Assessing mortgage default risk in full-recourse economies, with an application to the case of Chile

This paper is aimed at finding individual and systemic determinants of housing default in a full-recourse economy. We propose an empirical strategy that accounts for rare events and choice-based sampling bias and estimates the contribution of idiosyncratic and systemic determinants, as well as their interactions. To support our empirical approach, we adapt and extend a well-known model of mortgage default for nonrecourse economies (Geanakoplos and Zame, 2014), adding a nonpecuniary cost for defaulting to account for possible loss of utility due to the full-recourse framework. This model applies to economies such as Spain, Australia, and Chile, where defaulters can be prosecuted until their debts are completely settled. Under the proposed model, we obtain an analytical expression involving default determinants for micro and macro mortgage loans. As a case study, we estimate this relationship for the Chilean economy using information from the Chilean Survey of Household Finance (EFH). As stated by our micro–macro model, household financial conditions and their interactions with systemic determinants account for an important part of the cross-sectional probability of mortgage default.