Apple

IThe business

Apple designs a phone, a computer, a tablet, a watch and a pair of earbuds, writes the software that runs on all of them, and then sells a second business on top: the apps, music, storage, insurance, advertising and payments that people use on those devices.³ It does not run factories. Contract makers in China mainland, India, Japan, South Korea, Taiwan and Vietnam build almost everything.³ About 166,000 people work for Apple itself.³

The number I would watch above all is the installed base. On the July call it said active devices passed 2.5 billion, an all-time high, with paid subscriptions above 1.5 billion.⁵ A customer who owns an iPhone is not the customer Apple wins once. That customer is worth an upgrade every few years, an iCloud plan, a case of apps and a stream of fees.

Fiscal 2025, which ended on 27 September 2025, brought $416.2B of net sales. iPhone was $209.6B of it, or 50%. Services were $109.2B, or 26%. Mac, iPad and Wearables, Home and Accessories made up the rest, at $33.7B, $28.0B and $35.7B.³ The fiscal 2026 10-K isn’t out. Apple reports its fourth quarter on 2 November, after the date limit of this page, so I work from the twelve months to 27 June 2026, which I build from the 2025 10-K and the June 2026 10-Q.²³ Over that period sales were $466.8B and operating income $154.9B, a 33.2% margin C (calculated).²³

What it sells

The nine months to June 2026 were strong. Sales rose 16%, to $364.4B. iPhone grew 22%, Mac 9%, iPad 3%, Wearables, Home and Accessories 2% and Services 14%.² The iPhone 17 line, launched in September 2025, is selling in numbers that surprised Apple: the company says demand ran ahead of its forecast and that supply of advanced chips, not demand, is the limit on the September quarter.⁵² Mac got a lift in the June quarter from laptops, including a new model called MacBook Neo launched in the March quarter.²¹²

The chart shows how much of Apple is still one product. In fiscal 2017 iPhone was 61% of sales; in fiscal 2025 it was 50% C (calculated).¹¹³ The share fell because Services and Wearables grew faster, not because iPhone shrank. It is still half the company. If iPhone has a bad year, Apple has a bad year.

Services

Services is five things that look unrelated. Advertising includes the licensing payments from search partners and Apple’s own ad platforms. AppleCare is paid support and insurance. Cloud services store people’s photos and files. Digital content includes the App Store and the subscription services for music, video, games, news and fitness. Payment services include Apple Pay and the Apple Card.³ Apple doesn’t split Services revenue among those lines. What it does say is that fiscal 2025 growth of 14% came mainly from advertising, the App Store and cloud services, and that growth in the nine months to June 2026 came mainly from advertising and cloud.³² App Store sales were softer in mobile games.⁵

Two of the largest parts of that profit are not Apple’s alone. The advertising line includes what Google pays Apple to be the default search engine in Safari. In his August 2024 opinion, Judge Amit Mehta found that Google’s payment to Apple in 2022 was an estimated $20 billion, nearly double the 2020 payment, which he said equalled 17.5% of Apple’s operating profit.⁷ Apple doesn’t disclose the figure. If the 2022 figure still held, it would be about 17% of fiscal 2022 operating income and about 24% of fiscal 2025 Services gross profit, or 15% of fiscal 2025 operating income C (calculated).⁷¹⁰³ It would also be hard to replace: an internal Apple assessment from 2018, recorded in the opinion, concluded that even with a search engine of its own Apple would lose more than $12 billion of revenue in the five years after separating from Google.⁷ I come back to this in the next section, and again in Valuation, because it is the single largest pure-profit item I can find that someone else controls.

The App Store is the other. Apple keeps a commission on purchases made inside apps. Courts in the US and regulators in Europe are cutting at it. The 10-Q describes the injunction in the Epic Games case, which at one point barred Apple from charging any commission on purchases that users make outside an app; the Ninth Circuit held in December 2025 that Apple can charge some commission on those link-out purchases, and the Supreme Court agreed on 30 June 2026 to review the contempt standard.² In the EU, the Commission fined Apple €500 million in April 2025 and found preliminary breaches in a second case, where fines can reach 10% of annual worldwide net sales.² I put dollar sizes on these risks in the valuation.

