action layer, on-device inference, and Private Cloud Compute architecture. Its third-generation foundation-model family was nevertheless developed in collaboration with Google, and Apple has a multiyear arrangement involving Gemini. Neither company has confirmed the reported payment of about USD 1 billion annually. The evidence supports a hybrid conclusion: Apple owns the interface and orchestration layer, but it does not independently own every frontier-model capability underneath it.

The US search-antitrust outcome has become less damaging in the near term. The final judgment did not prohibit Google from paying Apple for default placement. It did prohibit exclusivity, restrictions on distributing rival search or generative-AI services, and arrangements that condition payments on maintaining placement for more than one year. Appeals remained active in July 2026, including an attempt by the US government and states to secure stronger relief. Apple has preserved the revenue stream for now, but the contract has become more contestable and less durable.

Europe moved in the less favorable direction. The European Commission fined Apple EUR 500 million in April 2025 over anti-steering restrictions. Apple subsequently introduced EU fee schedules that permit external destinations but preserve acquisition, store-service, and commerce charges under various combinations. In July 2026, the EU General Court dismissed Apple’s challenges to the gatekeeper designations covering iOS and the App Store. The economics have not collapsed, but Apple’s legal freedom to preserve the old App Store structure has narrowed.

Greater China improved most clearly. Fiscal Q3 revenue rose 22% to USD 18.816 billion, while nine-month revenue increased 30% to USD 64.839 billion, helped by iPhone demand and favorable renminbi translation. IDC estimated Apple’s mainland smartphone shipments rose 24.4% in calendar Q2 2026, lifting share to 18.1% from 13.9%. Huawei remained ahead at 22.6%, and Apple’s 2024 shipments had fallen 17%, so the current rebound is a product-cycle and share-recovery fact rather than proof that the structural contest has ended.

Apple’s July 31 close was USD 8.68, or 2.9%, above the prior report’s USD 300.23 anchor. The price had risen despite the post-earnings drop. Fiscal 2025 diluted EPS was USD 7.46; replacing the first nine months of fiscal 2025 with the first nine months of fiscal 2026 produces trailing EPS of about USD 8.72. Removing the latest USD 0.11 refund gives trailing underlying EPS of about USD 8.61. The resulting multiple is 35.9 times and the earnings yield is 2.79%.

The 10-year US Treasury yielded about 4.74% on July 31. That put Apple’s underlying earnings yield around 1.95 percentage points below the government-bond yield. The previous report’s comparison was 2.75% against 4.59%, a negative gap of 1.84 points. The yield disadvantage has widened slightly rather than closed, even though earnings grew.

Apple’s qualitative portrait is a mature cash-generating compounder in strategic transition. It is a premium consumer platform, not a conventional high-growth AI-infrastructure company. Current growth rests on a strong device cycle, expanding Services, buybacks, and the possibility that AI shortens replacement cycles. Its capital-light position lets it return cash while Microsoft, Alphabet, Amazon, Meta, and others commit unprecedented sums to computing infrastructure. That creates an advantage if models commoditize and an important deficit if economic control accrues to model and cloud owners.

Since May the evidence has improved enough to support existing ownership more comfortably: revenue growth accelerated, China recovered, the Google payment survived the first remedies decision, and a working Siri architecture entered developer testing. The purchase case remains weak at the current price because the stock still requires a premium multiple, sustained Services economics, successful AI execution, and no lasting margin damage from component scarcity. The final rating and independently constructed valuation bands appear in the cross-synthesis section.

Vertical history and financial evolution

Apple began in Los Altos, California, on April 1, 1976, when Steve Wozniak’s ability to design unusually economical personal computers met Steve Jobs’s product and commercial instincts. Ronald Wayne helped establish the original partnership but exited almost immediately. Mike Markkula’s early capital and business discipline helped transform a hobbyist-board venture into a company capable of producing and distributing the Apple II. Sales rose from USD 7.8 million in 1978 to USD 117 million in 1980, reflecting the spread of personal computing beyond clubs and laboratories.

Apple went public on December 12, 1980, at USD 22 per share, equivalent to USD 0.10 after subsequent splits. It sold approximately 4.6 million shares, raised a little over USD 100 million, and ended its first trading day with a market value around USD 1.8 billion. Investors initially understood Apple as a rapidly scaling personal-computer manufacturer distinguished by approachable design and integrated hardware and software.

The company’s history is best understood in six stages.

The first stage, from 1976 through the early 1980s, validated the personal-computer category. Apple II was expandable, comparatively accessible, and supported by applications such as VisiCalc. The key decision was to sell a usable computing system rather than a bare technical component. Commodore, Tandy, Atari, and emerging IBM-compatible vendors competed for the same new customers. Apple built brand recognition and an integrated-product instinct, but it did not yet possess a durable industry standard.

The second stage, from Macintosh through the 1990s crisis, established both Apple’s design advantage and the economic weakness of remaining a minority computing architecture. Macintosh popularized the graphical interface and mouse for a mass audience, but Microsoft and the PC ecosystem spread software development and hardware production across many vendors. Jobs’s departure in 1985 removed the company’s most forceful product editor. A broad and confusing product portfolio, weak operating discipline, and the scale advantage of Windows-based PCs pushed Apple toward distress. The lesson stuck: superior individual products were insufficient without a platform, developer base, and controlled customer experience.

The third stage began with the NeXT acquisition and Jobs’s return in 1997. Apple reduced its product grid, secured liquidity, adopted NeXT-derived software as the foundation for Mac OS X, and rebuilt around a small number of recognizable devices. The iMac restored relevance; the company’s retail stores gave it direct control over presentation and service; the iPod and iTunes extended Apple from computers into consumer media. The strategic turn was vertical integration around the user experience rather than a renewed attempt to win the commodity PC-volume contest.

The fourth stage began with the iPhone in 2007 and the App Store in 2008. Apple changed from a computer and music-device company into the controller of a mobile-computing platform. The iPhone combined premium hardware margins with an operating system, identity, payments, distribution, and a developer ecosystem. Each layer increased the value of the others. Apple’s 2007 removal of “Computer” from its legal name accurately described the change. The lasting asset was no single iPhone generation; it was the recurring ability to place a proprietary software and commerce layer between users and digital suppliers.

