Monday, August 22, 2005

Corruption, Firm Governance, and the Cost of Capital

Garmaise and Liu present a model that has several important implications in their paper
Corruption, Firm Governance, and the Cost of Capital.

In the paper they
"develop a model of a firm owned by shareholders and administered by managers who may be either honest or dishonest. When managers have an informational advantage but shareholders retain control, dishonest managers can make false reports that distort investment and thereby reduce firm cash flows. When dishonest managers have privileged access to both information and control, firm value is further reduced and profits are diminished especially in the worst states of the world."
Put another way, Garmaise and Liu separate managerial advantages into informational and control measures. Where managers have information (but not control) advantages, managers' ability to take advantage of the informational advantages are limited. However, where the manager has both informational and control advantages, investors have little ability to limit managerial actions.

The authors take this idea and model the relationships in several steps. They begin off with the assumptions that
"Managers have a preference for large projects and empire-building. They therefore have an incentive to misreport the private signal they alone receive, neglecting to inform the shareholders of bad news. "
However, in the
"information model, shareholders retain control of the firm, and they design an optimal investment policy that reflects the potential for false reporting on the part of management....[the result is that] managerial deception leads to investment distortions, but they are not correlated with market outcomes."
So if investors still have control, systematic risk will not increase.

The authors then look at what happens if managers have both informational and control advantages.
"...managers set the investment level, basing it on the report they make to shareholders. Dishonest managers hide bad outcomes and therefore set the investment level consistently too high. Since shareholders do not have control, they cannot adjust the firm'’s investment downward to reflect the possibility of signal falsification."
So far nothing really new. However, the authors then go on to investigate the implications of the informational and control advantages.
"In the control model, however, the overinvestment selected by dishonest managers essentially gives the shareholders a more levered claim on the market; ex ante overinvestment is ex post beneficial when the market outcome is good and ex post destructive when the market outcome is poor. Since the extent of overinvestment increases with the proportion of dishonest managers, we show that the firmÂ’s beta is increasing in this proportion."
Which deserves a wow. But the paper is not done yet. The authors next look at this empirically across countries and find that the empirical evidence fits the model.
"We find that corruption increases firm betas substantially, controlling for industry, leverage and per capita GDP. For example, increasing the degree of a countryÂ’s corruption from the level of Canada to that of South Korea would increase the industry-adjusted beta of firms by 0.35."
Predictably, in countries where stong "anti-director" rights, dishonesty has less of an impact on the systematic risk.
"dishonesty has a stronger effect on firm betas in countries with weak antidirector rights, and this is what we find empirically. In countries with very strong antidirector rights corruption has no effect on beta, but when rights are moderate or weak, corruption substantially increases beta."
A definite I^3 paper! (interesting, informative, and important)

Mark J. Garmaise and Jun Liu, "Corruption, Firm Governance, and the Cost of Capital" (February 28, 2005). Finance. Paper 1-05.

I was just told that the paper is also available through SSRN.

No comments: