The effect of Tobin's q on investment in a bank-based financial system: Evidence from Japan

This study replicates the empirical analyses of investment-q theory in Japanese settings. We focus on publicly listed firms in Japan because their financial system is a bank-based system, which is different from the market-oriented system in the U.S. First, we find that investment-q regression does not work better for high-tech firms in Japan, which are different from U.S. firms during 1984–2015. Second, the investment-q regression works better for high-tech firms post the bubble-bursting era, same as in U.S. Third, the value of fits of the investment-q regression in Japanese firms is smaller than that in the U.S. Based on the findings, we interpret that the investment-q theory may not work better under financial constraints in research-intensive firms in a bank-based system. •Investment-q regression does not work better for high-tech firms with higher volatility of q in Japan.•Investment-q relationships works better for high-tech firms post bubble-bursting era in Japan.•The relatively low R2 value in Japanese high-tech corporations can be explained by the financial constraints of younger high-tech firms.