Where it sells

The US was $151.8B, or 36% of fiscal 2025 sales. Europe, which for Apple includes India, the Middle East and Africa, was $111.0B. Greater China was $64.4B, Japan $28.7B and the rest of Asia Pacific $33.7B.³ Direct sales through Apple’s own stores and website were 40% of net sales; the other 60% went through carriers, wholesalers and retailers.³

China is the place to watch. Greater China sales were $48.5B in fiscal 2016, fell to $40.3B in 2020 and then jumped to $68.4B in 2021 and $74.2B in 2022, mostly on iPhone. They slid for three years to $64.4B in fiscal 2025.^41314103 Then they turned: sales rose 22% in the June 2026 quarter and 30% in the nine months, to $64.8B, driven by iPhone, with the renminbi helping.² The bull and bear cases for the next few years often come down to whether that rebound is real. Apple also reported that it received approval in China for its original Apple Intelligence features and still has work to do for Siri AI there.⁵

The two shocks of 2026

The first is tariffs. The US imposed new tariffs from the second quarter of 2025 on imports from China, India, Japan, South Korea, Taiwan, Vietnam and the EU, among others. On 20 February 2026 the Supreme Court struck down the ones imposed under the International Emergency Economic Powers Act. Apple applied for refunds and books them as a cut to the cost of products it sold.²³ In the June quarter the refunds added about 2 points to gross margin and $0.11 to earnings per share. The company guides to about one point in the September quarter.¹⁵⁵ The Commerce Department also published initial results of a Section 232 semiconductor review in January and the US has since imposed new tariffs under Section 301.² I strip the refunds out of my starting numbers, because they are a one-time gain from a past cost.

The second is memory. The 10-Q says Apple is in a period of supply constraints and rising costs for advanced chips, flash storage and DRAM, and that it expects the trend to intensify and to hit revenue, cost and gross margin.² Tim Cook called it a “100-year flood” on the call and said Apple had reluctantly raised the price of iPad and Mac.⁵ DRAM comes from three suppliers. Apple said it is evaluating all its options.⁵ That is an unusual thing for Apple to say. It guided September-quarter gross margin to 47% to 48%, including the point of refunds, against 50.1% in June, and said the memory cost increase was bigger than the whole step down at the midpoint.⁵

What does the September guide say about the year? Revenue up 9% to 11%, operating expenses of $19.1B to $19.4B and a tax rate near 16.5%.⁵ Nine months of reported results plus that guide give me fiscal 2026 revenue of about $477B, operating income of about $153B before refunds and a 32% margin E (estimate). Nobody outside the company knows the real figure until 2 November.

IIThe moat

A durable advantage is a cost or a habit that a rival with money and talent can’t copy in a reasonable time. Apple’s is easy to describe and hard to price: it designs nearly the entire product, in the words of its own 10-K, from the hardware and the operating system to the applications and the services.³ Few others do that. The 10-Q names Android, Windows and the game consoles as the platforms Apple competes with for developers; Android and Windows run on hardware made by many companies.² Apple does both halves and keeps the margin from both.

The loop works like this. A better product wins a customer. The customer adds a watch, a Mac, a subscription. Each addition makes leaving more expensive, and because the next phone is designed to work with the last, upgrading is easy. Apple’s services then earn a high margin on the same customer for free. The risk is not that the loop breaks; the risk is that a judge opens it from one side.

Evidence it is widening

• Gross margin rose from 39.1% in fiscal 2016 to 46.9% in 2025, and to 48.7% in the twelve months to June 2026 with refunds C (calculated).⁴³²
• Services gross margin went from 55.0% to 75.4% in eight years, and the line grew from $32.7B to $109.2B C (calculated).¹¹³
• Active devices hit a record in every major product category and every geographic segment in the June quarter.¹⁵
• iPhone, Mac and Services each set a June-quarter revenue record, and every geographic segment grew by double digits.¹⁵

Evidence it is narrowing

• The products gross margin has barely moved: 35.7% in fiscal 2017 and 36.8% in 2025.¹¹³ All the gain is mix toward services.
• Court orders already limit the commission on link-out purchases in the US, and the EU has fined Apple and may fine it more.²
• A court found in August 2024 that Google violated US antitrust law. On 2 September 2025 it ordered certain remedies, and both the Justice Department and Google have appealed.² Apple’s 10-Q says a reversal of the order could bring the remedies the Justice Department first asked for, which would ban Google from offering Apple commercial terms for search distribution.²
• The Justice Department’s monopolisation suit against Apple, filed in March 2024, is still pending.²
• Apple’s next foundation models will be based on Google’s Gemini models and cloud technology, under a multi-year collaboration announced in January 2026, and Apple concluded Google’s technology was the most capable base.¹⁶ That is sensible. It is also a sign that the model layer is something Apple does not yet own.

On balance I think the hardware-and-software integration is wide and stable, the services margin is wide and exposed, and the AI layer is open. The part that compounds is the installed base. The part the market prices as if it were certain is the services profit.

IIIThe competitive advantage

The cleanest comparison for Apple, on this site, is with the two other US companies that run a platform used by billions of people: Microsoft and Alphabet. None of them is a direct rival in every market, but they compete for the same dollars of attention, for developers, and for the same AI talent and chips.

