The balance sheets of Japanese banks in the late 1990s were damaged by the enormous amount of non-performing loans (NPLs) that had accumulated over the previous decade. NPLs at the end of fiscal year for 1997 (March 1998) reached 30 trillion yen, or 5.5 percent of loans supplied by domestically licensed banks. 1 The write off of NPLs against equity leading to a sharp fall in the ratio of equity capital to assets (the book based capital to asset ratio) of domestically licensed banks in March 1998 was followed by a long lasting fall in the domestic lending growth (Figure 1). Domestic loans fell by 20 trillion yen, or about 4 percent during the three year period from April 1997 to March 2000. The BOJ’s tankan “lending attitude of financial institutions” diffusion indices also experienced sharp declines in March 1998.
Did this fall in bank capital, the “capital crunch”, cause the reduction in supply of bank loans, the “credit crunch” in the late 1990s? In order to satisfy the capital adequacy requirements, banks may have cut back on their lending in response to large losses of capital as issuing the new equity incurs costs associated with asymmetric information between investors and banks. When a banking system involves binding capital requirements, in addition to the standard reserve requirements, it will be possible that the limitation on the expansion of loans may be capital and not reserves.
The seminal work by Bernanke and Lown (1991) defines a bank “credit crunch” as a “significant leftward shift in the supply curve of bank loans, holding constant both the safe real interest rate and the quality of potential borrowers.” The interest rate was very low and stable throughout the late 1990s. The quality of borrowers, in particular their lending demand may have declined under the stagnant economic environment. It is identifying the supply side phenomenon of the “credit crunch” with the alternative hypothesis of declining lending demand that is the most important in our empirical analysis.
We construct the unique instrumental variable in order to capture the “capital crunch” that is independent of the present lending demand. According to Hoshi and Kashyap (2000), since large firms became almost independent of banks following the capital market liberalization of the late 1980s, and banks themselves were still confined to their traditional lending business, banks replaced their traditional keiretsu lending (relationship lending) with lending to opaque small and medium firms, taking land as collateral, while also expanding through even riskier real estate lending. Such a shift in the lending portfolio exposed banks to asset price risks, which did not become apparent until the bubble burst in the 1990s. Based on the stylized fact found by Ueda (2000) and Hoshi (2001) that the bank’s real estate lending in the 1980s best explains the bank’s NPLs in the late 1990s, we use within bank share of real estate lending in the late 1980s as an instrumental variable for bank capital. We examine the impact of capital “surplus” defined as the gap between the actual and the estimated bank specific target capital to asset ratios on lending supply by running year by year cross section regressions.
The “capital crunch” is regulatory driven. The reported bank balance sheets themselves are the reflection of regulatory toughness or the banks’ concessions to the regulator. It was in March 1998, when the MOF required banks to carry out a more rigorous self-assessment of their assets and the adequate loan loss write offs and the provisions that the negative capital shock was observed. As such, the econometric analysis of bank balance sheets needs to control for changes enforced by regulatory and institutional oversight bodies. Any valid interpretation of the results inevitably requires regulatory and institutional information.
We find that banks cut back on their lending supply in the fiscal year 1997 in response to a large loss of bank capital (the “regulatory driven capital crunch”) mainly caused by the rigorous self-assessment of assets requested by the regulator. Increased capital mainly due to an injection of public capital in FY 1998, in turn, likely relaxed the capital constraint of banks, thus lead to an increase in the supply of loans. This positive effect on lending, however, barely offsets the “credit crunch” of the previous year.
The remainder of the paper is arranged as follows. In section 2, we introduce the notion of credit crunch and review the literature. In section 3, we discuss the relevant regulatory background. In section 4, data and econometric issues are set out. In section 5, preliminary results are reported. In section 6, main results are reported and some policy implications are derived. Section 7 concludes the paper.
Download
Prudential Regulation and the “Credit Crunch”: Evidence from Japan