The fifth stage, spanning roughly 2012 to 2020, converted iPhone scale into a broader ecosystem and capital-return machine. Tim Cook’s operational discipline reduced inventory risk and secured component capacity. Apple Watch, AirPods, payments, cloud storage, subscriptions, and an expanding App Store increased revenue per device and made switching more inconvenient. The dividend resumed in 2012 and repurchases became a central use of capital. Five historical stock splits, including seven-for-one in 2014 and four-for-one in 2020, increased nominal share accessibility but did not alter value.

The sixth stage, beginning with Apple silicon and extending through the present AI transition, brought core computing technology back inside the company. Apple’s custom chips improved performance per watt and reduced reliance on Intel’s roadmap. The company simultaneously increased Services revenue from USD 85.2 billion in fiscal 2023 to USD 109.2 billion in fiscal 2025; Services gross margin rose from 70.8% to 75.4%. The unresolved part of this stage is whether Apple can repeat its hardware and silicon integration advantage in generative AI or must rent the most important intelligence layer from Google and other model suppliers.

Several key nodes still determine today’s investment case. The iPhone created the installed base and remains about half of quarterly sales. The App Store turned installed devices into recurring distribution revenue, but also created Apple’s most concentrated regulatory exposure. The shift to Apple silicon proved the company could internalize a foundational technology and improve the whole product portfolio. The 2012 capital-return program converted excess cash into per-share earnings growth. Then came the 2024–2026 Siri delays, which exposed a different execution record in modern AI: Apple retained control of the device and interface but fell behind frontier-model developers.

Management succession now forms another important node. Tim Cook remains CEO as of the report date but is scheduled to become executive chairman on September 1, 2026. John Ternus, a 25-year Apple veteran who led hardware engineering, will become CEO and join the board. Johny Srouji has already taken an expanded chief hardware officer role; Kevan Parekh has served as CFO since January 2025. The transition preserves institutional continuity, but it moves product and AI execution under a first-time public-company CEO at a moment when Apple must define its role in a new computing interface.

Apple’s vertically proven capability is the integration, industrialization, distribution, and monetization of selected technologies at enormous consumer scale, not invention in every technology category. That distinction explains both its successes and the current AI debate. Apple can arrive after pioneers and still capture a large profit pool when it controls the final product. The strategy becomes less reliable when the upstream technology itself acquires direct user relationships.

The financial record shows the transformation from hardware concentration to a mixed device-and-services model.

The five-year figures come from Apple’s annual filings; free cash flow is operating cash flow less purchases of property, plant, and equipment.

Revenue did not compound smoothly over this period. Pandemic-era device demand and work-from-home purchases boosted fiscal 2021 and 2022; currency pressure and electronics normalization then reduced fiscal 2023 sales. Fiscal 2024 growth was slight. Fiscal 2025 reaccelerated to 6.4%. The more durable change was mix: Services revenue expanded faster than products and lifted consolidated gross margin by 5.1 percentage points from fiscal 2021 to fiscal 2025.

Cash conversion has been strong over a full cycle. Aggregate operating cash flow for fiscal 2021–2025 was approximately USD 566.5 billion against USD 497.2 billion of net income, a ratio of 1.14. The annual ratio exceeded one in four of the five years and dipped just below one in fiscal 2025 because working-capital movements reduced cash generation. This is consistent with a business that receives consumer and channel cash rapidly, carries limited manufacturing assets, and expenses most research rather than capitalizing it.

Apple’s capital expenditure averaged about USD 11 billion annually over the five years, or only around 2.7% of fiscal 2025 revenue. The company relies on suppliers and contract manufacturers for much of the physical production base. Its most consequential investment appears instead in research and development expense, component prepayments, tooling arrangements, and long-term supply commitments. Fiscal 2025 R&D reached about USD 34.6 billion; fiscal Q3 2026 R&D rose 32% year over year to USD 11.7 billion.

The June 2026 balance sheet contained USD 39.5 billion of cash and USD 107.0 billion of current and noncurrent marketable securities. Commercial paper and term debt totaled about USD 84.3 billion, leaving approximately USD 62 billion of net cash and securities before operating-liquidity adjustments. Inventory rose to USD 11.1 billion from USD 5.7 billion at fiscal 2025 year-end, driven primarily by components, which rose from USD 2.1 billion to USD 7.6 billion. That build may partly protect the September product cycle from shortages, but it also shows that procurement conditions have changed materially.

Apple spent USD 62.1 billion on repurchases and USD 11.8 billion on dividends during the first nine months of fiscal 2026. Diluted weighted-average shares fell about 2% year over year in fiscal Q3. Repurchases have added meaningfully to per-share growth, though buying shares at a multiple in the mid-30s earns a lower prospective return than repurchasing them at the teens or low 20s.

ROE is an imperfect measure for Apple because decades of repurchases compressed book equity and at times produced negative retained earnings. The underlying economic return remains exceptional: Apple generates more than USD 100 billion of annual profit with comparatively little net property and equipment and limited working capital. The source is structural rather than cyclical alone (brand pricing, supplier financing, software economics, and customer prepayment), but the level can be overstated when regulatory rents and low capital needs are assumed permanent.

The stock’s valuation history reflects a change in how the market classifies the business. During the late-2018 selloff, Apple traded near 13 times earnings as investors treated it as a mature handset manufacturer exposed to replacement cycles and China. Services growth, recurring revenue, wearables, buybacks, and very low rates later moved the valuation center toward that of a consumer platform. The 10-year average P/E is approximately 25–26 times, while the current underlying multiple is about 35.9 times, roughly two-fifths above the historical median or average depending on the data series.

The market’s current valuation label is “capital-light platform and AI option,” not “cyclical hardware vendor.” The multiple embeds both proven Services economics and future benefits that have not yet appeared in Apple’s reported AI revenue. The re-rating is partly justified by business quality; its upper end also reflects market preference and scarcity value for a cash-generating mega-cap that is not spending tens of billions each quarter on data centers.