Apple stands out in two ways. First, it needs far less capital. Spending on property and equipment in the twelve months to June 2026 was $10B, 2.2% of sales C (calculated).² On my Alphabet page, equipment takes 89% of operating cash flow. Second, the gap between what the price needs and what I expect is the widest of the three: 15% a year against my 4.9% for Apple, 16% against 12% for Alphabet, and 25% against 18% for Azure at Microsoft. The business is not the problem. The price is.

What Apple does better than a phone maker

• It keeps the margin in a market that cuts prices. The 10-K says rivals have aggressively cut prices and lowered product margins.³ Apple’s operating margin was 32.0% in fiscal 2025 and 33.2% in the twelve months to June 2026 C (calculated).³²
• It spreads fixed cost over a huge base. R&D was $34.5B in fiscal 2025 and $42.9B in the twelve months to June 2026, up 33% in the nine months on infrastructure for artificial intelligence C (calculated).³² A company with a tenth of the sales would struggle to spend that.
• It has cash to wait. Cash and securities of $147B against $84B of debt, a net $62B.² A rival squeezed by memory prices can’t absorb the cost for long. Apple can.

What it does worse

• It depends on a few suppliers and a few countries. Most of the hardware is made in China mainland, India, Japan, South Korea, Taiwan and Vietnam, and many parts come from one source.³
• It is a minority player. The 10-K says Apple has a minority share in smartphones, PCs, tablets and wearables, and that some of its markets have at times stopped growing.³ Its edge is margin, not volume.
• It borrows its AI. Apple’s next foundation models are based on Google’s Gemini, and the 10-Q lists constrained supply of compute as a risk for its own AI features, which depend on third-party cloud capacity.¹⁶²

IVManagement

The big fact is that Apple has just changed chief executive for the first time in fifteen years. On 17 April 2026 the board named John Ternus, who runs hardware engineering, as CEO from 1 September. Tim Cook, who had run Apple since 2011, became Executive Chair. Art Levinson became Lead Independent Director.¹⁷

Almost everyone at the top is a long-service insider. Khan joined in 1995, O’Brien in 1988, Cook in 1998. The CFO and COO changed in 2025, as planned.⁹ Ternus is a hardware engineer. The last chief executive was an operations specialist, and the one before him was a founder. The lesson of the last transition, from Steve Jobs to Cook in 2011, is that a successor chosen from inside can be right. It says nothing about whether this one is.

Pay, and what it pays them to do

Cook earned $74.3M in fiscal 2025: a $3M salary, $57.5M of stock awards, a $12M cash incentive and $1.8M of other pay.⁹ The proxy says the committee targets his pay between the 80th and 90th percentiles of peer-company CEOs.⁹ That is a choice to pay more than most. The pay ratio was 533 to 1.⁹

The design matters more than the amount. The cash incentive is set by net sales and operating income, each weighted equally, with a payout from 50% of target at threshold to 200% at the maximum. Fiscal 2025 paid the maximum for every executive, because sales were $416.2B and operating income $133.1B.⁹ The stock is the big part. Three quarters of Cook’s grant were performance RSUs, which vest on Apple’s total shareholder return against the S&P 500 over three years. The 2022 awards vested at 187% of target because Apple’s return was at the 81st percentile.⁹

The new terms continue the pattern. Ternus gets a $3M salary, a prorated $2.5M award for fiscal 2026 and a target of $55M in stock for fiscal 2027, 75% of it paid on relative shareholder return and 25% as time-based units that vest over four years. Cook’s salary falls to $2M from 26 September 2026, and he gets a $45M target award, half performance-based. If he retires after a year, that award still vests on the original dates.⁸

So follow the incentive. Relative shareholder return pays executives to lift the share price faster than the index, whatever lifts it. That includes buybacks that shrink the share count, a rising multiple, and a mix shift toward services. It doesn’t reward return on capital as such, the cash return after stock pay, or the price Apple pays when it buys back stock. And when the price is high, it pays them more for the same business. I don’t think anyone is gaming it. But a plan like this can’t tell good capital allocation from luck.

At the February 2026 meeting, 91% of the votes cast for or against backed the pay plan.¹⁹ A shareholder proposal to audit the company’s ties to China got 129M votes for and 8.94B against.¹⁹

Their own money

Cook owned 3.28M shares in January 2026, worth about $1.13B at $344.63 C (calculated).⁹¹ Levinson held 4.13M, about $1.42B, and Khan 1.07M. All directors and officers together held 9.08M shares, about $3.1B and under 0.1% of the company C (calculated).⁹ The two largest holders are index managers: Vanguard with 9.63% and BlackRock with 7.10%.⁹ Berkshire Hathaway held 227,917,808 shares at 30 June 2026, the same count as at 31 March, about 1.6% of Apple; it is a holder, with no seat on the board. I cover it on its own page, where Apple is the largest line.²⁰ Nobody running Apple owns enough to steer it. Cook’s stake is large in dollars and small as a share of the company, and it is mostly stock he earned.

