Corporate Finance

Managerial compensation, revolving doors, and labor market mobility, for financial regulators

My recent work with John A. Cole entitled A Regulator’s Exercise of Career Option To Quit and Join A Regulated Firm’s Management with Applications to Financial Institutions fills a gap in the literature by posing regulatory capture and revolving door hypotheses in the context of managerial compensation, and labor market mobility, for regulators who design legislation that affect firm capital structure, and who subsequently exercise career options to quit and join firm management or a lobbying firm. Applied to a sample of commercial banks, we find that bank CEO vega between 1994 and 2006, and our managerial vega-leverage theory predicts an out-of-sample critical time point in 2007, and a range of critical time points for commercial bank failures thereafter. So our model would have predicted the 2008 financial crisis. slides

The Public—Private Partnership Puzzle: Overinvestment and Underinvestment Implied by Agency Costs of Access to Credit (in progress)

This paper examines the effects of agency costs on public-private sector partnerships established (by public sector contracting with private equity (PE) and venture capital funds) to alleviate credit rationing via access to credit. We show that with venture capitalists as delegated managers of access to credit, agency costs of loan provision induce a public-private partnership puzzle: A Stiglitz-Weiss type underinvestment problem triggered by interest rate sorting, and a DeMarza-Webb type overinvestment problem triggered by subsidized loans. The under/over investment puzzle stems from agency conflicts that undermine the benevolent intent of the partnership. Moreover, the range of loans offered through the partnership is negatively correlated with the expected risk associated with high risk firms. We extend the analysis to a sequential contract setting in which venture capitalists use a credit scoring mechanism to assign firm type. We show how firm effort affects posterior credit risk assessment by the venture capitalists. There is no welfare loss from the partnership when the marginal cost of a firm characteristic is weighted by the relative effect of that characteristic in posterior evaluation of firm credit risk.

Learning And Termination Delays In Corporate Investments: Why Managerial Compensation Schemes, Control Rights And Psychology Matter (in progress) (with H. Morgan & O. Ngwenyama)

We present a model of the termination decision of managers in the context of software infrastructure projects. Starting with the optimal-stopping-time model of Ryan and Lippman (2003), we consider a particular application that is informed by findings from a laboratory experiment conducted by Croson, Elfenbein and Knott (2014). The latter authors found that participants with an equity stake and control rights over project continuation were more likely to delay project termination beyond the optimal stopping time relative to those without equity and such control rights. For this reason, in this paper we explicitly consider how option-like or project-contingent managerial compensation schemes coupled with control rights over project continuation may contribute to deviations from an optimal termination rule. This approach identifies agency conflicts in the Ryan-Lippman model. For instance, the value of the option to exit increases when volatility of firm profit decreases. However, managerial compensation schemes are typically based on stock option, i.e., equity stake in the firm. Moreover, the value of the option increases with increased firm volatility which is known to increase with increased spending on software infrastructure projects. This leads to higher profitability thresholds for the firm. Whereupon we show how the value of the early exercise premium for the option, in the context of optimal stopping times for managerial compensation with an American style option, decreases with higher profit thresholds. This escalation of commitment result increases delay in project termination. Contrary to Ryan-Lippman posterior probability representation, in a behavioural context subjects do not update probabilities according to Bayes rule. Instead they employ a representativeness heuristic which leads to different posterior probabilities than that posited by Ryan-Lipmann. We illustrate this with a simple application in the context of Ryan and Lippman (2003) two-states model. We show how probabilistic myopic loss aversion and managerial confidence affect the threshold for the expected rate of profit, and that that also introduces behavioural delay in exit times different from that predicted by Ryan and Lippman (2003).