Business model, moat, industry, and peers

Apple earns money through two economically different engines. Products acquire and refresh the installed base: iPhone, Mac, iPad, wearables, home products, and accessories. Services monetize access, identity, payments, storage, applications, entertainment, advertising, warranties, and licensing across that installed base. Apple does not disclose operating profit by product, but it discloses product and Services gross margins, which reveal where the profit pool is concentrated.

Apple’s filing supplies revenue and segment gross margins. The reported consolidated margin includes the tariff-refund benefit discussed earlier.

At disclosed margins, products generated approximately USD 31.5 billion of fiscal Q3 gross profit and Services approximately USD 23.2 billion. Services contributed about 42% of segment gross profit on 28% of revenue. This estimate does not allocate corporate costs or account for different operating expenses, but it explains why a USD 481 million Services miss matters more than the same miss in lower-margin hardware.

iPhone remains both the acquisition engine and concentration risk. It represented just under half of quarterly revenue, and many Services purchases originate from iPhone users. Mac and iPad broaden the installed base but remain replacement-driven categories. Wearables increase ecosystem attachment. Services raises lifetime revenue and margin without requiring Apple to manufacture another physical device for every additional transaction.

Apple’s cost structure contains a relatively flexible hardware component and a less flexible innovation component. Manufacturing, freight, components, warranty expense, and channel payments vary with unit sales. R&D, software development, operating-system support, retail infrastructure, content commitments, and regulatory compliance are harder to reduce without damaging future competitiveness. The 32% increase in quarterly R&D and 23% rise in total operating expenses show that Apple is currently investing through the AI transition rather than harvesting the installed base with a static cost base.

Operating leverage depends on mix. Services growth and premium-device mix can raise group margins even at moderate unit growth. Component shortages, memory inflation, foreign exchange, lower App Store commissions, or a reduction in search licensing can reverse the effect. The tariff-adjusted fiscal Q3 margin of about 48.1% illustrates the underlying model more accurately than the reported 50.1%.

Apple’s strongest moat is the combination of installed base, operating-system control, trusted identity, distribution, and custom silicon, not brand in isolation. More than 2.5 billion active devices create a large audience for developers and service providers. Users accumulate paid applications, photographs, messages, health data, accessories, subscriptions, and device familiarity. Each element makes switching incrementally inconvenient, and Apple’s control of hardware and software lets it coordinate security, performance, battery life, and support.

The second durable moat is product-level pricing power. Apple can charge premium prices because customers buy an integrated experience, resale value, security support, retail service, and ecosystem compatibility rather than an isolated list of specifications. Recent Mac and iPad price increases have not prevented strong Mac growth, although the willingness to absorb higher prices should not be extrapolated without limit.

The third is supply-chain scale and silicon control. Apple can reserve advanced-node capacity, design chips around its software needs, and spread development costs across hundreds of millions of devices. The current constraints reveal the boundary of this moat: scale improves priority and bargaining power but cannot create advanced-node or memory capacity quickly when an industry-wide AI buildout absorbs supply.

The fourth is distribution control. The App Store, default settings, payment credentials, and operating-system APIs give Apple access to economic activity generated by third parties. This is highly profitable but is the moat most vulnerable to law rather than ordinary competition. US search remedies, the EU Digital Markets Act, UK proposals, and other proceedings seek to make distribution more contestable.

Brand without ecosystem attachment would be a marketing moat. The actual moat is stronger because customers repeatedly reveal willingness to upgrade, attach accessories, subscribe, and remain inside the system. The risk is that AI agents become a new cross-platform layer through which users discover, communicate, shop, and complete tasks. If that occurs outside Apple’s interface, some switching costs and distribution power migrate upward to the agent provider.

Apple’s consumer-electronics markets are mature. Smartphone demand depends largely on replacement, premiumization, camera and processor improvements, carrier economics, financing, and regional share shifts. PCs and tablets combine replacement cycles with education and enterprise demand. Wearables retain category-growth potential but are no longer at their first-adoption stage. Services can grow faster through rising installed devices, price increases, subscriptions, cloud storage, payments, advertising, and more transactions per user.

Apple participates in several cycles at once: the consumer cycle, device-replacement cycle, semiconductor-capacity cycle, technology-iteration cycle, foreign-exchange cycle, and regulatory cycle. It crosses ordinary consumer downturns better than mass-market vendors because its customers are wealthier and its balance sheet supports marketing and financing. Its largest fragility is a synchronized shock in which higher component costs, weaker discretionary demand, China share loss, and a lower valuation multiple occur together.

The peer comparison is less about which company makes the closest phone. Apple’s economic competitors occupy different control points.

Microsoft became the enterprise software, cloud, and AI-infrastructure provider. Customers choose it for corporate identity, productivity software, Azure integration, and distribution through existing enterprise relationships. Alphabet became the search, advertising, Android, cloud, and frontier-model company; its economic relationship with Apple combines competition, supply, and a multibillion-dollar distribution payment. Meta owns consumer attention and advertising across social applications, while investing heavily to control models and devices. Amazon owns commerce and a major cloud platform. NVIDIA captures the scarcity value of accelerated computing. Samsung and Huawei compete more directly in devices, while TSMC, Broadcom, Qualcomm, memory suppliers, and assemblers enable Apple’s hardware economics.

† Apple P/E uses trailing EPS adjusted for the USD 0.11 tariff-refund benefit. Alphabet’s multiple is unadjusted, and its June-quarter net income of USD 112.193 billion included USD 97.983 billion of non-operating other income, so the trailing comparison overstates Apple’s premium to Alphabet. ‡ Apple figure covers the first nine months of fiscal 2026, while Microsoft and Alphabet figures are single quarters; the comparison is intended to show scale, not matched-period accounting. § Microsoft figure refers to fiscal Q3 2026 because its subsequent release was published immediately before the research date and detailed comparisons were still being processed. Market data and company disclosures support the table.

The capex contrast is extraordinary. Apple spent USD 6.8 billion on property and equipment over nine months, while Microsoft spent USD 41 billion and Alphabet USD 44.9 billion in their latest quarters. Microsoft said about two-thirds of its latest capex involved shorter-lived CPUs and GPUs; Alphabet said most spending supported AI technical infrastructure and that Q2 free cash flow was negative USD 5.9 billion. Apple generated USD 117.0 billion of nine-month operating cash flow and returned roughly USD 74 billion through buybacks and dividends.

