INVESTMENT AND CASH FLOW: EMPIRICAL STUDY USING PANEL DATA FROM US MANUFACTURING INDUSTRY

This study aims to illustrate the investment-cash flow sensitivity of constrained and unconstrained firms. Using data from 2,233 US manufacturing firms over 1999 to 2008, several findings present. Firstly, cash flow is found to significantly affect firm’s investment behaviour. Secondly, when firms are labelled as constrained and unconstrained by dividend and firm size dummy, unconstrained firms are found more sensitive if the sample is split by dividend dummy. Meanwhile, the result is reversed if firms in the sample are separated by firm size. Finally, the research on splitting firms by the two criterions interacted reveals that regardless of firm size, high dividend payers are always more sensitive. Similarly, for firms at identical dividend payment level, small firms are more sensitive. Several econometric techniques are employed to generate the above results. OLS, fixed-effect model, first-differencing transformation and instrumental variables by GMM are utilized to test the significance of cash flow on investment. For the test of constrained and unconstrained firms, fixed-effect and instrumental variables by GMM are employed. However, GMM is found to exhibit weak estimation power