Empirical Policy Function Benchmarks for Evaluation and Estimation of Dynamic Models
University of Minnesota - Finance Department
R. Jay Kahn
University of Rochester - Simon Business School
Toni M. Whited
University of Rochester - Simon School of Business; National Bureau of Economic Research
This paper presents a set of benchmark moments for evaluation or estimation of quantitative capital structure models. The moments are directly related to the models being studied: the main features of each models' empirical policy functions. The paper describe a general method for estimating these benchmarks and shows that they ‘capture’ a substantial part of the actual variation in firms actions in the data. Two versions of these benchmarks are presented: one dimensional ones and two dimensional ones. In both cases we express these as the total change in the control variable and the change relative to the change in the state variable. The empirical policy functions turn out to be smooth and mostly monotonous. Three key numbers that we suggest quantitative dynamic models have match closely are that within firms, for every 10% increase in debt relative to assets investment relative to assets declines 3.7%, debt issuance relative to market value decreases 1.1% and equity issuance relative to market value increases 0.5%.
Number of Pages in PDF File: 34
Keywords: dynamic models of capital structure, policy function, value function, model evaluation
JEL Classification: C14, C52, C61, G31, G32working papers series
Date posted: March 17, 2012 ; Last revised: November 19, 2013
© 2014 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo3 in 0.531 seconds