This capital-light position is both advantage and strategic deficit. It is an advantage if foundation models become competitive commodities: Apple can purchase intelligence, preserve the consumer interface, avoid depreciation risk, and continue buying back shares. It is a deficit if proprietary models, data centers, and agent ecosystems create durable control over user intent. In that case, Apple would be paying rent to suppliers whose products increasingly mediate the relationship between Apple and its customers.

Apple trades at a higher P/E than Microsoft, Alphabet, and Meta despite having less direct AI-infrastructure revenue. The premium reflects lower capital intensity, consumer loyalty, buybacks, and the perceived stability of its installed base. It also assumes Apple can remain the owner of the user interface while sourcing parts of the underlying model. That assumption is more demanding than a simple comparison of near-term earnings growth suggests.

Alphabet is the most important horizontal comparison because it attacks and supports Apple simultaneously. Google supplies search payments, a major portion of the new model stack, Android competition, and cloud intelligence. Microsoft has stronger enterprise AI monetization but less direct leverage over Apple’s consumer interface. Meta can reduce the time users spend in Apple-controlled discovery channels. Amazon and NVIDIA benefit from AI spending that Apple largely avoids. Samsung and Huawei can pressure device share, but neither currently duplicates the entire global combination of premium hardware, Services economics, custom silicon, and retail distribution.

Apple’s ecological niche is the premium consumer endpoint and operating-system gatekeeper. It takes profit from handset vendors, component suppliers, developers, media providers, payment intermediaries, advertisers, search engines, and carriers by controlling the customer’s device and access layer. Model providers are the parties most likely to take some of that pool. Regulation can redistribute another part to developers and alternative payment channels.

Current fundamentals and the three unresolved questions

Apple’s last four reported quarters show a sharp improvement from the sluggish growth of fiscal 2023–2024. Fiscal Q4 2025 revenue grew 8% to USD 102.5 billion. Fiscal Q1 2026 revenue rose 16% to USD 143.8 billion, with an installed base above 2.5 billion active devices. Fiscal Q2 produced USD 111.2 billion and double-digit geographic growth. Fiscal Q3 returned to 16% growth. The pattern supports a strong iPhone 17 and Mac product cycle rather than a one-quarter pull-forward alone, although September guidance indicates the rate will slow.

Fiscal Q3 product performance was uneven beneath the total. iPhone and Mac provided nearly all of the acceleration. iPad declined 6%, while Wearables, Home and Accessories grew 6%. Americas revenue rose 11%, Europe 22%, Greater China 22%, Japan 13%, and the rest of Asia-Pacific 16%. The strength was broad by geography but concentrated by product.

Product gross margin increased to a reported 40.1% from 34.5% a year earlier, reflecting mix and tariff refunds, partly offset by memory costs. Services gross margin held at 75.6%. R&D increased faster than revenue, and inventory nearly doubled from the fiscal year-end as Apple accumulated components. The current income statement thus combines better mix and demand with a more difficult procurement environment.

The market reacted to a forward margin-and-availability problem, not merely to a backward-looking Services miss. Management expects September-quarter revenue growth of 9–11%, an FX headwind of approximately 2.5 points, iPhone growth in the mid-teens, Services growth broadly similar to June after adjusting for currency, and gross margin of 47–48%. That margin guide itself includes about one percentage point of expected tariff-refund benefit, implying an underlying range closer to 46–47%.

Supply constraints are rooted principally in advanced semiconductor-node availability for Apple system-on-chips and rising memory scarcity, rather than a disclosed shortage of one finished product. Cook said demand had exceeded Apple’s expectations and that advanced-node supply had less flexibility to respond. The AI data-center buildout is also competing for memory, wafers, packaging resources, capital equipment, and supplier attention. Apple expects the constraints to affect iPhone, Mac, and iPad more significantly in the September quarter.

The constraint is partly a demand-strength signal. Mac grew 29% despite price increases, and iPhone grew 22%. It is also a cost and lost-revenue problem. If Apple cannot obtain enough advanced chips, units cannot ship; if it secures supply at higher prices, gross margin falls unless prices or product mix compensate. The component-inventory increase provides some buffer, but it cannot fully solve shortages extending into a major product ramp.

The market’s current narrative has shifted. Before the results, investors were rewarding Apple for escaping the capital intensity affecting other AI names, a strong device cycle, China improvement, and the prospect of a successful Siri release. Apple briefly exceeded USD 5 trillion in market value in late July. The earnings release reminded investors that the AI infrastructure boom can still hurt Apple indirectly through scarce components and that capital-light participation provides less direct AI revenue.

The most important bull-bear disagreement concerns whether Apple’s current device growth is the beginning of an AI-supported upgrade cycle or the peak of an unusually favorable product and pricing period. Bulls point to double-digit revenue growth, record iPhone and Mac quarters, China recovery, a 2.5-billion-device base, and a Siri architecture that can distribute AI at global consumer scale. Bears point to a Services miss, worsening supply, refund-assisted margins, a high multiple, and dependence on Google for both search income and parts of the AI stack.

Services deserves separate treatment because it connects the moat, regulation, and valuation. Trailing-12-month Services revenue is approximately USD 120.5 billion, calculated by replacing the first nine months of fiscal 2025 with the USD 91.7 billion reported for the first nine months of fiscal 2026. Fiscal Q3 growth was driven primarily by advertising and cloud services, while the call also identified records in cloud and payment services. Apple does not disclose App Store, search licensing, cloud, payments, media, or warranty revenue separately, limiting precise assessment of organic composition.

The Google default-search payment is commonly estimated at about USD 20 billion annually. Neither Apple nor Alphabet confirms the exact amount, so the figure should be treated as an external estimate. At USD 20 billion, it would equal about 16.6% of trailing Services revenue and roughly 4.3% of trailing group revenue. Analysts previously estimated that losing it could reduce Apple profit by 4–6%, although a full model depends on tax, replacement monetization, and Apple’s negotiating response.