What they did with the money, over a full cycle

I take fiscal 2017 to 2025 as the cycle: a trade war, a pandemic, a boom in 2021, a drop in 2023 and a memory shortage. In those nine years Apple generated $858B of operating cash. It spent $98B on capital goods, paid $130B in dividends and $683B on buybacks, and paid $40B in taxes on stock it handed to staff C (calculated).⁶³ Stock pay was $75B C (calculated).⁶ Net cash went from $151B at the end of fiscal 2016 to $34B at the end of 2025 C (calculated).⁴³

Three things stand out. First, the buyback was enormous and the timing was good for a long time. Apple spent $651B from fiscal 2018 to 2025 and retired about 6.5B shares at an average of about $100 a share, split-adjusted C (calculated).⁶ Those shares were bought at $45 in fiscal 2018 and $79 in 2020, which was cheap. By fiscal 2025 the average was $226, and in May and June 2026 it was about $297.⁶² My base value is $179. At those prices Apple is buying stock above what I think it is worth. Second, all of this was paid for with net cash. Cash after stock pay was less than payouts in each of the last eight years, and the gap came out of the balance sheet.⁶ In the first nine months of fiscal 2026 that reversed, because cash flow jumped. Third, the company renewed its authority twice: $100B in May 2025 and another $100B in April 2026. As of 27 June 2026 $138B remained.¹²²

The record is large, with one caveat. A buyback adds to value only when the price is below value. Over the first half of the period it plainly was. Now I think it isn’t. Relative-return pay gives management no reason to stop: fewer shares lift earnings per share and probably the share price.

VFinancials

Apple’s accounts are plain. There is one class of stock, a large and conservatively invested cash pile, and a simple income statement. The two judgements that matter are the tax provision, which swung from 14% to 24% in recent years, and the revenue that is deferred and released over time.³²¹

Margins

Operating margin went from 27.8% of sales in fiscal 2016 to 32.0% in 2025 and 33.2% in the twelve months to June 2026 C (calculated).⁴³² Almost all of the improvement came from gross margin, which rose 7.8 points, because services grew from 14% of sales to 26% C (calculated).¹¹³ Operating expenses were 14.9% of sales in fiscal 2025 and stayed near that until the June 2026 quarter, when R&D alone rose to nearly 11% of sales on investment in artificial intelligence C (calculated).³²

Earnings quality

Net income grew 10.5% a year from fiscal 2016 to 2025 and earnings per share 15.3% a year C (calculated).⁶ The difference is the shrinking share count: diluted shares fell 32% in those nine years C (calculated).⁶ Stock pay was $12.9B in fiscal 2025 and $13.7B in the twelve months to June 2026, 11% of net income, and it is already counted as a cost in net income.³² As of June 2026, $26.7B of unrecognized compensation cost for restricted stock units was waiting to be expensed.²

The tax rate is the other thing to read with care. It was 24.1% in fiscal 2024 because of a $10.2B charge for the European Commission’s State Aid decision, and 15.6% in 2025 without it.³ It was 17.6% in the nine months to June 2026, and Apple guides to about 16.5% for the September quarter.²⁵ I use 17.5% in my base case E (estimate).

The tariff refunds are the third. In the June quarter they added $0.11 to earnings per share, about $1.6B after tax.¹⁵ I take that out of my starting earnings: underlying net income in the twelve months to June 2026 is $127.3B, or $8.65 a share, against $128.9B reported C (calculated).²³¹⁵

From profit to cash

Operating cash flow in the nine months to June 2026 was $117.0B against $101.5B of net income, partly because cash taxes paid fell from $37.3B to $26.6B between the two nine-month periods.² The earlier period carried the State Aid settlement, funded from $15.8B held in escrow.³ The new period includes the last $8.8B of the deemed repatriation tax from the 2017 tax law, and a $5.5B build of inventory ahead of the iPhone cycle.² Both nine-month figures carry large one-offs in opposite directions, which is why I start from profit, not from cash flow.

The balance sheet adds an unusual source of cash. Accounts payable were $64.5B at 27 June 2026 against inventories of $11.1B, 5.8 times as much C (calculated).² Apple sells to customers before it pays its suppliers. As long as it keeps growing, that float grows too. If sales fell, it would shrink and take cash with it. I assume no benefit from it.

What an owner could take out

The number I care most about: if Apple stopped growing and just kept its products and services running, how much cash could it hand to its owners each year? I start from net income, which is already after stock pay, take out the refund, add back depreciation and amortization, and deduct what it takes to stand still.

The answer is about $127B, or $8.65 a share E (estimate). That is within 1% of underlying net income, because Apple’s capital needs are small next to its profit. The cross-check, cash flow less capital spending and stock pay, gives $123B, lower because it includes about $14B of one-off outflows, the transition tax and the inventory build.² So reported profit is a fair guide to what an owner could take out. An owner buying at $344.63 gets about 2.5% of the price back each year from today’s business, against 4.94% on a ten-year Treasury.²² Everything above that has to come from growth. At a 40% fall in the share price, the yield would be 4.2%, still below the Treasury.

