سرمایه گذاری شرکت، انتقال پولی و مشکلات ترازنامه در ژاپن: یک تحقیق با استفاده از داده های میکرو

This paper investigates what can be learned about the effects of monetary policy on firm investment after the collapse of the asset price bubble in Japan. By estimating firm investment functions based on corporate panel data, the paper reveals that the monetary easing worked through the interest rate channel, but its effect through the credit channel was blocked because of a deterioration in balance-sheet conditions. The paper finds that this deterioration in balance-sheet conditions, especially in bank balance-sheet conditions, hampered investment by smaller non-bond-issuing firms more severely than that by larger bond-issuing firms.

This paper investigates what can be learned about the effects of monetary policy on firm investment after the collapse of the asset price bubble in Japan. For this purpose, based on corporate panel data, the paper estimates accelerator-type firm investment functions augmented with variables relating the balance-sheet conditions both of firm themselves and of their main banks. This approach relates the paper to the existing literature as follows: • Our paper is one of few examples that estimate accelerator-type investment functions using Japanese data. Most of the previous studies employing Japanese corporate panel data estimate Q-type investment functions. These include Asako et al., 1989, Hoshi and Kashyap, 1990 and Hayashi and Inoue, 1991. More recently, Ogawa and Kitasaka, 1998 and Suzuki, 2001 follow this tradition. • Our results are comparable with findings in other industrial countries. Outside Japan, there are several studies that estimate accelerator-type investment functions to gauge the effects of monetary policy. For instance, a series of studies organized by the European Central Bank (ECB) adopt the same type of functional forms to model firm investment behavior in their countries.1Chirinko et al. (1999) also estimate accelerator-type investment functions, using U.S. corporate panel data. • Our paper follows Sekine (1999) in that balance-sheet conditions of both firm themselves and of their main banks are controlled simultaneously. In order to observe the impact of asset price fluctuations, some researchers augment their investment functions either with variables reflecting balance-sheet conditions of firms themselves (Ogawa et al., 1996) or those of their main banks (Gibson, 1997). Sekine (1999) combines these two approaches by examining the balance-sheet conditions of both at the same time. The paper finds that the monetary easing after the bubble burst worked through the interest rate channel, but its effect through the credit channel was blocked because financial constraints became tighter following the deterioration in balance-sheet conditions. The paper demonstrates this by showing that the deterioration in balance-sheet conditions, especially in bank balance-sheet conditions, hampered investment by smaller non-bond-issuing firms more severely than that by larger bond-issuing firms. The rest of the paper is organized as follows: Section 2 introduces the reduced-form investment functions used to examine monetary transmission channels. Section 3 outlines our micro data set. Section 4 presents our estimation results. Section 5 concludes the paper by discussing some policy implications of the research, which is followed by a data appendix (Appendix A).

It is sometimes argued that the monetary easing in the 1990s was largely ineffective, because it failed towork either through the interest rate channel or through the credit channel. The interest rate channel is said to have become less effective as the nominal short-term interest rate approached the zero bound. The credit channel is also said to have become less effective as non-performing loan problems prevented banks from expanding credit. By estimating firm investment functions based on corporate panel data, this paper finds that the monetary easing worked through the interest rate channel, but its effect through the credit channel was overwhelmed by the deterioration of balance-sheet conditions after the bubble burst. It finds that the coefficients on the user cost, which reflects the interest rate, are similar to those found for other industrial countries such as Italy and France. Moreover, by quantifying the impact of changes in the user cost upon firm investment, it shows that the interest rate channel was effective at least in the first half of the 1990s.18It then demonstrates that the credit channel was blocked by showing that the deterioration in balance-sheet conditions, especially in bank balance-sheet conditions, hampered the investment of smaller non-bond-issuing firms more severely than that of larger bondissuing firms. Looking ahead, in order to restore the functioning of the credit channel, the balancesheet conditions of both firms and banks need to be repaired. This requires either a reduction in outstanding debts or an increase in market-valued assets, which can be achieved by, say, pursuing enterprise reform and disposing of non-performing loans. Although the paper does not look at the lending data, the finding that investment was hampered by the deterioration in bank balance-sheet conditions is consistent with the view that investment was depressed by a ‘capital crunch.’ However, our findings are somewhat at odds with Woo (1999), who approaches the problem by looking at bank lending data: he finds that the capital crunch became evident only after the financial crisis in 1997. This differs from our analysis which suggests that the capital crunch was depressing firm investment even before the crisis period. More empirical research is warranted to resolve this issue.19 This paper has implications for the debate concerning the asymmetric effects of monetary policy. Due to the presence of a ‘financial accelerator,’ monetary policy is thought to be more effective when firms face tighter financial constraints. In fact, Hosono and Watanabe (2002) find empirical evidence to support this proposition using Japanese data. However, in the light of our findings in this paper,we believe that this theory requires a degree of modification in order to accommodate recent Japanese experience. Where tighter financial constraints have resulted from a fall in asset prices after the collapse of the bubble, it is far from clear that an easy monetary policy is as effective as the theory suggests.