The following sensitivity assumes USD 20 billion of annual search-payment revenue, a 90% incremental operating margin because the payment requires little associated product cost, a 15% effective cash tax rate, and approximately 14.7 billion diluted shares. These are research assumptions, not disclosed Apple economics.

A full elimination would likely turn present Services growth into contraction unless other components accelerated substantially. The estimated USD 18 billion operating-income reduction would equal roughly one-tenth to one-eighth of Apple’s current annual operating-income scale. The effect could be smaller if Apple replaced Google with an auction, its own advertising, a paid search arrangement, or an AI partner; it could be larger if losing the payment also weakened user or developer economics.

The current legal position is more nuanced than “Google payments survive.” The December 2025 final judgment permits payments, but makes default arrangements nonexclusive and limits contractual mechanisms that lock placement beyond one year. Google must also refrain from restricting partners’ distribution of competing search, browser, or generative-AI products. Appeals filed by both sides leave the final remedy open. Apple keeps the cash flow in the immediate period, while its bargaining position and revenue visibility are structurally weaker than under a durable exclusive arrangement.

The first prior uncertainty, AI entry-point ownership, has improved but remains open. Apple controls Siri’s invocation, permissions, personal data context, app actions, on-device models, and Private Cloud Compute. The new foundation models include five model classes and were developed with Google collaboration. Apple’s developer framework can also use cloud models from providers such as Gemini and Claude. The architecture lets Apple choose suppliers and keep the interface. It also confirms that the company views external models as necessary components rather than peripheral add-ons.

The commercial terms are insufficiently disclosed. Reuters reported a multiyear Google arrangement; a separate report cited an approximately USD 1 billion annual model-access payment, but neither company has confirmed that price, the revenue-sharing terms, inference commitments, or data-center obligations. Apple may be obtaining favorable economics because it supplies distribution to an enormous user base. The opposite risk is that Google gains model leverage over the world’s most valuable consumer interface.

The decisive evidence has not yet arrived because the rebuilt Siri is not broadly available. Developer testing can establish technical integration, but investment relevance depends on user adoption, reliability, task completion, latency, geographic availability, developer support, and whether AI changes device replacement behavior or Services spending. Until these are measured, Apple is positioned as both the owner of the AI entry point and a distribution channel for suppliers beneath it.

The second uncertainty, EU antitrust and revenue sharing, has resolved partly unfavorably. Apple’s current EU terms permit developers to direct users to websites, alternative marketplaces, or other applications. Depending on the selected framework and service tier, fees can include a 2% initial-acquisition fee, a 5% or 13% store-services fee, and a 5% core-technology commission; reduced rates apply to certain small businesses and longer subscriptions. Earlier alternative terms also included a EUR 0.50 core-technology fee above one million annual first installs during the transition.

These structures preserve economics, but at lower effective rates and with greater complexity than the historic 15–30% model. The EUR 500 million Commission penalty, failed gatekeeper-designation challenge, and pending national actions indicate that regulatory pressure is structural. The direct financial loss remains uncertain because Apple does not disclose EU App Store commission revenue, developer migration, or external-payment conversion. The direction of travel is clear: developers receive more steering freedom, and Apple must justify a narrower fee base.

The third uncertainty, Greater China, has resolved favorably over the latest two quarters but remains strategically open. Fiscal Q3 revenue was USD 18.8 billion, about 17% of group sales, and nine-month growth reached 30%. IDC’s share data confirm that the rebound reflected real handset shipment gains, not currency alone. Huawei remains the market leader, domestic ecosystems are improving, and political or government-use restrictions could affect demand independently of product quality.

A one-quarter share rebound does not restore Apple’s former structural position. The company lost 17% of mainland shipments in 2024 and faced intense competition from Huawei and other domestic brands. The current iPhone cycle, stable pricing relative to Android devices affected by memory inflation, and promotional support all contributed to the recovery. China should now be treated as an improving but volatile earnings variable rather than a continuously deteriorating one.

The three prior uncertainties can be classified plainly. AI has moved favorably from delay to a credible, testable architecture but remains unresolved until broad shipment and usage. US search-payment risk has resolved favorably for near-term cash flow while remaining open on appeal and contract durability. EU regulation has moved unfavorably for Apple’s legal control, though not yet catastrophically for revenue. Greater China has resolved favorably in the near term and remains open over a three-to-five-year horizon.

Valuation, risks, catalysts, and tracking

At USD 308.91, Apple trades on approximately 35.9 times trailing underlying EPS of USD 8.61. The corresponding earnings yield is 2.79%, below the July 31 10-year Treasury yield of about 4.74% by 1.95 percentage points. The stock’s 0.35% dividend yield does little to close that gap. Investors require continuing EPS growth and a persistent premium multiple to earn more than the government-bond yield.

Historical comparisons reinforce the conclusion. Independent series place Apple’s 10-year average or median P/E near 25–27 times, with the current headline multiple about 32–40% above those benchmarks. The current multiple is below the extreme levels reached around late 2024 and parts of 2026, but remains above the normal post-Services re-rating range.

The premium can persist because Apple has more predictable consumer demand, less capital intensity, exceptional cash generation, and a huge installed base. It becomes difficult to justify if Services growth settles below 10%, regulatory revenue is lost, AI does not accelerate upgrades, or the risk-free rate remains close to 5%. Peer expensiveness cannot substitute for absolute return.

Cash-flow passthrough is favorable. Over fiscal 2021–2025, operating cash flow was 1.14 times net income. For the trailing period through June 2026, operating cash flow is approximately USD 146.7 billion, calculated as fiscal 2025 operating cash flow less the first nine months of fiscal 2025 plus the first nine months of fiscal 2026. Trailing cash capital expenditure is about USD 10.0 billion, producing conventional free cash flow near USD 136.7 billion.

Apple does not disclose maintenance versus growth capex. The valuation assumes approximately 75% of trailing capex, or USD 7.5 billion, is maintenance expenditure. This is intentionally conservative relative to Apple’s asset-light model but recognizes that stores, data centers, tooling, laboratories, and facilities require replacement. Owner earnings under that assumption are about USD 139.2 billion, or approximately USD 9.47 per diluted share. Deducting all capex gives free cash flow per share of roughly USD 9.30.

