◆ What the Market Isn't Pricing In

Before venturing an opinion on what a stock does not reflect, an honest man goes and reads the latest results, because a thesis that quarrels with the tape is just a prejudice in a good suit. So I went back to Apple's March-2026 quarter, announced on the last day of April, and it promptly embarrassed a draft of my own thinking. I had leaned on the tired line that this is a sleepy grower wearing a rich multiple, a business creeping ahead two or three percent a year while the crowd pays close to forty times earnings for it. The quarter says otherwise, and plainly. Revenue rose seventeen percent to $111 billion; iPhone jumped twenty-two percent to $57 billion on the iPhone 17 lineup; Services set an all-time high at $31 billion, up sixteen; and every geography grew double digits — Greater China, the perennial worry, up twenty-eight¹. Management then guided the June quarter to fourteen-to-seventeen percent growth². The June quarter then delivered sixteen percent, to $109.4 billion¹⁷. Whatever Apple is at this moment, it is not the stagnant annuity the bear leans on. What follows is not a price forecast — that turns on a hundred things a moat cannot see — but a note on where the tape and the story diverge.

It is worth being precise about that word 'sleepy,' because until recently the bear's caricature was largely fair, and the multi-year record is what proves it. Strip out the pandemic surge of fiscal 2021 and Apple's revenue barely moved for three years: $394 billion in fiscal 2022, $383 billion in 2023 — an outright decline — and $391 billion in 2024, with earnings per share stuck near $6.10 the whole way³. iPhone, the supposed engine, sat between roughly $200 and $210 billion a year across that entire stretch, going nowhere⁴. What quietly carried the company was Services, compounding at a low-teens pace from $78 billion in 2022 to $109 billion in 2025 and crossing a hundred billion for the first time⁵. Then the line broke upward — fiscal 2025 revenue rose about six percent, and the iPhone 17 drove the December and March quarters of fiscal 2026 to sixteen and seventeen⁶. So the reacceleration is not one lucky quarter laid over a steady grower; it is a genuine break from three flat years, which is exactly what makes it worth arguing about.

The valuation history is still the right place to start, and it still holds. At its 2016 low the stock fetched around ten or eleven times earnings⁷ — the true hardware multiple, a box-maker priced for the next cycle to disappoint. It now trades at several times that. What the crowd repriced was the recognition that Apple had become an installed-base annuity; the surprise in the latest numbers is that the annuity has, for now, started growing like something better than an annuity again. That reframes the real question. It is no longer 'does the market wrongly see a gadget-maker' — it plainly sees a compounding ecosystem — but 'is the market now extrapolating a supercycle?'

There is the genuine gap, and it cuts both ways. A twenty-two percent iPhone quarter off a single hot lineup, flattered by roughly two and a half points of currency⁸, is exactly the sort of figure that can be a peak as easily as a floor. If the iPhone 17 merely pulled upgrades forward and next year hands some back, then a full multiple laid over peak earnings is the classic way money gets lost in fine companies. But if this is the leading edge of an AI-driven replacement wave — buyers upgrading because the newest silicon is the price of admission to on-device intelligence — then the growth has legs the multiple has not fully banked. The market is holding that question open, not answering it, which is another way of saying the outcome is unpriced.

On one point the fresh detail flatly improved on what I had written. I had cast the App Store commission as a slowly eroding, regulator-besieged stream. The quarter cut the other way. Services gross margin runs near seventy-seven percent against thirty-nine for hardware⁹, and Apple is opening new toll booths rather than losing old ones: search ads across the App Store are now a stated growth driver, and this summer paid ads arrive inside Apple Maps in the U.S. and Canada¹⁰. Regulation did not so much as come up on the call. The long-run legal pressure on the commission is real and I would not retract it, but the near-term direction of travel is plainly more monetization of the installed base, not less — and that rich services mix is the quiet thing lifting the blended gross margin toward the high forties.

Two corrections in honesty's favor cut against the bull. First, margins are not a one-way street: management guided the June quarter down to 47.5-to-48.5 percent gross margin and flagged 'significantly higher memory costs' that worsen beyond it¹¹ — a dull, physical headwind the AI excitement tends to skate over. The June quarter printed 50.1 percent only because about two points of tariffs were refunded, and the September quarter is guided to 47-to-48 percent¹⁸. Second, I oversold the buyback. It is vast — the board just authorized another hundred billion¹², and the share count grinds steadily lower — but this quarter it was the earnings that did the work: net income rose nineteen percent while the count fell only about two¹³. Over a decade the buyback has retired about a third of the shares¹⁴ and remains the durable backstop; it is simply not the engine doing the lifting right now, and it pays to be precise about which is which.

Two items in this report are new enough that the multiple can scarcely have digested them. Research spending rose thirty-four percent year over year¹⁵ and is, in Cook's telling, accelerating faster than the company itself — a visible wager that the AI upgrade cycle is real, and a cost the income statement now carries. Notably, Apple is not going it alone: it leans on a collaboration with Google for foundational models while doing its own on-device work, which hedges the downside but dilutes the purest bull case, the one where Apple's private intelligence is something no rival can cheaply answer. The larger unpriced item is human. On the first of September Tim Cook became executive chairman and John Ternus, a hardware engineer, took the chief executive's chair¹⁶ — a handoff at the precise moment the story is turning on services, advertising and AI. Markets do not price a new hand on the tiller until they are made to.

So why claim anything is unpriced when the multiple has so obviously caught up to — and the business has, for a quarter, sprinted past — the story? Because a rich multiple resting on a supercycle is the most fragile kind: it has extrapolated the good quarter and has not yet met the full memory-cost headwind, a new chief executive's first year, or the year the iPhone gives some of 2026 back. The annuity is real, the advertising runway is real, the buyback is real — but the price now leans on three bets at once: that the reacceleration holds, that the margin holds, and that a new chief executive keeps the machine humming. The gap worth watching is no longer hardware-versus-services, a quarrel the tape has settled. It is whether a supercycle is being mistaken for a new baseline — and it is there, not at the consensus, that a patient owner earns their keep.