Agency theory: a model of investor equilibrium and a test of an agency cost rationale for convertible bond financing

TR Number
Date
1982
Journal Title
Journal ISSN
Volume Title
Publisher
Virginia Tech
Abstract

The conflict that may arise among holders of competing claims on firms' assets is being studied under the heading of "agency theory."

The primary purposes of the research done in this study were to: (1) economically model the individual investor's consumption-investment decision as it is modified by the agency problem, and (2) to econometrically model the firm's decision to issue convertible versus nonconvertible bonds using explanatory variables which measure the extent of the agency problem.

Individual investors are assumed to maximize expected utility of consumption by choosing consumption and investment amounts over a single period. A mathematical model of the investor's consumption-investment decision was derived in an environment characterized by agency problems between stockholders and bondholders. It was demonstrated that if the capital markets exhibit conditions known as spanning and competitivity, then the only investors affected by the agency problem are those holding the affected securities prior to the act of expropriation. It was also shown that the agency problem does not vanish in general, even if investors attempt to avoid the expropriation by holding balanced portions of all outstanding claims on a firm's assets.

Implications of the theoretical development were then tested by econometrically modelling the firm's choice of convertible versus nonconvertible debt. The explanatory variables included in the model included measures of the more popular reasons for convertible financing, such as the "debt sweetener" hypothesis and the "delayed equity" rationale discussed in most basic finance textbooks. In addition, measures of agency costs were included, since one possible solution to the agency problem is the issuance of convertible bonds. The empirical results showed that the model accounted for a significant portion of the discrimination between convertible and straight debt, and that the variables designed to measure agency costs were marginally significant.

Description
Keywords
Citation