| This study tests empirically contracting theory in financial market. The model is based on Krasa and Villamil model (1992) which is modified to let each investor has k units of endowment, each bank has equity B, and monitoring cost depends on project size. In this model loan size per entrepreneur and the number of loans become choice variables (this makes the intermediary able to economize monitoring cost by investing more in one project). The modification allows us to determine the optimal bank size, the effect of a bank's assets on its return, and the effect of a bank's capital adequacy ratio on its asset portfolio and returns. The model is calibrated using Indonesian data, because there is no deposit insurance and banks are not restricted to branch. Thus, it is expected that the market distortion observed will not be significant. The result of the study is interesting, because the model can explain why many banks exist, predict real deposit rates and interest rate spreads. This research also finds that in general, a bank's size does not influence bank's return on equity. Finally, we can analyze the government policy to limit the bank's exposure to a single borrower or connected parties in Indonesia using the calibration result. |