Apple After Cook: Ternus, $100bn in Buybacks, and the AI Question

Khan Capitals | May 2026

---

Key Takeaways

• Apple announced that John Ternus will succeed Tim Cook as chief executive on 1 September 2026, with Cook taking the executive chairman role; this is the most significant leadership transition at Apple since 2011.
• Q2 FY26 results beat across every line: revenue $111.2 billion (+17 per cent), iPhone $58 billion (+22 per cent, a March-quarter record), Services $31 billion (+16 per cent), and EPS $2.01 (+22 per cent).
• The board authorised an additional $100 billion in share repurchases and raised the quarterly dividend by 4 per cent to 27 cents, extending the longest unbroken capital-return programme among the mega-caps.
• Apple has deliberately stayed out of the hyperscaler AI capex cycle, choosing partnerships and on-device inference over training infrastructure, and now spends a fraction of the combined $700 billion-plus that Microsoft, Alphabet, Amazon and Meta will deploy in 2026.
• The strategic question for Ternus is whether Apple’s discipline becomes a moat or a liability as AI experiences increasingly sit inside ecosystems controlled by the hyperscalers, with implications for both the multiple and the long-run terminal value.

Part of: The AI Infrastructure Supercycle – Khan Capital’s hub on the AI infrastructure build-out.

An Earnings Beat, a Succession, and a $100 Billion Buyback

Apple closed mega-cap earnings week with a result that, in any other quarter, would have dominated coverage. Revenue of $111.2 billion grew 17 per cent year on year, comfortably ahead of the $109.7 billion consensus. Diluted earnings per share of $2.01 rose 22 per cent and beat the $1.95 estimate. Net income reached $29.58 billion. The board authorised a further $100 billion in share repurchases and raised the dividend.

The succession announcement, however, was the more market-moving disclosure. Tim Cook, who has run Apple since August 2011, will hand over to John Ternus, the current Senior Vice President of Hardware Engineering, on 1 September 2026. Cook becomes executive chairman. The transition has been telegraphed for some time inside the analyst community, but the formal date and the choice of Ternus over other internal candidates close the speculation that has hung over the company for the past two years.

The Numbers Behind the Headline

The quarter’s segment performance is detailed below. iPhone unit demand was the headline driver, with the iPhone 17 lineup described by Cook as “the most popular lineup in our history”. Services growth of 16.3 per cent ran ahead of the 14 per cent rate consensus had assumed and represented the highest quarterly Services revenue in the company’s history. Mac revenue grew at a more modest single-digit rate, while Wearables, Home and Accessories was broadly flat.

Segment	Q2 FY26 Revenue	Y/Y change	Consensus
iPhone	$57.99bn	+22%	~$54bn
Services	$30.98bn	+16.3%	$30.4bn
Mac	~$8bn	Mid-single digit	~$8.0bn
Wearables, Home, Accessories	~$8.5bn	Flat	~$8.7bn
iPad	~$5.7bn	Low single digit	~$5.6bn
Total	$111.2bn	+17%	$109.7bn

The Services result deserves particular attention. At a $124 billion run rate, Services would, if reported separately, sit between the largest pure-play software companies. Operating margins on Services exceed 70 per cent, against a corporate average closer to 32 per cent. The Services trajectory has been the single most important driver of Apple’s gross margin expansion since 2018, and the Q2 print suggests the trajectory remains intact.

Why Ternus, and What It Means

The choice of Ternus over alternatives such as Jeff Williams (Chief Operating Officer) or Greg Joswiak (Senior Vice President of Worldwide Marketing) is internally consistent with Apple’s product-led culture. Ternus joined the company in 2001 and has run hardware engineering since 2021, overseeing the M-series Apple Silicon transition, the iPad family, the Vision Pro programme and, most recently, the in-house chip work that powers the company’s AI strategy.

Three implications follow. First, the succession reinforces the centrality of silicon design to Apple’s strategy: Ternus is the Apple Silicon executive, and his promotion suggests the next chapter of the company is built around continuing to extract performance and efficiency advantages from custom silicon rather than chasing scale in cloud infrastructure. Second, the choice over a more operationally focused candidate signals continuity in the product cadence rather than an inflection toward a different kind of company. Third, Cook’s continuing presence as executive chairman provides a stabilising bridge of the kind that helped Microsoft’s Bill Gates handover and IBM’s Lou Gerstner transitions in their respective eras.

For shareholders, the principal risk in any chief executive transition is strategic discontinuity. The Ternus profile makes that risk small. The principal opportunity is that a new chief executive is sometimes able to make decisions a long-tenured incumbent cannot, particularly around capital allocation and underperforming product lines. Whether Ternus exercises that latitude is the open question, and the next twelve months will provide the first signals.

The AI Capex Decision: Discipline or Mistake?

Apple’s relative absence from the hyperscaler AI capex narrative is, in 2026, the most strategically distinctive aspect of the company. Microsoft, Alphabet, Amazon and Meta will collectively spend more than $700 billion on AI infrastructure in 2026 alone. Apple’s total capex is forecast at roughly $14 billion. The company’s AI strategy rests on three pillars: distribution partnerships with frontier-model providers (including the existing OpenAI integration on iOS, with reports of additional partners under negotiation), on-device inference using Apple Silicon’s neural engines, and a relatively modest internal training programme.

The argument for the strategy is structural. Apple does not sell cloud compute; it sells devices and services. Building hyperscale training infrastructure to compete with Azure and Google Cloud would require entering a market where Apple has neither distribution advantages nor existing customer relationships. The depreciation drag visible at Microsoft and Meta would be uncovered by a corresponding revenue offset. The strategy also leverages Apple’s existing privacy positioning: on-device inference is differentiated, defensible and consistent with the brand.

