The author develops a model of bank-firm relationships on the basis of the following general idea: Banks want to prevent moral hazard on the side of their customers. In particular they want to prevent their business customers to use bank credit for purposes different from those that have been negotiated thus damaging the bank's interest. The idea of this model is relatively simple. Banks do not extend a loan if the project for which the money is intended will probably be un­ profitable. They extend the loan if the success of the project is highly probable and if the revenues from that project are greater than the expenses of the bank for monitoring the customer. Assuming as Miarka does that the results from a successful project are certain, this model is an equivalent to minimizing moni­ toring costs. In fact, this is the outcome of the model. The banks are known to monitor their loans. They thereby signal to the capital market that they have tested the project. Therefore, the buyer of bonds of the company on the capital market may rest assured that the project is financially sound. The buyers of bonds thus avoid monitoring costs and can grant better credit conditions than the banks. Pur­ chasers of bor. . ds are free riders on the monitoring of the banks. Miarka tests his model econometrically. The results are amazingly supportive of the model.