THREE EXITS: BREAKDOWN

Read time: 8 minutes

Apple — John Ternus and the Impossible Job

He has two options. One leads to slow decline. The other might get him fired. Congratulations on the promotion.

SITUATION

What the stock non-reaction actually told you

On April 20, 2026, Apple announced that Tim Cook steps sideways to become executive chairman, and John Ternus takes the CEO role from September 1. Fifteen years of Cook. The most successful operational run in corporate history. A $4 trillion machine built from a $350 billion starting point. Revenue from $108 billion to $416 billion. And when they announced his successor?

The stock moved less than 1%.

Most coverage read that as confidence. Orderly transition. Safe pair of hands. The institution is bigger than any individual. Which is one reading. Here is another: when you replace the chief executive of a four-trillion-dollar company and the market barely notices, you haven’t been told that the new CEO is trusted. You’ve been told that the CEO is not the variable. The system is the variable. And the system, as far as anyone could tell from the outside, had not changed.

Investors weren’t expressing confidence in Ternus. They were expressing confidence that nothing Ternus does will be allowed to disturb what the system currently produces.

That is a very different thing. Ternus, to his credit, probably knows it.

THE TRAP

What he’s walked into, and why both exits are blocked

Cook built something extraordinary. He also built it to a point where the straightforward growth is largely gone — and where the two obvious responses to that problem are both structurally unavailable.

iPhone revenue has plateaued at around $200 billion annually. Upgrade cycles are estimated to have stretched past 35 months — customers are keeping their phones longer, which is both a testament to how good the product is and a slow compression of the engine that drives everything else. The installed base sits at 2.5 billion devices. To put that another way: if every Apple device on earth were a person, they’d be the third-largest country in the world and almost all of them already have a phone. The uncaptured upside is not enormous.

Services were supposed to be the answer. Revenue north of $100 billion annually, gross margins around 47%, growth of 12% year-on-year — genuinely impressive. Except services growth depends on the installed base continuing to expand, and the installed base is a function of hardware, and hardware has plateaued. You can’t infinitely monetise a customer base that isn’t growing. And the App Store fee model that makes those 47% margins possible is under sustained regulatory pressure in multiple markets. If that gets meaningfully disrupted, the services story changes character quickly. The Vision Pro — Apple’s most visible attempt at a new category — was repositioned within a year of launch at $3,500 a unit.

So the ceiling is real. The question is which of the two obvious responses Ternus takes. And both of them are bad.

Path A: run it as Cook did. Protect the margins. Release premium iterations of existing lines. Keep the ecosystem tight. Let services compound. This is the path the board has effectively mandated. It is also the path that leads, with reasonable certainty, to growth compression. The market currently prices Apple as a premium-growth franchise. If growth quietly runs out — not catastrophically, just slowly, quarter by quarter — the multiple contracts. A company producing excellent results at 40% margins on flat revenue is a very good business. It is not a four-trillion-dollar business. The gap between those two things is where a very large amount of shareholder value currently lives, and it closes slowly and then quickly.

Path B: change the culture and try to build the next thing. Push into AI seriously. Ship experimental products. Tolerate public failure as the cost of category creation. This is the path that might actually address the ceiling. It also requires Apple to operate in ways that are structurally foreign to everything it currently rewards. Apple’s culture is built on secrecy, control, and late-stage perfectionism — not as an accident, but because that’s what produces products people pay a premium for. That is precisely the opposite of the speed, failure tolerance, and iterative public experimentation that building genuinely new categories requires. And even if Ternus wanted to do it, announcing that you’re going to sacrifice margin for speed and accept public product failure as part of the process would remove the foundation on which the current multiple rests. The stock would reprice before he’d shipped anything.

Don’t change and the ceiling arrives. Change and the market punishes you before you get there. Those are the two options. The board built both of them.

HOW IT WAS BUILT

Three governance decisions, made over nine months, that closed the exits before Ternus arrived

The board didn’t make this call in April. They made it in stages, starting in July 2025, and each decision foreclosed something before the next one was announced.

Gate 1 — July 2025

Jeff Williams, Cook’s COO and the most credible internal candidate for a different kind of succession, retires. Williams was the operational brain of the Cook era — supply chain, health initiatives, the disciplined scaling of the hardware model. His departure removed the one executive who might have balanced product ambition with genuine continuity, and the most likely internal advocate for a bolder succession process. The decision not to replace him from outside, or to ask what a post-Williams COO role might look like differently, was itself a decision. The board wasn’t exploring options. They were narrowing them.

