Posted on Harvard Business Review: March 22, 2010 5:43 PM
I had been prepared to switch subjects from my five-post critique of the fundamental flaws in the prevailing thinking about executives, compensation and the capital markets, but my brilliant colleague and writing partner Mihnea Moldoveanu read the blogs and had a terrific idea that I wanted to share.
While I might wish that stock-based compensation would go away, it probably won't no matter what nasty arguments I make against it. So I should be open to clever ways to make stock-based compensation work better. Here is my rendition of the Moldoveanu argument for doing so.
Recall that the notion of stock options as a preferred form of executive incentive compensation arose out of the seminal 1976 paper by Mike Jensen and Bill Meckling (Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, Journal of Financial Economics, 1976).
The takeaway was that options align the interests of executives with the shareholders: if the stock goes up, option value skyrockets, making both shareholders and executives happy. However, the problem has become that options became largely costless for the modern executive. Executives generally got paid a healthy cash compensation with options slathered on top in order to 'align incentives'. This provided an encouragement for executives to swing for the fences. The upside for executives is that they could get wildly rich if they hit a home run. And if they strike out instead, they have their base salary to count on.
However, if swinging for the fences and striking out puts the firm in financial distress (for example like most of the US financial services industry in 2008-9 and most of the US high-tech industry in 2001-2) then the shareholders and bondholders, who actually have to put up their own money to earn returns, get hurt badly, losing part or even all of their principal.
When the negative impact of options as incentives was recognized, it was addressed by replacing stock options with deferred stock units (DSU, alternatively referred to as restricted stock units or RSU). This vehicle gave the executive the right to the value of a specified number of shares sometime in the future, either a particular number of years hence or at retirement.
The notional advantage of the DSU was that it exposed the executive to the whole distribution of upside and downside, not just the upside half of the return distribution as with options. The idea was that this would make executives more conscious of the downside of their behavior, which would impact not only the shareholders' stock value but their DSU value as well.
Sadly, this reasoning is pretty lame. Since these DSU are given over and above the healthy base salary, the downside for the executive is not particularly damaging. If the stock price falls by 50% between award of the DSU and its vesting, so what? The executive would earn years of salary plus half of the DSU value at award while the shareholders would watch their investment drop by a half. The same incentive remains to swing for the fences. The switch from options to DSU accomplished very little in improving the alignment of incentives of executives and capital-providers.
We need to return to the 1976 Jensen and Meckling article for the clue to a solution. What the article actually showed (in contrast to the popular conception) was how the incentives of an option holder are equivalent to those of a holder of a piece of equity in a debt-levered firm. A sliver of levered equity has the following payoff structure: you make no money until you pay off the debt and you make dollar for dollar returns on the stock thereafter. For an option, you make no money until the stock reaches the strike price and you make dollar for dollar returns thereafter.
So, options would seem to give managers incentives that combine cash discipline (required to pay off debts) and equity value maximization, right? Wrong: the downside for executive is zero while for the debt holders in particular, the downside is huge.
What then can be done? We should take seriously the Jensen and Meckling focus on the sliver of levered equity. If we really want to align the interests of executives with the providers of capital, as part of their long term incentive package, we should sell executives a sliver of the firm's debt (senior and junior in proportion to actual) stapled (permanently, so the debt can't be stripped off and sold) to a sliver of its equity (with proportions of debt and equity equal to current actual). We could finance the executive's purchase, but with recourse, not non-recourse, debt.
This instrument has the upside payoff of a DSU but additionally focuses the executive on the cash discipline required to make principal and interest payments on the debt. There will be no more swinging for the fences with the sliver of levered equity. Interestingly, Jensen suggested just such an instrument in a much less cited 1986 American Economic Review article (Michael C. Jensen, Agency Cost of Free Cash Flow, Corporate Finance, and Takeovers, American Economic Review, Vol 76, No 2, May 1986).
Bondholders would be wise to put a covenant requiring exactly this sort of executive compensation structure in their debenture agreements. Their biggest risk exposure is to executives swinging for the fences and they typically have no useful protection for that.
Thanks, Mihnea, for the inspiration behind this solution.