The Value of Bank Relationship: A Study on the Hokkaido Takusyoku Bank's Failure

Modern banking theories assume that banks have a specific part to play in managing some problems resulting from imperfect information on borrowers. To make their monitoring and information processing activities more efficient, banks and borrowing firms often develop long-term relationships, and some argue that banking relationships involve a degree of lock-in that makes it costly for borrowers to switch lenders (Sharpe [16]). Following several empirical success for the theories in the United States (James [12], Slovin, Sushka and Poloncheck [17]), Yamori and Murakami [18] tried to evaluate the economic value of Japanese bank relationship by investigating the stock price reactions against large scale bank failures in Japan, i.e., Hokkaido Takusyoku Bank (HTB) failure in Nov.1997. Their finding seems to roughly support the hypothesis that the Japanese bank relationships, well-known as the main bank relation, has an economic value. The purpose of our study is to empirically reassess the argument that the Japanese bank relationship has an economic value and main bank failure may cause client firms to incur a large economic loss. Although the preceding studies follow the conventional procedure of the standard event study, their attempt is just an indirect evaluation based on a compound hypothesis (with a sort of efficient market hypothesis), and the result hinges upon variations of limited number of observations. Through careful examination of firms' stock returns and financial statements around the HTB failure, we argue that the result of preceding studies overestimated the value of Japanese main bank relationship, and conclude that the value should not be overemphasized in the context of bailout policy discussion and systemic risk management. This paper consists of two parts. In the first part (section 1), we extend the study on stock returns around the HTB failure by Yamori et al., and show their findings are partly due to a certain compositional effect that is irrelevant to the value of bank relationship. In the second part (section 2 and 3), making full use of rich information contained in the financial statements of more than two thousands firms (including 400 unlisted firms in Hokkaido), we directly evaluate the effect of the HTB failure on client firms. Section 4 concludes.