Distorted Advice in the Mortgage Market: Theory and Structural Estimation (with L. Guiso, A. Pozzi, L. Gambacorta, P. Mistrulli)

Abstract:
Complex financial decisions, such as choosing the mortgage type, require some degree of sophistication to form expectations about the future evolution of economic conditions. This provides banks, who often advise households, with an opportunity to steer unsophisticated households choices in a way that benefits them though not necessarily be in the household best interest. We set up a structural model of financial advice and estimate it exploiting administrative data on the universe of mortgages issued to Italian households. In the model, there are two types of households: Sophisticated households are perfectly informed and know which mortgage is best for them; naive households are instead susceptible to advice and will take the type of mortgage recommended by the bank. Each bank has an ideal mix of issued fixed and adjustable rate mortgages and attempts to achieve it by both appropriately setting the rates and strategically providing advice to its clientele. Banks compete in setting rates and households search. We recover the primitives of the model, in particular the share of “sophisticated” and “naive” households in the population, and use them to quantify the welfare implications of biased financial advice. In a counterfactual exercise, we study the effect of a financial literacy initiative which improves the general financial knowledge of the public reducing the scope for biased advice.