Financial constraints and investment: An alternative empirical framework

Abstract

In the empirical literature on financial constraints, firms are usually assigned to distinct groups, according to whether or not they are supposed to be financially constrained (FC). Several recent empirical papers studying the relationship between firms’ cash flows and investment, have found mixed results regarding whether or not more FC firms show higher or lower sensitivity of investment to cash flows (CFSI). We show that these mixed results may be attributable to the use of different “ex-ante” sample-splitting schemes. In response to these issues, we develop a novel research strategy that allows for estimating firm-specific investment-cash flow sensitivities and identifying the whole distribution of sensitivities across firms. Specifically, we apply entropy-based estimation methods, which outperform classical estimation methods under various general conditions. Our empirical results, based on a large panel-data set of U.S. firms, demonstrate some of the limitations of prevailing classification schemes. The approach we propose offers new and interesting prospects for re-examining the existing empirical evidence on financial constraints.