That puts the owner-earnings multiple at approximately 32.6 times and the all-capex free-cash-flow yield at about 3.0%. This is modestly better than the 35.9-times underlying accounting P/E, but the difference is less than 10%, far below the framework’s 30% threshold for abandoning earnings-based valuation. The scenarios use a blend of normalized owner earnings, free cash flow, and forward P/E.

The valuation below is independently derived and does not adjust the prior report’s bands.

† Includes an assumed dividend contribution near the current yield and is not a forecast of market timing.

This is valuation-scenario analysis within a research framework, not investment advice.

The conservative case assumes the current device cycle fades, memory and advanced-node costs keep underlying margin near 46–47%, Services slows, and the multiple returns toward Apple’s long-run center. The base case assumes component conditions normalize during fiscal 2027, Services remains in the low-to-mid teens in constant currency, Siri ships broadly without transforming the revenue mix immediately, and buybacks add roughly 1–2 percentage points to EPS growth. The optimistic case requires meaningful AI-assisted upgrades, sustained China strength, stable search payments, and a valuation premium near recent highs.

Apple’s present price lies near the upper half of the base fair-value range, not below the conservative case. Upside depends increasingly on the optimistic scenario. The strongest operating quarter in years has improved forward earnings power, but much of that improvement is already capitalized.

The most fragile base-case assumption is the 28–30-times multiple. Cutting the midpoint multiple of 29 times to 70%, or 20.3 times, reduces the base valuation from about USD 300 to roughly USD 211–215, even without cutting normalized owner earnings. This is why interest rates and narrative durability matter almost as much as product revenue.

If underlying earnings remain flat for three years and the multiple does not change, the annual return is approximately the 0.35% dividend yield. That is far below the 4.74% 10-year Treasury yield. If the multiple returns to 26 times, the share price would approach USD 224 before dividends, implying a negative annualized return around 10%. There is no margin of safety at this buy price.

The current price is also above the conservative fair-value range by roughly one-third. The margin of safety relative to that scenario is zero. This is a good business at a demanding price; waiting for a lower price offers a better prospective return, though the opportunity cost is missing further gains if Siri succeeds quickly and supply tightness reflects more durable demand than assumed.

Margin-of-safety sufficiency verdict: none.

Five risks can plausibly cause permanent loss.

The first is valuation compression: high probability, medium-to-high impact. Watch the 10-year Treasury yield, Apple’s forward P/E, and the gap between earnings yield and bond yield. If rates remain near 5% while EPS growth returns to mid-single digits, a decline from 36 times to 25 times would remove approximately 30% of equity value even if earnings do not fall. The transmission path is direct through the discount rate and narrative rather than operating insolvency.

The second is Services-rent erosion. Probability is medium and impact high. The indicators are Services growth, disclosed App Store fee changes, Google-search remedy developments, developer adoption of external payments, and changes in advertising or licensing commentary. A 50% cut to the estimated Google payment could reduce annual EPS by roughly USD 0.52 under the model above. Simultaneous App Store pressure would reduce both growth and the margin mix that supports Apple’s premium multiple.

The third is AI-interface disintermediation. Probability is medium and impact high over three to five years. The indicators are rebuilt-Siri availability, task-completion reliability, usage frequency, developer integration, external-model branding, and the share of discovery or commerce that begins in third-party agents. If users primarily interact with Google, OpenAI, Meta, or another cross-platform agent, Apple could retain device sales while losing part of the distribution and Services profit pool.

The fourth is a prolonged component and pricing shock, medium-to-high probability over the next year and medium impact. Component inventory has already increased sharply, and management expects worse constraints and higher memory costs. Watch product lead times, memory prices, advanced-node allocation, inventory, gross-margin guidance, and price increases. Lost units reduce revenue; higher input costs reduce margin; compensating price increases can weaken future demand.

The fifth is renewed China deterioration: medium probability, high impact. Greater China contributes around 17% of quarterly revenue, while Huawei retains the larger smartphone share. Indicators include quarterly regional revenue, IDC shipment share, local promotions, government-use restrictions, and the timing of Apple Intelligence approval. A share decline would hurt units and mix; a political restriction could also compress the valuation because investors would treat the supply chain and demand exposure as correlated.

Leadership transition is a sixth, lower-probability but consequential variable. Ternus has long product experience and Cook will remain executive chairman, reducing immediate discontinuity. The risk is that a hardware-centered succession preserves operational excellence but moves too slowly in AI software and services. Observable evidence will include senior-team turnover, product delays, acquisition strategy, and whether Apple’s model dependence falls or increases.

Positive catalysts include broad shipment of the rebuilt Siri with high reliability; easing advanced-node and memory shortages; sustained iPhone growth without major price elasticity; Services reacceleration above 14% in constant currency; continued mainland China share gains; and an appeal outcome that preserves search payments. A lower Treasury yield would also support the multiple, though that would be a capital-markets catalyst rather than a business improvement.

Negative catalysts include September-quarter revenue below the 9–11% guide, an underlying gross margin below 46%, another Services miss, major iPhone price increases followed by weaker demand, a new US remedy against search payments, wider developer migration from Apple payments, delayed China or EU Siri availability, and evidence that external AI agents are replacing App Store or search discovery.

† October 29, 2026 is an unconfirmed market-calendar estimate; Apple had not announced the date as of August 1.

The financial indicators should be read together. A Services slowdown with stable total margin is less alarming if hardware mix offsets it; a Services slowdown accompanied by falling margin indicates that the profit engine is weakening. Inventory is protective while demand exceeds supply, but becomes a warning if unit growth slows. The Treasury gap is a valuation discipline rather than a trading signal: it measures how much growth investors must receive before accepting equity risk.

The legal dashboard should track the Google-search appeal, annual default-contract disclosures, EU App Store enforcement, UK steering rules, and developer migration. The AI dashboard should track broad Siri release dates, supported countries and languages, benchmarked task reliability, third-party model branding, and whether Apple discloses any measurable upgrade or Services effect.

