Erik Heitfield

Monitoring, Moral Hazard, and Market Power: A Model of Bank Lending

We model the relationship between market power and both loan interest rates and bank risk without placing strong restrictions on the moral hazard problems between borrowers and banks and between banks and a government guarantor. Our results suggest that these relationships hinge on intuitive parameterizations of the overlapping moral hazard problems. Surprisingly, for lending markets with a high degree of borrower moral hazard but limited bank moral hazard, we find that banks with market power charge lower interest rates than competitive banks. We also find that competition makes banking industry risk highly sensitive to macroeconomic fluctuations by making banks more vulnerable to borrower moral hazard. This finding offers an explanation for the dramatic rise and subsequent decline in bank failure rates during the 1980s and 1990s.

Parametric, semi-parametric and non-parametric models of telecommunications demand: An investigation of residential calling patterns

Abstract We investigate long distance telephone calling patterns using billing information and demographic data for a large cross-section of residential households. The joint distribution of the number and duration of toll calls is estimated using a non-parametric kernel model, a semi-parametric Cox proportional hazard model, and a fully parametric Poisson–Weibull model. Particular attention is paid to the estimation and analysis of call duration hazard functions. We find that call duration is quite inelastic with respect to price, and that while evening and night/weekend rate calls share vary similar duration characteristics, they differ substantially from peak rate calls.