The purchase that does not show in capital spending

One thing in the 10-Q needs saying plainly. Net intangible assets rose from $13.3B at 27 September 2025 to $25.4B at 27 June 2026, and gross intangibles from $24.95B to $38.2B.² Other non-current liabilities rose by $13.5B over the same nine months, and other purchase obligations, which include licensed intellectual property and content, doubled to $29.3B.² Payments for property and equipment in the nine months were only $6.8B, and the cash-flow line for other investing was $1.8B.² The filing doesn’t name what Apple bought or from whom. The pattern fits licensed rights or technology bought on instalments.

I can’t tell what it is, so I treat it as what it looks like. I deduct $13.0B from net cash as a debt-like claim E (estimate). And I count the amortization of these rights inside my standstill capital, because depreciation and amortization already include it, and in my cases I keep total capital needs, spending on property plus licensed rights, at 4.2% of sales in fiscal 2027 and about 3.3% by fiscal 2036 E (estimate). That is a higher ratio than the 2.2% of sales in the plant-and-equipment line alone. If I am wrong and these are one-time, the value of each share rises by about $0.89, or less than 1% C (calculated).

The balance sheet

At 27 June 2026 Apple held $39.5B of cash and equivalents and $107.0B of marketable securities, $146.5B in all, mostly Treasury, agency and corporate bonds and mortgage-backed securities.² Against that it owed $2.0B of commercial paper and $82.3B of term debt, $84.3B in all, so net cash was $62.2B C (calculated).² After the $13.0B deduction, my figure is $49.2B. Unrealized losses on the securities were about $2.1B net.² Apple also has $57.0B of manufacturing purchase commitments, nearly all due within a year, which I treat as ordinary working capital.²

The company guides to a dividend of $0.27 a quarter, which it raised 4% in May 2026.¹² That costs about $15.7B a year, or 12% of owner cash. The rest goes to buybacks.

VIValuation

Start with what would go wrong

Before any numbers, I ask what would most likely make this a bad purchase from here, and work back from that. For Apple the business is not the weak link. Three things are, and I put a size on each.

1. The business does fine and the multiple falls. Apple earns about 2.0 times as much per share in fiscal 2036 as today in my base case, on a share count that keeps shrinking, but the market pays 25 times earnings instead of 40. The return on $344.63 would then be about 2.7% a year, with dividends, less than a Treasury E (estimate). Nothing has to break for this to happen. In October 2016 the market paid 13.7 times and in 2018 18.4 C (calculated).¹⁶
2. The courts take the services profit. Google’s payment for Safari search was an estimated $20B in 2022, and the App Store commission is under injunction in the US and under investigation in the EU.⁷² If $10B of annual pre-tax profit went and the loss grew with the business after fiscal 2028, my base value would fall by about $9 a share; if $20B went, by about $17, or 10% E (estimate). That is large and survivable. It is also the part of the case I can’t put a probability on, and the 10-Q says a reversal on appeal could bring the Justice Department’s original remedy, a ban on Google offering Apple commercial terms for search distribution.²
3. Costs outrun price. Memory is rising, Apple’s own AI needs computing power it may have to rent, and R&D was 10.7% of sales in the June quarter against 9.4% a year before C (calculated).² Each point of operating margin lost across my whole forecast is worth about $5 a share, and three points about $16, or 9% of the base value E (estimate). Operating margin was 27.8% in fiscal 2016 and 24.1% in 2020, so a return to the high 20s is not a stretch.⁴¹³

These can happen together, since a squeezed Apple might raise prices, which could slow device sales, which in turn weakens Services. I can’t put honest odds on the second and third. I can show what the first one does, and I do below.

What the price implies

On consensus, Apple earns $9.60 a share in fiscal 2027, up 10% from $8.75.²³ My base gives $9.03 on today’s share count and about $9.16 on a count that shrinks 1.5% a year, so I am 5% below consensus E (estimate). That gap is the memory squeeze and AI spending, which I weigh heavily. The rest of the disagreement with the price is about the ten years after, and about what someone pays in 2036.

Over ten years the price rose 12.1 times while earnings per share rose 4.2 times C (calculated).¹⁶ Of the 12.1, the market paying more for each dollar of profit accounts for 2.9 times. That is the part that can’t repeat from here without the multiple going above 40. It is also the part that explains why a decade of excellent profit growth did not look as good in the last two years: the price moved ahead of earnings.

