Executive compensation: The fragile foundations of stock options

Jean-Pierre Danthine, John Donaldson , 2 August 2008

a

A

As boards of directors have sought to align the interests of managers and stockholders, executive stock options have become an ever-larger fraction of the typical CEO’s total compensation (Murphy 1999). Occasionally this practice has led to aggregate compensation payments that are so large as to mock the very connection they are supposed to encourage.

What does economic theory have to say about executive compensation in a dynamic context? From a conceptual perspective, how effective is the granting of stock options in promoting the correct managerial decisions? How confident can we be that when a large fraction of a manager’s compensation assumes this form he or she will be led to undertake the same labour hiring and capital investment decisions that the shareholders would themselves want to undertake if they were similarly informed?

Managerial incentives and the design of compensation contracts

When the systemic implications of executive remuneration are taken into account, that is, in a general equilibrium context, one finds that for a contract to induce managers to take the correct business decisions in the above sense, it must naturally have the following three features.1

  • A significant portion of a manager’s remuneration must be based, in one way or another depending on the context, on her own firm’s performance.

This concurs with the general message of a wealth of microeconomics studies. But this is not sufficient.

  • The general contract characteristics must also be such that the manager is not, as a consequence of this first requirement, enjoying an income stream with time series properties that are too different from the time series properties of the income stream enjoyed by shareholders.

This later restriction arises because, as is well known, the income and consumption position of a manager will determine his or her willingness to undertake risky projects. Optimal delegation requires that this risk attitude is not too different from shareholders’ own.

  • The second feature may have to be modified if the manager’s risk tolerance is inherently different from that of the shareholders.

The typical motivation for stock options (as opposed to pure equity positions) is precisely that the (recurrent) lack of income diversification of a manager may make her excessively prudent (in pursuit of a “quiet life”). This is the idea behind setting executive compensation according to a “highly convex” contract, i.e. one where the upside is really good, but the downside is not so bad. This asymmetry is necessary induce risk averse managers to make the right investment decisions from the perspective of well-diversified stockholders.

Are options-dominated contracts warranted?

Shareholders receive both wage and dividend income, with the wage or salary component being, on average, the larger of the two. This is an implication of National Income Accounting. In the typical modern economy, about 2/3rds of GDP is composed of wages, with capital’s income account for only 1/3. Points 1 and 2 above therefore imply that an optimal contract will have both a salary (with properties close to those of the wage bill) and an incentive component (with properties naturally linked to the income accruing to capital owners) with the former being about twice as large as the latter.

The incentive component may take the form of a non-tradable equity position (giving the right to regular dividend payments) or it may be more closely tied to the firm’s stock price itself. Furthermore, both of these components enter linearly into the manager’s compensation function.

In today’s business world, the salary component appears to be too small relative to the incentive component. Hall and Murphy (2002) report that the grant date value of stock options represented 47% of average CEO pay in 1999. Equilar, Inc., an executive compensation advisory firm, reports that stock options awards represented 81% of CEO compensation for the largest 150 Silicon Valley firms in 2006.

What happens to incentives if the salary component is too small relative to the incentive component?

Such an imbalance between the components of a manager’s compensation will lead to excessive smoothing of the firm’s output from the shareholders’ perspective. They typically prefer a highly pro-cyclical investment policy whereas, without further inducement, the manager will be much more reluctant to exploit the good opportunities and instead select a mildly pro-cyclical or, even, possibly an anti-cyclical investment strategy.

This problem is well recognised, and it is the main justification for using highly convex managerial compensation contracts (i.e. options). Convex contracts overcome this possibility by reducing the personal (expected) cost to the manager of increasing the firm’s investment when times are good. If the manager’s preferences are well represented by a logarithmic utility function of consumption, however, then this latter argument does not apply; the manager’s actions will be insensitive to contract convexity. That is, even a compensation contract that is heavily laden with options will not induce managers to alter their behaviour one whit.

A straightforward application of this logic produces an even more striking result. If the manager happens to be more risk averse than would be dictated by log utility – an entirely plausible configuration – the only way to induce optimal managerial behaviour is by using a highly unconventional remuneration package in which the manager’s compensation is inversely related to the firm’s operating results. This would mean a contract that pays high compensation when profits are low and vice versa.

In this situation an options laden compensation package will induce the manager to behave in a manner directly opposite to what the shareholders would like. More generally, the degree of contract convexity must be related to the relative risk aversion of the manager as compared to the shareholders and if these quantities are not precisely estimated large welfare losses will ensue.

Conclusion

From a theoretical macroeconomic perspective, the circumstances under which a highly convex compensation contract, for example, one that has a large component of options, will properly guide the manager in making the correct hiring and investment decisions are very narrowly defined.

It would be surprising if these circumstances were fulfilled in the typical contract case.

References

Danthine, Jean-Pierre and John B. Donaldson, ‘Executive Compensation and Stock Options: An Inconvenient Truth,’ CEPR Discussion Paper 6890.
Hall, Brian J. and K. Murphy, “Stock Options for Undiversified Executives,” Journal of Accounting and Economics, 33 (2002), 3-42.
Murphy, K., “Executive Compensation”, in O. Ashenfelter and D. Card (eds) Handbook of Labor Economics, vol. 3B, North Holland, 1999, 2485-2567.


1 See the articles cited in reference for details, limitations and proofs.

Topics: Labour markets
Tags: CEO compensation, stock options

Mario J. Gabelli Professor of Economics and Finance at the Graduate School of Business, Columbia University
Professor of Economics and Finance at the University of Lausanne and CEPR Research Fellow