This paper puts forward some hypotheses to try and explain the recent fragility of the Japanese banking sector in a corporate-governance perspective. One hypothesis is that the banks themselves were not effectively monitored and disciplined. A second hypothesis is that the banks were able to show a good performance not because they monitored and disciplined their client firms, but because the firms were themselves disciplined by international competition. When those client firms reduced their reliance on bank credit in the 1980s, the banks were forced to extend their funds to non-traded-goods industries, such as real estate and finance, without having prepared themselves for this. The consequence was a serious non-performing-loan problem in the 1990s.