Reverse DCF: what the price needs

Held at my base margins, the price needs net sales to grow about 15% a year for ten years, to about $1.9T in fiscal 2036, against $0.48T this year C (calculated). The last nine years gave 7.6%, with a pandemic cycle in the middle.⁶ With my bull margins it needs about 13% C (calculated). Fifteen percent for ten years is not impossible in principle. In the nine years of growth rates I have, Apple has had two above that, fiscal 2021 at 33% and fiscal 2018 at 16% C (calculated).⁶

There are two other ways to read the same price. One is that it equals my base cash flows at a discount rate of about 6.3% C (calculated). That is the return the price offers if my base is right, 1.4 points above the Treasury. The other is to ask what a buyer in 2036 pays. If my base earnings come through and the market still pays 40 times them in 2036, the return from $344.63 is about 7.5% a year E (estimate). That is the real bet: not a faster business, but a multiple that never comes down.

The case for the market being right

The price reflects a debate, and the other side has an argument. It has three parts. First, the rate. Apple’s cash flows are among the steadiest on this site, and at 6.3% my base cash flows equal the price. A buyer who needs only 6.3% is not being irrational; my 9% is a choice, and the three-year beta, at 0.76, supports 8.1%, which gives $213 C (calculated).²⁴²²²⁵ Second, growth. Analysts expect earnings per share to rise about 13% a year from $8.75 to $12.59 by fiscal 2029.²³ If that pace held for ten years, earnings would be about $29 a share, and at 25 times that is $734, a return of about 7.9% a year plus a dividend yield near 0.3% E (estimate). That is close to my required rate. Third, the platform. More than 2.5 billion devices are a way to sell AI features and services that I can’t price, and I give them no value beyond the services growth in my cases.⁵ My answer to all three is that the price already pays for them, but it is a judgement, not a measurement.

The discount rate

I discount at 9.0%, the rate on my Microsoft page. Alphabet is at 10.0%. The capital asset pricing model, with the 4.94% ten-year Treasury and Damodaran’s 4.20% implied equity risk premium, gives 8.1% on Apple’s beta against the market over 36 months, 9.4% over 60 and 9.7% over 119 C (calculated).²²²⁵²⁴ I take 9.0% because it sits at the average of the three, 9.1%, and matches the page I wrote for Microsoft, whose ten-year beta is 0.97. The ten-year reading is higher than my choice, and the table shows what a higher rate does.

No row of the table reaches the price. At 9.5% the base is $165 and at 10% $152 E (estimate).

Three cases

My base case is Apple carrying on much as it has, a bit slower. Products grow 5.0% in fiscal 2027, on the iPhone cycle and the price rises, and then about 3% a year. Services grow 11% in 2027 and fade to 6.5%. The products gross margin steps down to 37% in 2027 from about 38% this year on memory costs and settles at 36.5%. Services margin slips from about 76.1% this year to 74.5%, because I assume some of the App Store and search income erodes E (estimate). That gives revenue of $770B and operating income of $264B in fiscal 2036, about 5.6% a year against 9.2% in the last nine years C (calculated). The operating margin rises to 34.3% from 32.2% because services reach 37% of sales.

The bull case has Apple win more of what it is already winning: products growing 8.5% in 2027 and 4% by 2036, services growth of 14% fading to 8.5%, a products margin near 38.5% and a services margin of 77%. That is 7.2% sales growth for ten years from a much larger base, and an operating margin of 37%. The bear case has memory and tariffs bite, the iPhone cycle fade into falling product sales, China slip and the courts take a chunk of the services profit: services growth slows to 7% and then 4%, the services margin falls to 71% and the operating margin to 30%. Even the bear is a very profitable business. It is only a poor purchase at $344.63.

What I can’t judge

Some inputs I can’t know, and I would rather say so than pretend. I can’t tell how long the memory shortage lasts; the 10-Q expects it to intensify, and nothing in the filings says when it turns.² I can’t tell how the Supreme Court will rule on the contempt standard in Epic, or whether the Google appeal reopens the payment.² I can’t tell what the licensed rights bought in 2026 are for, or what Apple will pay for them in later years. I can’t tell what Apple’s artificial intelligence strategy costs at full scale: Cook said on the July call that he did not want to claim a complete plan for computing costs, and that AI spending was growing in both operating expense and cost of sales.⁵ And I can’t know what multiple the market will pay in 2036. That last one matters more than all the others together, which is why I show the returns at several multiples instead of picking one.

Multiply it out

My base needs several things at once. Services growth of 9% or more a year for five years: I would give that 55%. iPhone and Mac revenue growing about 3.5% a year: 60%. A services gross margin held near 75% despite the court rulings: 65%. A products gross margin held near 36.5% through the memory shortage: 60%. No lasting loss in Greater China: 80%. Multiplied, that is about 10% E (estimate). They overlap, since a strong iPhone cycle helps services and the margins, so the true figure is higher, perhaps a fifth. And missing one doesn’t ruin the case. But the price needs more than my base. It needs my base and a multiple near today’s in 2036. If I give that a one-in-five chance, the joint odds are about 2% before overlap E (estimate). The argument for the price is not simple. It needs many things to go right together.

