Saving Stock-Based Compensation From Itself - 22 March 2010

By
Roger
L. Martin

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.

Roger L. Martin has served as dean of the Joseph L. Rotman School of
Management at the University of Toronto since Sept. 1, 1998.