Book excerpt: In the thirty or so years since the Modigliani-Miller theorem, scholars have worked to relax the theorem's assumptions in order to obtain a better understanding of the capital structure of firms. This work has produced some important insights but has not yet delivered a fully coherent theory of optimal capital structure. For example, at present we do not understand very well the distinguishing features of debt and equity or why these claims, as opposed to the many instruments that could be chosen, are most frequently issued by firms. Given this state of affairs, existing explanations of the debt-equity ratio must be seem as still preliminary, as must efforts to use these explanations to understand global trends such as the large increases in leverage in the United States and United Kingdom during the 1980s.In the first part of this paper, I will argue that one reason progress on understanding capital structure has been limited is that relatively few analysts have adopted an explicit agency-theoretic or managerial discretion perspective. In particular, although the literature, starting with the work of Michael Jensen and William Meckling, frequently refers to conflicts of interest, most of it does not emphasize the conflict of interest between a firm's management and its security holders. But I argue that this particular conflict of interest - that is, the idea that management is self-interested - is critical. In the absence of this conflict, optimal capital structure would look very different from what is observed in the world. In particular, firms would not issue senior or secured debt, whereas in fact a considerable amount of corporate debt has at least one of these features. That is, standard departures from the Modigliani-Miller framework that focus on the role of taxes, asymmetric information, or incomplete markets but ignore managerial self-interest are not sufficient to explain observed capital structure.In the second part of the analysis I will discuss what has been learned from the relatively few studies that have explicitly adopted an agency-theoretic perspective. This body of work, although itself quite preliminary, can explain the use of senior or secured debt or both, as well as shed light on some observe patterns of capital structure, including a number of findings from studies that measure the response of security prices to important events that affect optimal capital structure (quot;event studiesquot;).