Quality and price are two questions

How good is the business? Among the best ever built: a base of more than 2.5 billion devices, a gross margin near 47%, owner cash of about $127B a year and net cash that no rival matches. How good is the price? Poor, on my numbers. At $344.63 the price is above my bull case, and the stock closed at its highest level in the ten years of prices I have on 29 October 2026.¹ A great business can be a poor purchase at the wrong price, and on this evidence that is what this looks like.

It helps to compare it with the next best use of the same dollar. A ten-year Treasury pays 4.94% with no risk to the dollars promised.²² Apple returns about 6.3% a year on my base, 1.4 points more than the Treasury, with all the risk of a share, and 2.7 points below the 9% I require. On this site the price is 93% above my base value for Apple, 91% for Costco, 44% for Microsoft and 43% for Alphabet. Of those, Microsoft is the best comparable, because it also sells a platform with high margins to a huge base. Its premium is under half of Apple’s. Neither is cheap on my numbers, and Apple is the least cheap.

The margin of safety

My margin-of-safety price is $125, 30% below my base value. At $125 my bear case returns about 7.6% a year, and the base about 11.4% C (calculated). A 30% cushion is what I use for a business this predictable. It would be wider for one I understood less well. The price is 176% above it.

Where this could be in 5 and 10 years

The value at a later date is what the owner cash after that date is worth then, at my 9% rate, per today’s 14,594M shares and in that year’s dollars. It leaves out the $49B of net cash I add today and the cash paid out before the date.

In the base, value rises from $179 to $213 in five years and $248 in ten, and the company pays out $116 a share of owner cash on the way. Four numbers carry that: sales growth of 4.9% a year, an operating margin that climbs to 34% as services grow, capital needs of about 3.5% of sales, and the 9% rate. At five years it needs Services to keep growing near 10% and the products margin to hold near 36.5%. At ten it needs Apple to still earn this margin in 2036 when the courts have had a decade to act. I can see the next five years better than the five after.

Bear and bull differ most in growth and margin. In the bear, sales barely grow and margin falls to 30%, so value stays near $105 and $113. In the bull, services grow in double digits for seven years and margin reaches 37%. The return column is low because the price today is 93% above my base: if the market in 2036 values Apple at my 9% rate, a holder earns about 0.6% a year in my base.

The most likely way the base is wrong is the services margin. A ruling that took the Google payment or cut the App Store commission would lower every value here; today’s base falls by about $9 to $17 a share for each $10B to $20B of annual profit lost E (estimate). These are estimates of value, not forecasts of the share price. The market can sit above or below value for years.

VIIThe case for the price

What most people miss

Apple’s suppliers fund its stock. Accounts payable were $64.5B at 27 June 2026 against inventories of $11.1B, 5.8 times as much C (calculated).² Apple collects from customers before it pays for the parts, so growth releases cash instead of using it. The second fact is the quarter that China delivered. Greater China sales rose 30% in the nine months to $64.8B after three years of decline, and the company says Mac had an all-time revenue record there.²⁵ The third is the cash. Operating cash flow in the nine months was $117.0B against $81.8B a year before, a rise of 43%, and net cash grew about $28B while Apple bought back $62.1B of its own stock.²

The bull case

The case is that Apple turns its installed base into an AI platform. Siri AI was introduced at the developer conference in the June quarter, and Apple says its next foundation models will be based on Google’s Gemini, running on Apple’s devices and private servers.²¹⁶ Cook said on the July call that Apple is “off the charts excited” about it, and that installed base, iPhone, Mac and Services all set records.⁵¹⁵ Analysts’ average forecast is earnings per share of $9.60 in fiscal 2027, $10.96 in 2028 and $12.59 in 2029, a rise of 13% a year from $8.75.²³ Berkshire Hathaway’s 13F shows 227,917,808 shares at both 31 March and 30 June 2026, so it neither added nor sold in that quarter.²⁰ I found no verified public statement in 2026 from a named outside investor on Apple that I could cite, so I cite none.

Against reality

Much of the bull case is true today. The base is large and growing, the quarter was the best June quarter on record, and China is back.¹⁵² My bull case gives it all of that: products growing 8.5% in 2027, services in double digits to 2033, and a margin of 37% by fiscal 2036. That is worth $253 E (estimate). If Apple does all that and hits its targets, it is a much bigger business in 2036 and it still falls short of $344.63 at 9%. The price needs net sales growth of about 15% a year at my base margins, against 7.6% over the last nine years, or a buyer in 2036 who pays above today’s 40 times earnings.

So the honest bull case at this price is about the multiple as much as the business: it went from 13.7 to 39.8 times earnings in ten years C (calculated).¹⁶ That can continue. Nothing in the filings says it will.

