The Steward

In December 2008 Brian Moynihan finished designing his own job out of existence. "I designed my own job elimination and I eliminated my job and I was on my way out the door," he said later, and the departure press release was drafted and ready to send.¹ Bank of America offered him a different job instead, and shortly after the financial crisis he was running the company.¹

Nothing in that sequence is a founding. There is no product, no first customer, no stake taken at risk, no idea that had to be argued for. The Steward is the leader whose claim on the seat rests on having carried an institution through a crisis rather than on having created anything, who works by refusal rather than by invention, holds his own words constant on purpose, and states a time horizon that runs past his tenure with other people standing inside it. Every clause of that is checkable against what a person says out loud, and the last clause is the one that excludes.

The authority is the wreck, not the invention

Asked how he handles a sitting president publicly attacking his bank for refusing him as a customer, Moynihan answers with a founding date that is not his own. "It's part of the job. It's part of what we do... this company's been around since 1784. Our job is to work with every administration, make them successful because in America's successful, we're successful."¹ He then recites the institution's works the way another executive would recite a resume: the funding of the Erie Canal, and the federal government borrowing from banks to rebuild the White House after it was destroyed.¹ A founder under attack points at what he built. The steward points at what the building has already outlasted, and the pronoun he uses for all of it is "we."

The crisis question gets the same treatment. What he takes from the financial crisis is not a strategy but a demonstration, and the proof he offers is that nothing broke: capital and liquidity, businesses that offset each other, "but I also say how you have to have a team that in times of stress actually outperforms and that's what we showed in the stresses since."¹ That is the plate's foundation and its strangest feature, because no other plate in this catalog rests a person's legitimacy on survival. The platform consolidator earns his by finding the broken five percent nobody else priced. The wave recognizer earns his by committing everything at the one moment his skills fit. The necessity-forged entrepreneur earns his by turning each business's problems into the next business. Every one of them made a thing that did not exist. The steward's claim is that he was in the chair and it is still standing.

Strategy is a list of refusals

Ask a founder for the strategy and the answer is a list of things built. Ask Moynihan and the answer is a list of things declined: capital, liquidity, and "the range of activities we have, keeping it narrow, staying out of certain business which are inconsistent with the biggest consumer brand and financial services in the country."¹ Pressed for a concrete case, he gives the sharpest one a bank can give. "We don't do subprime lending and it's not because other people can't do it. It's just it's inconsistent with us being able to maintain the customer relationships at the level we have to maintain them."¹

The reason matters more than the refusal. He does not say the business is unprofitable or that the bank could not run it. He says it is inconsistent with something the bank already is, which means the strategy is not a plan for what the institution should become. It is a fence around what it must not stop being. Subtraction of that kind is a positive act, and it is only available to someone who inherited something already worth protecting.

The same words, on purpose

Asked which phrase he uses too often, he does not deny the charge. He explains the method. "The ones I use too often I have to use too often because I'm trying to get a message across... you have to be consistent. CEOs who make mistakes change their words in a way that may not be as meaningful as they think but then people say we're doing something different."¹

This is the plate's most transferable discipline and its least intuitive one. For most leaders vocabulary is expressive, and a fresh formulation is a sign of a live mind. For the steward vocabulary is load-bearing, and a fresh formulation is an unplanned announcement: the change gets read downstream as a change of direction he never authorized. The reader is directed to the file on repetition doesn't spoil the prayer, which argues the identical discipline from the opposite end. There a founder repeats a new message roughly twenty times until a market finally learns it. Here an incumbent repeats an old one so that nobody reads a course correction into a synonym. Same mechanic, opposite direction of travel, and the reversal is what makes the pair worth holding together.

Nobody's client

The independence doctrine is stated flatly, and it is aimed at the party a bank has the most obvious reason to court. "We shouldn't be dependent on what the regulators think about our company, what they're rating it. We have to be able to run our company no matter if they love us or don't like us."¹

The stress test follows from it. Most institutions treat a hostile examination as a tax to be minimized. Moynihan treats it as the only evidence worth having, and for exactly the reason others resent it: the examiner is not on his side. "When you give this information to somebody who doesn't really have any interest in making you look good. The Federal Reserve when they run the stress test they're tough tests. They want to make you look bad."¹ A founder's instinct is to control the narrative until it can be released on his own terms. The steward's is to hand it to a party that wants it to fail, and to count the survival of the handling as the finding.

