Keiretsu Style Main Bank Relationships, R&D Investment, Leverage, and Firm Value: Quantile Regression Approach

Abstract

Using quantile regression, our results provide explanations for the inconsistent findings that use conventional OLS regression in the extant literature. While the direct effects of R&D investments, leverage, and main bank relationship on Tobin’s Q are insignificant in OLS regression, these effects do show significance in quantile regression. We find that firms’ advantages with high R&D investment over low R&D monotonically increase with firm value, appearing only for high Q firms; while firms’ advantages with low R&D over high R&D monotonically increase with firm value for low Q firms. Tobin’s Q is monotonically increasing with leverage for low Q firms; whereas it is decreasing in high Q firms. Main banks add value for low to median Q firms, while value is destroyed for high Q firms. Meanwhile, we find the interacted effect of main bank and R&D investment which increases with firm value, only appears in medium quantiles, instead of low or high quantiles. Results of this work provide relevant implications for policy makers. Finally, we document that industry quantile effect is larger than the industry effect itself, given that most of the firms in higher quantiles gain from industry effects while lower quantile firms suffer negative effects. We also find the results of OLS are seriously influenced by outliers. In stark contrast, quantile regression results are impervious to either inclusion or exclusion outliers.