What a 10x would take. Ten times $344.63 is $3,446 a share, about $50.3T of market value on today’s share count C (calculated). At 30 times earnings that needs $1.68T of net income, 13 times the $127B of the last twelve months. At today’s 27% net margin it needs about $6.1T of revenue, 13 times current sales C (calculated).²³ With no rise in the multiple, earnings would have to grow about 29% a year for ten years, against 15% for earnings per share in the last nine, helped by a shrinking share count C (calculated).⁶ Apple rose 12.1 times in ten years, from about $0.6T C (calculated).¹⁶ A tenfold move from five trillion dollars is not plausible, and I build no case on it.

Where I come down

The cash and the suppliers moved me: Apple needs very little capital to grow, which is why my base has a return on new capital of 30% for ever. The bull case didn’t move me on value, because even my bull falls short of the price. What would raise my values most is services growth holding in double digits for five years while the legal fights end in Apple’s favour, which would show the profit pool is safer than I assume.

VIIIRisks, and what would change my view

The three ways this goes badly wrong are at the start of Valuation. These are the rest.

• Succession. Ternus has been CEO since 1 September 2026, and Cook stays as Executive Chair with a $45M target award.⁸ The 10-K says the company depends on key personnel.³ A split between a chair and a CEO who disagree on pace would show up in product timing.
• Where it is made. Substantially all hardware is built by outsourcing partners in China mainland, India, Japan, South Korea, Taiwan and Vietnam.³ New US tariffs under Section 301 are a live risk, and so is any disruption in the region.²
• Components and computing. The 10-Q says the memory shortage may intensify and that Apple may be unable to buy enough computing power on commercial terms for its AI features.²
• The EU. The Commission found preliminary breaches in a second Digital Markets Act case and can fine up to 10% of annual worldwide net sales, which on fiscal 2025 sales would be $41.6B C (calculated).²³ Apple has said interoperability rules may stop it launching features such as Siri AI in some places.²
• Antitrust in the US. The Justice Department says Apple monopolised smartphone markets and asks for equitable relief.² I can’t size that remedy.
• Dependence on Google. It pays Apple, and Apple’s next foundation models will be based on its Gemini models.⁷¹⁶ A dispute or a ruling affects both flows.
• My own assumptions. I assume no benefit from the float in payables. I assume 17.5% tax. And I assume the licensed rights are a cost like any other. If they are far larger, my standstill capital is too low.

What would raise my base value: the fiscal 2026 10-K and the first quarter of fiscal 2027 showing services growth of 10% or more, a products gross margin above 37% without refunds, and a favourable ruling in the Supreme Court. What would lower it: services growth below 7%, a products margin falling toward 35%, or a ban on the Google payment.²

Geopolitical risk, and what it means for the value

Apple owns no factories. The 10-K says a significant majority of manufacturing is done by partners in China mainland, India, Japan, South Korea, Taiwan and Vietnam, that many parts have one source, and that global supply chains can be highly concentrated.³ Greater China, which includes Hong Kong and Taiwan, was $64.4B of fiscal 2025 sales and about $79.3B in the twelve months to June 2026, 17% of the total C (calculated).³² The narrowest point is the most advanced chip. Apple’s filings do not name its chip maker. Cook said on the July call that Apple sourced over 100 million components from the Arizona fab this year, part of a $600B US commitment, and that supply of advanced nodes limited Mac.⁵ I take the Arizona fab to be TSMC’s. I found no filing saying how much leading-edge silicon is made outside Taiwan, so I assume most of it still is. I also found no filing that gives India’s share of assembly.

Washington sets tariffs: the 10-Q lists Section 301 duties now in force, a Section 232 semiconductor review whose January results added no tariff on Apple’s products, and the Supreme Court ruling of 20 February 2026 that struck down the emergency-powers tariffs.² Beijing controls access to its customers, and the 10-K lists rare earths among the inputs that trade restrictions can cut off.³ Nobody I can name decides Taiwan. Export controls and sanctions matter less, because Apple sells consumer devices, but the 10-K says restrictive measures can force it to change suppliers or stop offering products in a country.³

My odds are judgement. Some tariff stays: better than even, since the legal basis has changed once and the duties came back under another. Greater China sales halve: about one in ten over a decade. They fell 13% from the 2022 peak to 2025, not half, and rose 30% in the nine months to June 2026.³² A year without advanced chips from Taiwan: one in twenty to one in ten, with no base rate behind it.

None of these is in my base, which strips out tariff refunds and assumes no lasting loss in Greater China. I see no sign the price holds them either. It sits $92 above my bull value, so no row above is needed to explain the gap, and none closes it. Even the two-year Taiwan case leaves a base of $138, but it counts only Apple’s own lost sales. A war would also hit every other company and raise the discount rate, which I allow for only in part.