A Critique of Quantitative Structural Models in Corporate Finance

I contend that structural quantitative models are better suited for situations in which the important forces are a priori well understood. This is rarely the case in corporate finance. It is also more difficult to detect and correct for misspecification in these models. Their existing empirical tests have largely been perfunctory. They have not entertained powerful and often simpler alternatives. They have not been tested out-of-sample. They have not been tested in the context of quasi-experiments, such as tax law changes. Moreover, even insample, the empirical evidence strongly rejects current models. They attribute capital structure behavior primarily to forces that seem to be, at best, of minor importance. Managers intervene in their capital structures very actively for reasons not yet understood. ∗This paper would make suitable reading for a Ph.D. course in empirical methods or corporate finance, especially together with Fischer, Heinkel and Zechner (1989), Hennessy and Whited (2005), and Strebulaev (2007). Please note that it is the purpose of a critique to focus on shortcomings. Thus, the perspective of this paper is inevitably biased.