Cross-synthesis, final conclusion, uncertainties, and sources

Vertically, Apple has proven that it can turn technical components into coherent mass-market systems, use design and distribution to command premium pricing, build proprietary silicon, and convert an installed base into recurring cash flow. The company survived its 1990s crisis by narrowing focus and choosing integration. It then used the iPhone to create an operating-system and commerce platform, and Tim Cook’s supply-chain discipline to scale that platform with limited asset intensity.

Past success combined era tailwinds with distinctive management capability. The personal-computing and mobile-internet waves created enormous markets, but many participants failed to capture durable economics. Apple’s contribution was to control the endpoint, make complicated technologies accessible, build direct customer distribution, and preserve pricing power. Its success was neither pure luck nor pure invention. It repeatedly arrived after foundational technologies existed and captured value through integration.

Most of those capabilities remain present. The brand is intact, the installed base is larger than ever, custom silicon is deeply embedded, cash conversion is strong, and the company can spend more than USD 40 billion annually on R&D without weakening its balance sheet. Fiscal 2026 has shown that iPhone and Mac can still generate double-digit growth at enormous scale. China’s rebound and record geographic revenue weaken the argument that Apple has already entered structural decline.

The missing capability is independent proof in generative AI. Apple has built the architecture that matters to its traditional strengths: personal context, permissioned application actions, device integration, on-device processing, and private cloud execution. It has not yet shown that the system works reliably for ordinary users at scale or that it can command the model economics. Collaboration with Google may be rational procurement, just as Apple procures displays, fabrication, and memory. Models differ from ordinary components because they can possess their own brands, developer ecosystems, cloud economics, and direct user relationships.

Horizontally, Apple’s advantage over Microsoft, Alphabet, Meta, Amazon, and NVIDIA is its control of the premium consumer endpoint without matching their infrastructure burden. It can distribute AI to billions of devices while committing a fraction of peers’ capex. Its weakness is that those peers own the cloud capacity, frontier models, enterprise relationships, advertising systems, and accelerator economics that are capturing the first wave of AI spending.

The capital-light strategy will look brilliant if model performance converges and inference costs fall. Apple could source competing models, select the best provider for each task, preserve privacy and interface control, and return excess cash. The strategy will look defensive if model providers become the primary brands through which users communicate and transact. Apple would then own valuable hardware while renting part of the intelligence that determines engagement.

The current valuation rewards both proven history and unproven future success. A 35.9-times underlying P/E is not the valuation of a mature hardware company expected to grow in line with nominal GDP. It requires sustained high-margin Services, continued buybacks, resilient pricing, and a successful role in AI. The earnings-yield deficit to the Treasury market confirms that investors are prepaying for growth.

The market may be misjudging two things in opposite directions. It may understate how valuable Apple’s control of permissions, personal context, and application actions will be in agentic computing. Foundation-model benchmarks alone do not determine which interface users trust to send messages, access health data, make payments, or operate devices. Apple owns those privileged operating-system pathways.

The market may simultaneously overstate how much of that value Apple can retain when model suppliers provide the underlying intelligence and regulators weaken default and payment control. The Google relationship illustrates the tension: Google may pay Apple billions for distribution while Apple pays or otherwise compensates Google for model capability. The net economics and strategic balance are opaque.

For the next year, the critical variables are supply availability, memory costs, underlying gross margin, September-quarter execution, and broad Siri shipment. These will determine whether fiscal 2026’s acceleration extends into fiscal 2027 or represents the peak of the product cycle.

For the next three years, the critical variables are Services composition, search-payment durability, App Store economics, AI-assisted replacement rates, and China share. Apple can tolerate moderate hardware volatility if Services compounds and regulatory leakage remains contained. It cannot justify the current premium if Services loses both growth and margin.

For the next five years, the decisive variable is ownership of the consumer AI interface. Apple’s installed base and operating-system control give it a strong starting position. The threat becomes structural if users regard a third-party agent as the primary environment and the Apple device as interchangeable access hardware.

The prior report’s three uncertainties have not produced one common verdict. Greater China has improved materially. Search-payment risk has improved in the near term, although annualized, nonexclusive contracts and appeals reduce durability. EU regulation has moved against Apple’s historic control. AI has progressed from delayed product to credible developer test, but the broad user and economic evidence remains absent.

The post-earnings selloff should not be treated automatically as an entry point. The share price remains above the May reference price, and the market reacted to information that lowers near-term revenue visibility and underlying margin. The selloff removed part of the most optimistic premium; it did not create a conservative valuation.

The company becomes a better investment under one of two conditions. The first is price: a decline toward USD 175–185 would provide at least a 20% discount to the conservative fair-value range. The second is evidence: broad Siri adoption, sustained Services growth above 12%, normalized component costs, and durable search/App Store economics could raise conservative earnings power enough to justify a higher purchase range.

The judgment should be overturned positively if Siri establishes high engagement and task reliability, Services growth reaccelerates without licensing dependence, and gross margin remains above 48% after refunds disappear. It should be overturned negatively if underlying gross margin stays below 46.5%, Services growth falls below 10%, the Google payment is materially reduced without replacement monetization, or China returns to sustained contraction.

Bull reasons:

• Fiscal Q3 underlying EPS still grew about 22% after removing the USD 0.11 refund, while revenue grew 16%, showing that the acceleration was not created by the one-off item.
• iPhone, Mac, and Greater China revenue grew 22%, 29%, and 22%, respectively, giving Apple a broader demand base than it had at the time of the prior report.
• Services generates approximately 42% of segment gross profit on only 28% of revenue, and its 75.6% gross margin continues to raise group earnings quality.
• Apple can distribute AI across more than 2.5 billion active devices while spending a small fraction of hyperscaler capex.
• The US search judgment preserved Google payments in the near term, removing the most severe immediate licensing scenario.

Bear reasons:

• The stock trades at approximately 35.9 times underlying trailing EPS while its earnings yield is nearly two percentage points below the 10-year Treasury yield.
• September guidance points to worsening supply constraints, a 2.5-point FX drag, and an underlying gross-margin range near 46–47%.
• Services missed consensus by about USD 481 million, while an estimated USD 20 billion search payment may represent roughly one-sixth of annual Services revenue.
• The rebuilt Siri had not broadly shipped as of the report date, and Apple relies partly on Google-developed model capability under undisclosed commercial terms.
• EU regulation has already produced a EUR 500 million fine, alternative fee structures, and a failed challenge to Apple’s gatekeeper designation.

