Apple’s CEO Transition: From Tim Cook’s Supply‑Chain Mastery to John Ternus’s Engineering‑Led Growth
Tim Cook's final day as Apple CEO is today. On September 1, John Ternus—a 25-year hardware engineering veteran who oversaw the iPad, AirPods and the Apple Silicon Mac transition—takes the title. The stock, around $314 and down roughly 1.8%, is pricing the succession as a non-event. Investors should look harder.
Apple simultaneously promoted Johny Srouji to Chief Hardware Officer, consolidating silicon, batteries, cameras, modems, displays and sensors under one executive. Sabih Khan, a three-decade Apple insider, stays as COO and keeps running procurement, manufacturing, logistics and fulfillment. Cook will serve as executive chairman. Phil Schiller, who ran the App Store and events, is stepping away from those roles the same week; Eddy Cue's Services organization absorbs App Store oversight.
Two institutional chapters—Cook's supply-chain architecture and Schiller's App Store empire—close simultaneously. An engineer now sits above a dedicated chip chief and a dedicated operations chief.
The R&D Numbers Have Already Decided
The spending acceleration that might define the Ternus era started before he got the job:
| Metric | Value |
|---|---|
| R&D, first 9 months of FY2026 | $34.04 billion (up 33% YoY) |
| R&D, full FY2025 | $34.55 billion |
| Fiscal Q3 R&D | $11.73 billion (~11% of quarterly revenue) |
| 9-month R&D as % of sales | 9.34% |
Given Apple's September-quarter revenue guidance, the company needs only about $9 billion of Q4 R&D to finish the year above 9%. It spent $11.7 billion last quarter. The ratio is effectively locked. A more demanding test:
- Does R&D stay above 9% of revenue in fiscal 2027, when revenue should still be growing?
- Does annual R&D approach $50 billion?
Those would confirm that the higher engineering intensity is a permanent regime under Ternus, rather than a pre-transition catch-up.
Services Still Grow—But Regulators Are Capturing the Upside
Apple's Services segment produced $30.74 billion in Q3, up 12%, with a gross margin of 75.6%. Saturation is the wrong diagnosis. The problem is that regulators are compressing the economics. Under Apple's new EU terms, effective October 1, standard App Store commissions fall as low as 15% for link-out purchases and 5% for certain transactions distributed outside the store.
The arithmetic bites because Services dollars are so much more profitable:
- Each dollar of lost Services revenue destroys about $0.76 of gross profit.
- Replacing that with product revenue at a 40.1% product gross margin requires roughly $1.89 in incremental hardware sales.
New devices cannot merely sell; they need large average selling prices, high attach rates, or IP that amortizes across hundreds of millions of units.
Why Procurement Alone Stopped Working
Cook, the world's most celebrated supply-chain CEO, conceded during his final earnings call that Apple faces "very significant" supply constraints with limited flexibility to fix them. AI data-center construction is swallowing advanced semiconductor and memory capacity faster than any buyer—Apple included—can offset by negotiating harder.
That admission is a precise transition marker:
| Cook era | Ternus era |
|---|---|
| Competitive advantage: procure scarce capacity better than anyone | Emerging advantage: engineer around scarcity |
| Differentiating factor: procurement | Differentiating factor: chips, modems, sensors that extract more performance per transistor, per watt, per dollar |
Mac revenue—up 29% year-on-year in Q3—is the proof case. Apple Silicon simultaneously improved battery life and performance, gave Apple leverage over Intel, and re-energized an entire product line. That playbook, applied to new form factors (a foldable iPhone expected at the September 9 event, reported smart glasses, camera-equipped AirPods, home robotics prototypes), is the template.
Apple Is Running an AI Capital-Intensity Arbitrage
Buried inside Apple's financials is a divergence that very few investors are discussing:
| First 9 months FY2026 | |
|---|---|
| R&D spending | +33% (to $34.04B) |
| Capex (PP&E) | -28% (from $9.47B to $6.80B) |
Apple is becoming dramatically more technology-intensive without becoming proportionately more asset-intensive. That is the opposite of what Microsoft, Meta and Alphabet are doing.
The mechanism: Apple's next-generation foundation models will be based on Google Gemini models and cloud technology, executed across Apple devices and Private Cloud Compute. Apple is renting frontier-model capability—something that may commoditize rapidly as providers multiply—while retaining ownership of the layers that are genuinely hard to replicate:
- Custom silicon
- OS integration
- Privacy architecture
- A global hardware distribution network
- A direct billing relationship with over a billion customers
| If model providers remain substitutable commodities | If a single provider becomes non-substitutable |
|---|---|
| Apple gets improving AI without stranding $100B in proprietary training clusters | Apple has a problem—but that risk can be managed by keeping model-provider APIs loosely coupled with product logic |
Where Apple does commit capital: scarce manufacturing steps that constrain its product roadmap—Broadcom agreement (over $30 billion for custom silicon and wireless components), Amkor's $7 billion Apple-backed packaging facility, TSMC Arizona's advanced-node commitments.
The bet is that owning bottleneck access and chip IP compounds better than owning general-purpose GPU farms.
For a company valued at roughly 36 times trailing earnings on modest consensus growth, this approach carries real stakes. The margin question under Ternus comes down to whether rising R&D creates reusable platform technology—Apple Silicon's model—or funds expensive devices that remain niche. Executives and investors should track that ratio with more attention than they give the next keynote.
not investment advice