سیاست های پولی، اعتباری و فعالیت واقعی : شواهد از ترازنامه شرکت های ژاپنی

We analyze the mechanism of monetary transmission in the Japanese economy by using the quarterly time series data disaggregated by firm size. In particular we examine the channels through which monetary policy influences the firm's fixed investment with special focus on the firm's land. We estimate the vector autoregressive model where we encompass two competing hypotheses on the monetary transmission: monetary and credit channels. Our evidence is in support of the credit channel. We find that land has played a vital role in the monetary transmission, especially for small firms. Moreover, we find that fall of land value in 1990s weakened the efficacy of monetary policy considerably. J. Japan. Int. Econ., December 2000, 14(4), pp. 385–407. Institute of Social and Economic Research, Osaka University, Osaka, Japan Copyright 2000 Academic Press.

The official discount rate in Japan has fallen to a historic low of 0.5% per annum since September 1995 in order to boost the domestic economy. Despite such an easy monetary policy, the real GDP growth rate of Japan for 1998 was the lowest in the postwar period, indicating how stagnant the Japanese economy has become. Increasingly alarmed at the prospect of the Japanese economy plunging into a deflationary spiral, the Bank of Japan (BOJ) started to ease the monetary policy further in September 1998 and in February 1999. The overnight call rate eventually fell to 0.03% per annum or a virtually zero level, allowing for brokerage fees. The predicament of the Japanese economy naturally raises the question why an expansionary monetary policy has been ineffective in boosting the Japanese economy. It has been frequently asserted that the huge burden of bad loans in the banking sector is blamed for the prolonged stagnancy of the economy.2 The underlying logic is as follows. Mounting bad loans hinder commercial banks from playing an intermediary role. This severely affects the real activities of bankdependent firms, most of which are small in size. It should be noted that the bad loan problem stems from excessive loans to the private sector secured by land in the late 1980s. Land assets were ideal for collateral since it was expected that the land price would continue to rise. However, this perception was completely shattered and the sharp decrease in land prices in the 1990s forced a large proportion of loans into default. Thus it appears that land is a key ingredient in understanding the mechanism through which monetary policy is propagated into the real economy. For a rigorous evaluation of the arguments above, it is important to analyze the channels through which monetary policy is transmitted into the real economy. This study is an empirical attempt to analyze the monetary transmission mechanism in Japan with special attention paid to the role of land. In conducting the research for the period including the turbulent late 1980s to the 1990s, the following point deserves to be taken into consideration. In the course of liberalization and internationalization of financial markets, large firms could raise the funds directly from equity or the bond market at a lower cost than bank loans. Therefore, the banks were forced to lend to new customers, most of whom were small firms. When there is asymmetric information between lenders and borrowers, banks will be cautious in sharing with newcustomers information which has not been accumulated within the banks. However, when the land price is rising and is expected to continue to rise, the default risk of loans will be alleviated to a large extent by taking land as collateral. However, if the expectation turns out to be wrong, it will lead to a massive default of loans. It is conjectured that this was exactly the situation in Japan in the 1980s and 1990s. If this is the case, then land will play a key role in the propagation of monetary policy for small firms.3 To incorporate this point into our study, we will investigate the monetary transmission mechanism by using the less aggregated data of firms classified by firm size.4 Specifically we analyze the channels through which monetary policy affects a firm’s fixed investment based on the vector autoregression (VAR) model. We use the quarterly data reported in the Quarterly Report of Financial Statements of Incorporated Business (QRFS) of the Ministry of Finance. The virtue of this data source is that the tangible fixed assets are decomposed into components and that they include the land stock, which is expected to play a central role in our analysis. Our sample period covers the first quarter of 1975 to the first quarter of 1998, covering the long booms and the severe recessions in the 1980s and the 1990s. Let us preview our findings. We find that in Japan land is a key building block in transmitting the monetary policy into the real economy. The monetary policy, measured by the interbank rate or the call rate, affects both total loans and fixed investments by way of land. This channel is especially noticeable for small and medium-sized firms. Our results support the credit view that bank loans as well as borrowers’ collateralizable assets play a central role in propagating the monetary policy into the real activities in the presence of credit market imperfections. We also find that the potency of an expansionary monetary policy is weakened in the 1990s due to a sharp decline of land prices. The paper is organized as follows. Section 2 reviews the theoretical arguments of the monetary transmission mechanism with the role of land explicit in the model. Section 3 describes the data set and sets up a VAR model for empirical analysis. Section 4 presents the evidence and derives implications for the monetary transmission mechanism at work in Japan. Concluding remarks are given in Section 5.