Pre-mortem. The first plausible 50% loss script begins in fiscal 2027. Memory and advanced-node shortages persist, Apple raises iPhone prices, unit demand slows after the iPhone 17 cycle, and underlying gross margin falls from about 48% to 45%. Services growth drops below 8% as external payments expand. Underlying EPS returns to roughly USD 8.50 and the market applies 20 times earnings instead of 36 times, producing a price near USD 170.

The second script unfolds through 2028–2029. Google, OpenAI, Meta, or another provider establishes a cross-platform agent that becomes the main environment for search, shopping, communication, and application discovery. US appeals materially reduce the Google payment, while European and UK rules lower App Store commissions. A USD 1.00-plus EPS licensing loss combines with slower device upgrades, reducing sustainable EPS to USD 8.00–8.50. At 18–20 times earnings, the shares trade between approximately USD 145 and USD 170, around half the current price.

Research uncertainties remain material. Apple does not disclose the Google payment, individual Services revenue lines, App Store regional profit, Siri model-inference costs, third-party model commercial terms, unit shipments, or the revenue lost to supply constraints. The USD 20 billion search-payment estimate and 90% incremental margin are external and analytical assumptions. The rebuilt Siri’s developer access provides little evidence on mass-market reliability. Future component costs are highly uncertain because advanced-node and memory capacity depend on investment decisions outside Apple’s control.

The source base prioritizes Apple’s fiscal Q3 2026 Form 10-Q, fiscal 2025 Form 10-K, earnings releases, conference-call materials, and official WWDC and machine-learning disclosures. Regulatory analysis uses the US Department of Justice, European Commission, EU court developments, and Apple’s developer terms. Market and competitive evidence uses US Treasury data, Reuters, IDC, official peer earnings materials, and dated market feeds.

Apple remains one of the strongest cash-generating consumer franchises ever built. Its fiscal 2026 acceleration, China recovery, installed base, Services margin, and capital-light structure support continued ownership. The current price nevertheless discounts continued product strength, stable regulatory rents, successful Siri deployment, and a durable premium multiple at a time when the risk-free rate is higher than Apple’s earnings yield.

At USD 308.91, existing holders receive a plausible but modest base-case return and retain meaningful upside if Apple becomes the trusted consumer AI interface. New buyers receive no conservative margin of safety. The main concern is the combination of a high multiple and strategic dependence: Apple must preserve control of distribution while purchasing some of the intelligence that could become the next distribution layer.

【Company-profile scores】

• Fundamental quality: high
• Growth: medium
• Moat: strong
• Financial soundness: strong
• Management credibility: high
• Valuation attractiveness: low
• Risk level: medium
• Suitable investor type: long-term growth investors able to tolerate valuation compression; not suitable for value investors requiring a current margin of safety

【Investment rating】

• Rating: Hold
• One-line thesis: Strong underlying growth and cash generation support ownership, but a 35.9-times adjusted P/E leaves no conservative margin of safety.
• Ideal buy price: see the required line below
• Acceptable hold price: USD 255–340
• Clearly overvalued price: USD 390–420
• Current-price classification: acceptable hold
• Whether to wait for a better price: yes for new capital; the preferred entry is USD 175–185, or a higher price only after broad Siri adoption and sustained underlying margins above 48%
• Opportunity cost of waiting: missing an AI-assisted upgrade cycle and further multiple expansion if Siri succeeds rapidly
• Target holding horizon: 3–5 years
• Expected annualized return: approximately -8.6% conservative, 2.9% base, and 12.7% optimistic
• Max-loss risk: approximately 45–53% if Services rents fall, underlying margin approaches 45%, and the P/E compresses to 18–20 times
• Reassessment-trigger signals: underlying gross margin below 46.5% for two consecutive quarters; Services growth below 10%; a material reduction in Google search payments; Greater China revenue contraction below 5% for two quarters; broad Siri delay beyond fiscal 2027 or persistently poor task reliability

【Ideal Buy Price】175–185 USD

The range is approximately 20% below the midpoint of the independently derived USD 220–235 conservative fair-value range. It provides protection against a return toward Apple’s historical valuation center, weaker Services economics, and lower underlying margins.

Final verdict: Hold, upgraded in substance from the prior Watch posture because fundamentals and China improved, but the stock remains unsuitable for fresh balanced-risk buying at USD 308.91.

【Valuation Range】

• current: 308.91 (close as of 2026-07-31)
• bear (conservative · ideal buy zone): [175, 185]
• base (fair · acceptable hold zone): [255, 340]
• bull (optimistic · above the clearly-overvalued line): [390, 420]

Other tickers mentioned

• US MSFT.US: enterprise software and cloud peer with direct AI monetization and far higher infrastructure spending
• US GOOGL.US: search-payment counterparty, AI-model supplier, Android competitor, and the most important strategic comparator
• US META.US: consumer-attention and AI-agent competitor investing heavily in models and infrastructure
• US AMZN.US: hyperscale-cloud peer whose AI exposure requires materially greater capital intensity
• US NVDA.US: primary beneficiary of accelerated-computing demand and an indirect cause of industry resource competition
• KR 005930.KRX: Samsung competes directly in premium smartphones, displays, memory, and devices
• TW 2330.TW: TSMC fabricates Apple silicon and is central to the advanced-node supply constraint
• US AVGO.US: custom-connectivity and semiconductor supplier participating in Apple’s supply chain
• US QCOM.US: modem supplier whose Apple exposure is declining as Apple internalizes more silicon
• HK 00981.HK: SMIC represents China’s domestic semiconductor ambitions and geopolitical supply-chain context
• CN 688256.SHG: Cambricon represents China’s domestic AI-accelerator ecosystem
• HK 01810.HK: Xiaomi competes for handset share and ecosystem spending in Greater China
• HK 00268.HK: Kingdee illustrates the broader Chinese software ecosystem that affects local platform competition

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.