Random monitoring in financing relationships

This paper examines a financier's optimal monitoring intensity in a multi-period financing relationship. We identify conditions under which the financier should sometimes misidentify the quality of an entrepreneur. Such an imperfect evaluation technology affects action choices by bad entrepreneurs. We first characterize the optimal monitoring intensity and show that it is one in which the investor monitors entrepreneurs randomly. Random monitoring in the first stage of a relationship induces bad entrepreneurs to reveal their intrinsic types. Second, because random monitoring reduces the share of bad entrepreneurs in the subsequent periods, we show that the financier can therefore realize substantial gains.