The argument against the strategy is that AI experiences are increasingly the locus of platform value. If users come to expect frontier-model capabilities embedded into their daily applications, and those capabilities are dominated by hyperscaler-controlled stacks, Apple risks ceding control of the most important interaction layer on its devices. A version of this risk has occurred before: Apple’s Maps strategy, its absence from cloud productivity, and the relatively limited reach of its services outside its own ecosystem all reflect similar choices that have, in retrospect, capped optionality.

Capital Return: The Quietly Defining Number

The $100 billion buyback authorisation is, in absolute terms, larger than the total annual capex of any company outside the hyperscaler cohort. Apple’s cumulative buyback since 2012 now exceeds $750 billion. The dividend increase was modest at 4 per cent but represents the thirteenth consecutive annual rise. Together, the capital-return programme returned approximately $30 billion in the quarter alone.

The structural significance is that Apple’s capital allocation is the inverse of the hyperscalers. Where Microsoft and Alphabet are reinvesting record cash flows into infrastructure capex, Apple is returning record cash flows to shareholders. The valuation implication is straightforward: Apple offers a different free cash flow profile than the cohort, and that difference is becoming larger rather than smaller as the AI capex cycle compresses hyperscaler free cash flow yields.

Geographic and Tariff Exposure

Apple’s results were achieved despite an active tariff backdrop. The company has navigated the 2025-26 trade environment by accelerating the diversification of its assembly footprint into Vietnam and India and by absorbing modest tariff costs in gross margin. Greater China revenue grew at a high single-digit rate in Q2, with management commentary noting strong iPhone 17 reception. The relative resilience of the China business, against a context in which several Western consumer brands have reported soft demand, is itself a useful read on the iPhone’s durable position in the premium segment.

What the Market Is Underappreciating

The first underappreciated element is the operating leverage in Services. At current growth rates, Services will be approaching 30 per cent of revenue and a meaningful majority of operating profit by FY28. That mix shift, on its own, supports a re-rating to a multiple closer to the software peer group rather than the hardware peer group, even before any consideration of AI optionality.

The second is the asymmetry of the AI risk. Apple’s strategy is exposed to one large negative scenario (the hyperscaler-controlled AI experience capturing platform value) and several positive scenarios (on-device inference becoming the privacy default, partnerships providing optionality without capex risk, the Apple Silicon advantage extending into AI workloads). The market has tended to focus on the negative scenario, but the positive scenarios are not implausible.

The third is the succession alignment with the Apple Silicon era. Cook’s tenure was defined by the supply-chain and services platform build. Ternus’s profile is defined by silicon. If the next decade of value creation hinges on chip-level differentiation (custom AI accelerators, server silicon, on-device neural engines), the succession is unusually well-matched to the strategic direction.

Investor Implications

Equities. Apple offers a profile that is distinct from the hyperscaler cohort: positive free cash flow yield expansion against a backdrop in which Microsoft, Alphabet, Amazon and Meta are seeing free cash flow yield compression. Allocators may wish to consider whether their mega-cap equity exposure adequately reflects this divergence. The Services trajectory and the Ternus succession both reduce the probability of a negative re-rating event over the next twelve months.

Cross-asset. Apple’s capital return programme makes the stock function partly like a bond proxy in an allocator’s framework. With the ten-year Treasury yield in the mid-4 per cent area and Apple’s free cash flow yield around 3.5 per cent, the relative valuation is less attractive on yield alone, but the embedded growth optionality and buyback support remain meaningful. Investors balancing duration and equity exposure may wish to consider Apple as part of a quality-tilt sleeve rather than a pure technology bet.

Currency and trade. Apple’s foreign exchange and tariff sensitivity remain non-trivial. Continued dollar strength at the DXY 98 level is a modest headwind to translated revenue, while any escalation in US-China trade tension would weigh disproportionately on the iPhone gross margin. Investors with concentrated mega-cap exposure may wish to consider FX-hedged equity allocations as a complement.

A Visible Capital Return Story

Conclusion

Apple’s Q2 FY26 print is the cleanest set of fundamentals among the mega-caps and the most consequential leadership announcement in fifteen years. The Ternus succession is well-aligned with the strategic direction and the Services trajectory remains the most durable margin tailwind in the franchise. The deliberate decision to remain a comparatively small AI capex spender is a calculated bet on partnerships and on-device inference rather than infrastructure scale.

Whether the bet pays off depends on the trajectory of AI experiences over the next two to three years. If frontier-model capabilities concentrate inside hyperscaler stacks, Apple’s strategy will be tested in ways the current results do not yet reveal. If the inference layer disperses, with on-device and partnership models capturing meaningful share, Apple’s discipline will look prescient. The market is currently priced for the optimistic scenario, but the asymmetry of capital return and free cash flow keeps the risk-reward balanced rather than skewed.

Sources: Apple Q2 FY2026 earnings release; CNBC: Apple Q2 2026; Variety: Services revenue at record $31bn; Apple Q2 2026 earnings call transcript; AppleInsider preview.

Related Reading. For the wider AI capex context that frames Apple’s deliberate restraint, see the companion piece on the $725 billion hyperscaler capex cycle. For prior coverage of the broader equity backdrop and concentration dynamics, see Khan Capitals’ Q1 2026 market correction wrap, the SpaceX IPO filing piece, and the discussion of private credit and AI disruption, all of which sit downstream of how the hyperscaler-versus-non-hyperscaler split evolves. For the strategic backdrop to Apple’s position, see our analysis of the Apple Siri Gemini deal.

This piece is updated in light of Nvidia’s Q1 FY27 print and the moment when the bar becomes the beat.

For the fundamentals behind this story, start with the accounting behind the AI build-out and how companies return cash.