We exploit a change in bankruptcy law in 1978 in the U.S. as an exogenous shock that increased the cost of external funds for public companies. In a quasi-natural experiment setting, we investigate the impact of an increased cost of debt on the investment-cash flow sensitivity of firms. Our results show that the sensitivity of investment to cash flow increased by one third after 1978, and for a sample of firms likely to be more financially constrained the effect was as high as 80%. Our findings suggest the market value of a dollar in cash holdings increased by 12 cents after the change in law, with a larger effect for financially constrained firms.

Introduction

We exploit a change in bankruptcy law in 1978 in the U.S. as an exogenous shock that increased the cost of external funds for public companies. In a quasi-natural experiment setting, we investigate the impact of an increased cost of debt on the investment-cash flow sensitivity of firms. The interaction between investment and financing decisions is of major importance in corporate finance. In a Miller-Modigliani type world, these two decisions are independent; all positive net present value projects are funded with either internal or external funds, which are perfect substitutes. It is now well understood that issuance costs, agency conflicts, and information problems drive a wedge between the cost of internal and external funds, and make the financing and investment choice interdependent (Stein, 2003). An important issue that has received much attention in the literature is the sensitivity of investment to internally generated cash flow. If a wedge exists between internal and external funds, it seems natural that firms will use internal cash flow to finance investment as much as possible. However, empirical estimates of the investment-cash flow sensitivity suffer a myriad of challenges. Proxies for the unobservable investment opportunity set of firms are inherently noisy, and cash flow is probably correlated with investment opportunities. Moreover, previous literature beginning with Fazzari, Glenn Hubbard, and Petersen (1988) has emphasized cross sectional comparisons of investment-cash flow sensitivities under the argument that more financially constrained firms will exhibit a higher sensitivity of investment to cash flow. Kaplan and Zingales (1997) criticize this approach and with a simple one-period model of investment show that the sensitivity may not be monotonically increasing in the degree of financing constraints in the cross section. Furthermore, Kaplan and Zingales (1997) argue that disentangling the effect of financing constraints from other firm-specific factors is challenging in the cross section.