Gate 2 — December 2025

Apple announces a broader leadership reshuffle. Kate Adams and Lisa Jackson retire. Jennifer Newstead joins from Meta as incoming general counsel. Roles redistribute through existing reporting lines. Read this as transformation and you miss it: this is personnel rotation inside a stable machine. The reporting lines don’t change. The capital allocation model doesn’t change. The system absorbs the new people without adjusting what it rewards them to protect. Nothing in December pointed toward a new strategic direction. Everything in December pointed toward continuity with different names on some of the doors.

Gate 3 — April 20, 2026

Three decisions inside one announcement. First: Cook becomes executive chairman — not non-executive, not retired. Actively involved in policy and supply chain diplomacy. The architect of the current system stays in the building with a formal role over the things that define the system. Second: Arthur Levinson shifts from non-executive chairman — a role he held for fifteen years — to lead independent director. The board architecture is preserved, not rearranged. Third: Ternus joins the board as an insider, not as a reform candidate arriving with an external mandate.

Individually, every one of those decisions is defensible. Collectively, they are a statement. The board has decided what it wants: the current system, run by someone who understands and respects it, with the person who built it available to advise on anything that might disturb it. That is not a setup for transformation.

It is a setup for preservation — and Ternus, 25 years inside the building, raised on the same rules, promoted because he fits them, is exactly the right person for that. Which is precisely the problem.

THE READ

What this means if you run a business — or if you’re the one being asked to lead someone else’s

1. If your growth has plateaued, check whether the model has hit its ceiling before you change the person running it.

Apple’s problem isn’t Ternus. It’s that the model that produced the growth has reached the limit of what it can produce. More of the same, executed better, won’t change the output — it will just deliver the compression more smoothly. If your revenue line has flattened and your explanation is execution, check whether the underlying model can actually produce more. Changing the operator doesn’t change the arithmetic.

2. Watch what a new leader is paid to do, not what they say they’ll do.

Cook’s fiscal 2025 compensation was $74.3 million, with a large portion tied to margin, revenue, and installed base growth. Those are the metrics of the current model. If Ternus’s compensation is structured the same way, he is being paid to run Path A whether he intends to or not. The incentive structure is the strategy. Everything else is commentary.

3. Understand the constraint before you announce the vision.

Ternus was not hired to transform Apple. He was hired because he fits the company as it currently exists. That is a legitimate board decision — continuity over disruption is a defensible call. But the constraint is real and structural. Until the board signals that it wants what transformation actually requires — lower near-term margins, public failure tolerance, cultural change that unsettles the operational model — the vision is decoration. If you’re taking the top job at someone else’s company, the first thing to understand is what the board actually wants, not what they say they want.

4. When the market prices you as a growth franchise, flat is the new down.

Apple at $4 trillion is priced on the assumption of sustained premium growth. The risk for Ternus isn’t that he fails spectacularly. It’s that he delivers something the market already suspected was coming but was hoping wouldn’t. He doesn’t need a disaster. He just needs a few years of quietly disappointing quarters and a valuation that prices in what was always there to see.

VERDICT

The board got exactly what they asked for. That’s the problem.

Ternus is not incompetent. The M-series chip transition. AirPods. The MacBook development programme. Twenty-five years of building things at Apple that worked. That’s not the question.

The question is whether any individual — regardless of capability — can produce a different outcome from inside a system that has been specifically designed to resist different outcomes. The board’s answer, delivered across nine months and crystallised on April 20, is: we’re not asking him to.

Path A runs until the ceiling arrives. Upgrade cycles past 35 months, a 2.5 billion device installed base, services under antitrust pressure, a valuation built on growth the model can no longer reliably produce — these are the current position, not the forecast. The compression is already running. The most likely outcome is a slow, well-managed version of Path A: excellent products, sustained profitability, growth that quietly disappoints a market priced for something more. That is not a disaster. It is a compression. And Ternus will be the person in charge when the compression becomes visible.

Path B requires the board to want something they have shown no sign of wanting, and the market to tolerate something the current valuation doesn’t price in.

The board picked the right person for the job they wanted done. The job they wanted done may not be the one the situation requires.

THREE EXITS · threeexits.com