Business

Know the Business

Apple is a vertically-integrated premium-hardware OEM that owns the only iOS distribution rail, so a 2.4-billion device install base feeds a 75%-gross-margin services line — that is the entire equity story. The market correctly prices the consolidated economics at a ~26x EV/EBITDA premium, but is likely overestimating the resilience of the Services take-rate under DMA and US v. Google remedies, and underestimating the cyclical risk that Greater China and the iPhone Pro mix re-roll downward in any meaningful global consumer downturn. Read this report as: the moat is real, the multiple is fragile.

1. How This Business Actually Works

Apple runs two economic engines on one P&L. Hardware sells a device at a high ASP and earns a premium product gross margin (36.8% in FY2025) — strong but ordinary. The interesting engine is the Services rail, which sits on the device and earns 75.4% gross margin by clipping a fee on every app, ad, subscription, search query, cloud GB, and warranty plan. Hardware seeds the install base; the install base monetizes through Services. Neither half explains Apple alone.

Services is 26% of revenue and 42% of gross profit. Every incremental percentage point Services adds to the mix lifts the blended company gross margin by ~38 bps — that is the lever that has carried the multiple since 2018.

The cost structure runs negative working capital: customers and channels pay inside 60 days, suppliers wait ~115 days, inventory clears in ~10 days. The cash conversion cycle is -43 days, so suppliers effectively finance growth. Manufacturing purchase obligations of $56.2B at FY2025 year-end (with $55.4B payable within twelve months) are the cleanest live read on the iPhone build plan; they swing with the next product cycle.

2. The Playing Field

Apple sits in a peer set where no one else owns the rail and the brand and the silicon in one entity. MSFT and GOOGL each own a rail but no premium hardware franchise. META rents Apple's rail and pays the toll (ATT). AMZN runs a different model entirely (third-party marketplace + AWS). HPQ is the pure-hardware comparator and shows what consumer-electronics manufacturing looks like without a rail underneath.

The bubble chart shows the actual ranking the market makes. MSFT prints a higher operating margin (45.6%) at a lower EV/EBITDA (22.5x) than Apple (32.0% / 26.3x). META is even cheaper (16.4x EV/EBITDA at 41% operating margin). Apple's multiple premium over MSFT is not paid for higher current profitability — it is paid for predictability of capital return and the cash-conversion-cycle/buyback flywheel. That premium narrows when Services growth slows or when iPhone units come into doubt.

HPQ is the diagnostic. Same end customer, similar manufacturing footprint, no rail: 5.7% operating margin, 0.6x EV/Sales, 11% FCF yield. The gap between AAPL (9.1x EV/Sales) and HPQ (0.6x) is the moat tax the market charges for owning iOS, Apple Silicon, the App Store, and the Apple brand on one P&L. If iOS gatekeeper economics get re-priced by regulators, Apple's multiple moves toward MSFT's, not HPQ's — but the move would still be a meaningful de-rating.

3. Is This Business Cyclical?

Yes — but the cycle hits mix and gross margin first, then unit volumes. Revenue has only fallen twice since the iPhone era began (FY2016 post-China slowdown, FY2019 trade-war + 5G air-pocket, plus a flat FY2023). Operating margin is much more responsive: FX, memory pricing, channel inventory, and tariff drag all run through gross margin within two quarters.

Two real cycles are visible. FY2013: iPhone 5 cycle bridge, gross margin fell 630 bps year-over-year as the company digested cost inflation on a refresh year. FY2016-FY2019: gross margin slid from 40% to 38% as iPhone unit growth stalled and FX pressured international. The post-COVID Services tailwind plus a richer Pro mix has steered margin to a 16-year high; that is where the current multiple is anchored.

The cycle ahead is already turning back up. Q2 FY2026 (March quarter) printed $111.2B revenue +17% YoY with Greater China at $20.5B vs $16.0B prior year, ending a two-year Greater China decline. iPhone alone was $57.0B, Services $31.0B (all-time record), and operating cash flow $28B in one quarter. A young analyst should treat that as a normal post-cycle recovery, not a permanent re-rating — replacement timing, AI-on-device prompts, and FX have all moved the right direction at once.

4. The Metrics That Actually Matter

Skip P/E, ROE, and EPS-print compare — the AAPL P/E and ROE are distorted by buyback math, and reported EPS includes a one-time Irish tax true-up in FY2024. Five metrics drive the equity, in order.

The most important single chart in this report. Services gross margin has expanded 940 bps in five years, doing nearly all the work of lifting blended company gross margin. Any regulatory ceiling on App Store take-rate, search default payment, or steering rules shows up here first. If Services GM rolls back toward 70%, the bull case loses ~3% on blended company gross margin and a couple of turns on EV/EBITDA.

5. What Is This Business Worth?

Apple is best underwritten as one economic engine, not a sum-of-the-parts. Services has no separable share class, no separate management, and no separable distribution; it exists only because the device install base exists. SOTP analyses that put a software multiple on Services and a hardware multiple on Products produce a "fair value" that double-counts what already sits inside consolidated EV/EBITDA. The right lens is price-to-FCF per share through the cycle, recognising that ~85% of FCF gets recycled through buybacks.

What actually moves intrinsic value:

Apple has returned $770B+ since FY2013 and reduced diluted share count from 26.5B to 15.0B (a 43% reduction). At the current ~2.6% FCF yield and ~85% payout ratio, the buyback alone delivers ~2.2% annual share-count compounding — which means a flat-FCF Apple still grows FCF-per-share at low-single-digits without help from operations. That math is what insulates the multiple in flat years.

6. What I'd Tell a Young Analyst

Watch Services gross margin, not Services revenue. Services revenue growth can be sustained even as take-rate compresses (install base growth offsets) — but margin tells you whether the rail is still earning rent. If Services GM stalls or contracts, the bull case is broken before the revenue line shows damage.

Watch Greater China iPhone, not total Greater China. The segment table blends Mac (helped by Apple Silicon momentum in China) and iPhone (the real signal). Listen to the call commentary, not the line item — Q2 FY26 confirmed iPhone leadership in the region, and that single fact unwound the bear thesis from FY2024.

The buyback is doing the heavy lifting on EPS — but only while FCF stays at $100B+. Modeling the FY26-FY28 EPS line without modeling buyback pace independently of FCF will produce systematically wrong answers. Build the path with explicit share-count assumptions.

The regulatory tail is the only thing on this story that can hurt the multiple cleanly. Cycles compress and recover. Tariff costs get repriced into the next iPhone. A binding US-Google search remedy or a DMA-style App Store rollout to the US is a one-way de-rate that operating performance cannot defend against. Sizing that probability is the single highest-leverage analytical question on Apple today.

One contrarian read. The market treats the consolidated 32% operating margin as the moat. The moat is actually the negative cash conversion cycle, 75% Services GM, and the install-base flywheel — three engines that operate independently. As long as those three keep working, the multiple holds even if any one earnings line wobbles. When any two break at the same time (e.g., Services GM rolls + iPhone Pro mix rolls), the stock re-rates fast. Watch them together.