경기변동이 기업의 투자와 고용 및 자금조달에 미치는 영향에 관한 연구

We investigate whether the impact of external finance premium and financial constraints on firms’ investment, hiring, and financing decisions varies over time depending on macroeconomic conditions, investor senti- ment, and economic policy uncertainty. McLean and Zhao (2014) study the effect of business cycle and investor sentiment on the cost of external finance. They find that both investment and employment are less sensitive to Tobin’ s q and more sensitive to cash flow during the recession and low investor sentiment periods, which is consistent with time-varying external finance costs. Following McLean and Zhao (2014), we examine whether the cross- sectional relationship between Tobin’s q and investment (or investment sensitivity to Tobin’s q) varies over the business cycle, over periods of high and low investor sentiment, or over periods of high and low economic policy uncertainty. As proxies for economic conditions that are likely to affect external finance premium, we use seven domestic and eight U.S. variables in the empirical analyses. They include changes in interest rates, stock market volatilities, major business cycle indices, sentiment indices and the economic policy uncertainty index from Baker, Bloom, and Davis (2016). Our sample includes 26,675 firm-year observations of non-financial firms listed on the Korea Stock Exchange for the period of 1992~2015 and the number of firms included in the sample is 2,283. We find that investment sensitivity to Tobin’s q is high during the period of expansion, high investor sentiment and low economic policy un- certainty. The results imply that external finance is less costly during these periods. We also find that investment sensitivity to cash flow decreases during the period of expansion, high investor sentiment, and low economic policy uncertainly, which is consistent with previous studies interpreting investment sensitivity to cash flow as evidence of financial constraints. As in McLean and Zhao (2014), we assume that firms with higher investment have higher demands for hiring and external financing, and examine employ- ment (equity․debt financing) sensitivity to Tobin’s q (cash flow). The employment and external finance sensitivities to Tobin’s q are procyclical, which is consistent with external finance playing a role in explaining the investment sensitives. The countercyclical relationship of the sensitives to cash flow is significant for equity finance but insignificant for employment and debt finance. As proxies for economic conditions that affect external finance pre- mium, we use U.S. variables as well as domestic variables, based on Li et al. (2015) who find that changes in U.S. interest rates and stock market volatilities affect corporate investments in emerging markets. We also find that corporate investment, hiring, and financing decisions in Korea show significant relationship with the U.S. economic conditions, which may not be surprising given the importance of trade in the Korean economy and the open capital markets in Korea. More specifically, changes in the U.S. in- terest rates appear to matter more than changes in domestic interest rates, and the U.S. investor sentiment index of Baker and Wurgler (2006) also shows statistically significant results. Out of all the proxies for economic conditions that are considered, two economic policy uncertainty indices of Baker et al. (2016) measured for the Korean and U.S. economy, respectively, show the most consistent and statistically significant results. Our findings are also robust to other model specifications. When we include yearly observations of macroeconomic variables instead of year fixed effect, the estimates show that the sensitivities to Tobin’s q are procyclical whereas the sensitivities to cash flow are countercyclical, con- sistent with our baseline specification. We also construct the aggregate macroeconomic variables using the first principal components of the macro- economic variables in each country. When we include both the domestic and U.S. aggregate macroeconomic variables in the regression, the results indi- cate that the changes in U.S. aggregate macroeconomic variable appear to matter more than the changes in domestic aggregate macroeconomic variable. When we include both the domestic and the orthogonalized U.S. economic policy uncertainty indices in the regression, the results indicate that the orthogonalized U.S. economic policy uncertainty index is still statistically significant.