The test is a date

Four disciplines in, nothing has excluded anybody. Plenty of founders refuse businesses, repeat themselves on purpose, and prefer an adversarial audit to a friendly one. A plate that cannot turn anyone away is not a plate. The Steward has exactly one disqualifying test, and it is a date: the steward states a time horizon that runs past his own tenure, and puts other people inside it, unprompted and on the record.

Moynihan states it without being asked. His first obligation, he says, is not strategy but talent, and the reason is a year he does not expect to be present for: the company needs people who could run it not only next year "if I got hit by the proverbial truck" but "for decades ahead because somewhere in this company are the people going to run it," a horizon he runs through 2035 and beyond.¹ The exit trigger is set on the same clock and refuses every conventional stopping rule. Not age, not tenure, not results. "If you don't thrill at this job every day, you don't really try to be curious and drive it, you don't try to maintain your ability to let people move and prosper and drive it. If you're blocking, that's when you got to go."¹ Every one of those conditions is a description of the seat rather than of the man in it.

Now run the test across the catalog and watch it refuse people. The fanatical owner-operator fails in the definition itself, which turns on a founder who will not franchise out or sell because the company is his to run: the horizon there terminates at the man, deliberately, and that is the plate's point rather than a defect in it. Survival alone does not qualify anyone either. A founder who carried a company through a near-death year and has never said a word about who holds the seat afterward has been through a crisis without becoming a steward, because the plate is not a prize for endurance. It is a claim about where a person puts the end of the story. Most founders put themselves there, and they are not wrong to; it is what makes them founders.

A plate for a man who founded nothing

The uncomfortable part should be said plainly rather than filed under a caveat. This is a taxonomy of founders, and it has just opened a plate for someone who founded nothing. Moynihan's own record here is filed as an operator, and its second paragraph calls him a steward rather than a founder. The catalog holds a few others in that position: Steve Ballmer is filed as an operator and carries no plate at all, which the taxonomy's governing rule permits, since no specimen is owed one and an unclassified record is an honest record.

So the question is why this one gets a plate when nobody is owed one. Not as a courtesy to a man who missed out on founding. The plate exists because the misfit located a boundary drawn in the wrong place. A taxonomy that can classify every way of starting a company and none of the ways of holding one is describing the first act and calling it the play. The steward is what a founder's company turns into once the founder is gone, and a catalog that cannot classify that cannot say anything about whether any of the companies in it will outlive the people who made them, which is the only question that eventually gets asked about all of them.

The cost of admitting it should stay visible. The collection is now, slightly, about people who hold institutions and not only about people who start them, and the title is that much less accurate than it was. That is the better trade. The alternative was the one this taxonomy has already had to undo once: stretching a founder's plate over a man it does not fit, and doing the stretching inside his own prose.

The nearest plate, and the one question that separates them

The closest neighbor is Coach Over Player, close enough that Moynihan was filed there until this plate existed. Both men subtract themselves on purpose. Both keep "am I still the right person for this" open as a live question rather than a rhetorical one. They diverge on what gets subtracted.

Daniel Ek's audit returns a position. Told his presence in product reviews was costing his team output, he left product and never went back, and his verdict on himself is a job description: "in a way, I may be a better coach than I am a player."² The coach hands away roles and keeps the seat, and the answer the audit comes back with is always another place on the field, because the residual zone is the whole reason the plate exists. Moynihan's audit returns a vacancy. What he is developing people to take is not a function beneath him. It is his chair.

One question separates them cleanly: when the subtraction is finished, is he still in the building? For the coach the answer is yes, and the plate depends on it. For the steward the answer is no, and the plate depends on that instead.

The audit that closes after he leaves

The shadow is easy to name and nearly impossible to catch in time. A horizon stated is not a horizon honored, and "developing the people who will run it" is also the most graceful available way of never naming one of them. Moynihan keeps succession out of public view, and the objection is procedural rather than evasive: he lived through the alternative, when "CNBC was running who's going to get it for three months."¹ The reason is a good one. The effect is still that the plate's central claim cannot be checked from outside until the day it is settled. That is the right note for this archetype to end on, because the steward's whole case is a promise about a room he will not be in, and the only audit that closes it is conducted after he has left it.