This dissertation aims to answer two questions: i) whether main banks holdup their clients in Japan? ii) if the answer to the first question is yes, what are the consequences? With a non-linear specification, I find empirical evidence supporting the holdup hypothesis. Taking the bank's holdup behavior into consideration, I examine the effect of a close main bank relationship on the firm's investment and financing decisions in Japan. With sufficient control power over their clients before the Japanese deregulation, main banks had incentives to prod firms to over-invest as long as their loss from the share holdings was small relative to their profit from information monopoly (holdup profit). Such distortion was not very severe when there's limited supply of bank capital. But when banks accumulated sufficient capital, the values of those bank dependent firms got largely destroyed due to more overinvestment. After the deregulation, main hanks gradually lost their control right. For the benefit of existing shareholders, firms would like to finance those projects with more downside risk (lose more if fail) with bank loans. Unfortunately, the external shocks in early 90s made the bank financed projects end up with the worst situation. As the result, the Japan banks fell into a deep distress.