One of the weird facts about executive compensation is that the numbers companies give to the SEC don't really line up with how much cash executives actually take home. In 2014, for instance, the top 500 highest-paid executives reported an average fair-value compensation of $19.3 million. But if you calculate how much they actually made based on exercised stock options, the average comes out to $34.3 million.
The difference lies in the delta between fair-value reporting – a pricing model that predicts how much stock options will be worth in the future – and actually realized gains based on stock sales by executives. Here's what that disconnect looks like over the past decade, courtesy of Hopkins and Lazonick:
There are a few theories about the anomaly. One is that the Black-Scholes options pricing model is the wrong tool for the job. Another is that executives just have a better sense of when to offload their vested options, an intuitive feel for the movements of their own company's stock price. Then there's buybacks.
If CEOs could artificially increase demand for their stock during the months they plan to sell it, well, why wouldn't they? Conveniently, most publicly traded companies regularly repurchase their own shares. With advance board approval, they can do so at will, so long as they hold their fire while in possession of material nonpublic information (thus the buyback blackout periods before earnings reports).
If executives can time their stock buybacks around when their stock options vest, any rational and profit-maximizing CEO would go ahead and do so. Yet until now no one has checked to see whether this was actually the case. So David Moore, a PhD candidate at the University of Kentucky, ran some numbers based on executive stock-vesting records and repurchase data.* Here's what his working paper found:
In months when CEOs are predicted to sell equity due to the vesting of equity grants they are more likely to repurchase shares and repurchase more shares, indicating a plausibly causal relation between equity sales and share repurchases. The economic significance is impactful. Vesting induced equity sales increase the likelihood of repurchasing in a month by 4.0-5.5%, an increase representing around 20% of the unconditional probability to repurchase.
In other words, the probability that a company will buy back shares in a given month is about one-fifth higher if the CEO's personal stock options happen to be vesting that month. This, Moore writes, is “consistent with managers strategically using share repurchases to personally benefit from the positive effects of repurchasing on the stock price.” Essentially, CEOs are using company coffers to purchase the stock they themselves are offloading.
And the bigger the options package, the greater the odds that the company will gobble up shares:
A 100% change in the value of equity sold results in an increase of share repurchases by 0.05% of shares outstanding, an increase equivalent to 30% of the average monthly share repurchase volume.
This shouldn't surprise us all that much. The incentives are there, and no one is really watching. In fact, the buyback effect was “significantly stronger” when the CEO also happened to the chair of the company (we're looking at you, Brian Moynihan).
Investors might see all of this and wonder: so what? If a company is authorized to buy back a bunch of shares, does the exact timing matter all that much? From the data Moore presents, companies seem to be just shifting buybacks around. Though repurchases tended to diminish in the month before a CEO's options vested, Moore sees “little evidence that firm value is destroyed.”
That's not going to satisfy everyone. The paper's time frame for determining subsequent returns was just a quarter out. Plenty of investors already feel queasy about the trillions of dollars of cash that flow out or corporate coffers via repurchases. Harvard Law School professor Jesse Fried has written extensively on insider trading via the corporation, concluding that buybacks carried out to benefit insiders do indeed harm public shareholders.
However one feels about the either completely benign or economy-killing effects of buybacks, this line of research should help us all be adults and acknowledge that yes, personal incentives matter in how corporations dole out their cash. At the very least someone should build an algo that takes advantage of the effect.
* Obvious caveat about this paper being written by a grad student and not yet being peer-reviewed.