Ebook Japanese Banks' Bad Loans: What happened?
The extent of the 1990s bad loan problems of Japanese banks has received extensive press coverage. By the end of 1995, it was reasonably well-agreed that these problem loans amounted to some 100 trillion yen, or roughly 15% of outstanding loans. However, there is still little systematic analysis of the causes of the loan problem.
That is the task of this paper. We present evidence that, in the late 1980s and early 1990s, the approaching 1992 BIS capital standards, based as they are on accounting measures, had the perverse effect of giving banks an incentive to increase the risk of their loan portfolios; that the loan loss and bad debt write-off procedures helped banks pursue that incentive; that the incentive was compounded by the decline in profitability of banks' traditional business; and that the incentives could be acted upon, given the degree of deregulation in bank lending, i.e. relaxation of "window guidance," in the early 1980s.
Some of these factors also contributed to the recent banking problems in other countries, e.g. Australia and New Zealand, Norway, Sweden, Korea, and the U.S.. This we should expect. Japan's banking industry has faced the same world-wide trend toward declining demand for banks' traditional services as have the banks in these other countries. Profit margins, which had been declining for Japanese banks since the early 1970s, were temporarily boosted in the late 1980s by the shift toward higher margin/higher ex ante risk loans. Our evidence is consistent with the hypothesis that the new BIS capital standards played an important part in this shift. For example, we find that on average the banks with the highest capital deficiencies vis-a-vis the then-impending 1992 BIS capital standard had the highest growth in risky loans and retained earnings.
The BIS capital standards require that Tier I plus Tier II capital be at least 8% of risk-adjusted asset exposure. The determination of Tier I capital, which is shareholders equity including retained earnings but excluding goodwill, and of Tier II capital, which includes items like loan reserves, is heavily dependent on accounting procedures. Moreover, the measurement of risk adjusted asset exposure involves the classification of assets into "buckets" which are each assigned a risk weight to multiply the book value of all the on-balance sheet assets in the respective buckets. This measure of risk-adjusted exposure is thus also heavily dependent upon accounting valuations, as well as the fiction that all assets falling within each of the broad asset classes have the same default risk.
In sum, both the numerator and denominator of the BIS capital-to-risk-weighted assets ratio are accounting based. Thus, faced with declining margins and the approaching BIS capital requirements, it is not surprising that banks would tend to make riskier loans with higher promised returns and thereby improve their capital positions so long as the higher expected defaults on riskier loans were not charged "up-front" against the loans. In fact, tax-deductible general reserves in Japan were a fixed 0.3% of domestic and OECD loan capital.
As the structure of loan risk changed, the fixed rate was not adjusted, or reserves accumulated, to compensate for the change in the pattern of likely defaults on the riskier, longer-term loans. This in spite of the fact that the increased risk seems to have been recognizable at the time, e.g. banks' credit ratings were successively downgraded from 1989 onwards. Even more ironically, if banks had had more freedom to choose the level of general reserves for loan losses, they would probably have had an incentive to over-provision, not under provision, because general reserves were both tax deductible and still counted as Tier II equity in satisfying the BIS requirements.
Download
PDF Ebook Japanese Banks' Bad Loans: What happened?
Posted in :
