The fundamental problem the modern corporation solves is matching capital in the abstract—a number stored in a database that you can decrement in exchange for access to goods, services, and investments—with capital in the form of buildings, equipment, workers, brand names, tacit knowledge, etc. But as soon as you put this into practice, with one group of people supplying the capital and someone else providing the management, you have to align incentives, because one side has a lot of money at risk and the other gets upside but doesn't have much downside.

There are a few cases where there's a simple, elegant solution: if a company was founded by someone who already had capital, or by a team that didn't need much capital to get going, they'll have a built-in incentive to maximize the value of their stock. Mark Zuckerberg has a $1/year base salary, but to the extent that he's profit-motivated, owning 343m shares of Meta probably gives him an incentive to get out of bed in the morning and make sure the value of those shares doesn’t evaporate.

But most big company CEOs aren't also founding CEOs, and the skill of building a company is not the same as the skill of managing an already big company. At that point, you have to strike a balance between two forces:

• You can give your managers a stake in the upside they produce, perhaps by granting them lots of options struck at the current share price, or giving them performance-based stock units that vest at higher prices. This means that they aren’t getting paid for what someone else built, but it also means that they have a different set of incentives compared to shareholders: performance hurdles mean that they're long volatility as well as price, and have an incentive to do crazy things that make the stock move fast rather than sane things that make it move up slowly.

• If you grant them pure equity, you've aligned their incentives going forward but also given them a windfall that's more tied to taking a job than to how they perform at it.

Fortunately, there are other forces pushing in the opposite direction. Volatility is good to the CEO-as-holder-of-out-of-the-money-calls, but bad to the CEO-who-has-to-explain-decisions-to-the-board. It's pretty easy to imagine choices that lead to binary outcomes that are good for someone who has stock options, but the board of directors will spend a lot of time picking apart the upside case and fixating on the downside if that volatility-maximizing decision doesn't directly lead to a higher share price.

And in a way, experienced CEO hires are bringing some preexisting equity to the job: if they got hired, it's because someone was able to make a credible case that they'd help the company perform better than some hypothetical replacement-level CEO. The ability to make this case is a function of track record, and that track record is at risk any time a CEO maximizes variance. Between the reputational pressure on managers and the incentive of boards, companies tend to have a pretty low expressed risk preference: it's easier to underwrite buybacks than expansion, easier to justify bolt-on acquisitions than transformative ones, and easier to nudge up pricing than to blow up the pricing structure entirely. Even when companies do take risks, they often taken them in a risk-averse way: when big tech companies talk about their AI capex decisions, they'll often frame them in terms of the risk of being left behind, not just the upside from being first to some futuristic new technology.

There are some odd cases where an executive already has a very motivating stake in the company, and gets a pay package offering more of the same. Elon Musk periodically informs the car company he's synonymous with that he needs a sum of money that would move the needle for the richest man on earth in order to keep his attention from wandering. Rivian has apparently decided that it's very important to follow industry norms around compensation, so they're offering their billionaire CEO billions of dollars in performance-based comp, too.

This is just a lot weirder than the other cases above, but the more confusing you find the behavior of centibillionaires, the less confused you get to be about why you aren't one. If a $100m bonus wouldn't motivate you to work very hard if you already had a billion dollars, you're probably going to stop short of $1bn.¹ And if someone really is wired in such a way that the second billion hits just as hard as the first, and the third billion just makes them wonder why they don't have four already—then they're pretty incentive-aligned with shareholders, and if they ever underperform it will be incredibly easy to make the case that they're overpaid.

One workaround for executive compensation in general is to stop thinking about it in terms of what share of the upside goes to shareholders and what share goes to managers, but to look at what underlying drivers will improve the share price over time. There are executives who get paid a fixed piece of the profits they generate, or who have pay tied to other KPIs like return on equity or even total revenue. These variables tend to be more stable than share prices, and they're more under the control of management—but as soon as they're the target, there will be cases where the pay scheme tells managers to do something that will make the stock price go down. Pay based on return on equity, for example, and you're telling CEOs to lever up; switch that to return on assets, and you're telling them to outsource the most capital-intensive parts of their business, which may be more strategically important than the asset-light bits. Pay based on total profit, and you're paying them to make dilutive acquisitions; pay based on earnings per share, and you're paying every company trading at 30x earnings to use stock to diversify into sectors where the average multiple is closer to 10x.

None of this ever makes compensation easy to figure out. Any cogent argument that some people will be motivated by a boatload of money will probably be best articulated by a sociopath who wants the money and can claim whatever motivation is necessary to get it. But overall, things work out pretty well: there are still plenty of complaints that big companies and captive boards overpay their executives, but it's not like CEO pay is the first cost a private equity firm cuts—though they 're more likely to put a cap on salaries and tilt comp much more to equity, which tends to make average pay even higher.² There will be standout cases where CEOs are paid extremely well for mediocre performance, but the go-to example of this, David Zaslav, is currently presiding over a bidding war that has already produced more for shareholders than he's earned at their expense. Figuring out the right amount of pay, and the right way to pay, is a fundamentally messy business.

Compensation and conflicts of interest are a long-running theme in The Diff, at the level of operating businesses and asset managers. For example:

• Sometimes figuring out the right pay structure is similar to designing a good game.

• Compensation norms are contagiuos ($).

• Paying with stock options is a surprisingly new feature of business (and yet another example of a practice that started out as regulatory arbitrage and then evolved into having decent standalone economics).

• If there's no correlation between CEO pay and stock performance, it can mean that there's no such thing as a good CEO, or that good CEOs charge exactly what they're worth ($).

Join the discussion!

1  Within this, of course, there's room for skill, luck, moral flexibility, inherited wealth, secret payments from the Trilateral Commission, and whatever other explanation you care to apply to the accumulation of extreme wealth. But, whichever of those explanations you take, there has to be some element of personality to it. Even if someone's incredibly lucky, there comes a point at which they don't need to roll the dice any more, unless they have a peculiar reward function. But this is true of high achievers in other domains: bands that drop one amazing album and writers who produce a phenomenal debut album will both usually go for an encore.

2  This will be true independent of any change in incentives. If your pay is in the form of cash, doing a bad job means losing your job. If it takes the form of stock, doing a bad job means losing your income and a chunk of your assets. So for someone with conventional sensitivity to risk, the amount of equity comp they need to offset a cut in cash comp, and the expected bonus they need to offset a reduction in base, will both be higher.