Saturday, 19 October 2019: 5:30 PM
We examine the firm-level cyclicality of research and development (R&D) spending using Korean corporate finance data from the KISVALUE database, a corporate finance database of South Korea produced by Korea Investors Service (KIS). Using a panel data analysis, we find that firms’ R&D spending is procyclical and asymmetric, and financial constraints effectively limit the R&D spending. In the analysis, we use the KIS credit rating as a proxy of financial constraints. For instance, a firm with a high (bad) KIS credit rating would face strict financial constraints. Furthermore, our estimation results show that R&D spending is procyclical even after the countering forces of financial constraints, and the financial constraints do not have a significant role in direction of R&D cyclicality. Instead, the financial constraints tend to restrict R&D spending growth when firms intend to increase R&D spending during an economic expansion. This finding contradicts the Schumpeterian explanation of procyclical R&D investment by financial constraints, even though it is supposedly inherently countercyclical (Aghion et al., 2012). When we consider alternative variables such as the debt ratio, we still find procyclical and asymmetric R&D spending, and financial constraints constrain R&D spending the most during expansions. Thus, this study adds to the prior empirical findings on procyclical R&D investment (Fabrizio & Tsolmon, 2014; Ouyang, 2011) and supports prediction of Barlevy (2007). Moreover, our results show that the negative impact of financial constraints on the procyclicality of R&D spending is in effect only when unlisted firms experience sales growth, implying heterogeneity in R&D cyclicality. This implies that unlisted firms are more likely to face constraints on R&D spending due to bad credit ratings, particularly in expansionary periods. Our finding of asymmetric cyclicality and the role of financial constraints on cyclicality is relatively unknown and provides interesting insights for further theoretical development.