Residual Mortgage Debt, Insurance, and Defaults in the Netherlands

Working paper 824
Working Papers

Published: 16 December 2024

By: Madi Mangan Mauro Mastrogiacomo Hans Bloemen

Mortgage defaults are commonly linked to affordability and borrowers’ income; less often, to a decrease in home value. However, some studies talk about “strategic defaults”, a form of moral hazard whereby people who can afford their underwater mortgage choose not to pay. In this way, they clear their excess debt, as single recourse systems act as insurance. Our focus is on a type of mortgage insurance, available for houses with values below a certain threshold, that varies over time. We examine how this mortgage insurance affects decisions to default. We combine a quasi-natural experiment with the estimation of a structural model, more precisely an optimal stopping model. Our findings reveal that the (utility from) future value of home equity negatively influences the likelihood of default. We show that the discontinuity around the qualification threshold is linked to borrowers’ income, due to loan-to-income caps. The model indicates that while the insurance does not cause defaults in general, it does lead to more defaults for borrowers who separate from their partners, possibly indicating moral hazard.

Keywords: Residual Mortgage Insurance; Non-performance; Structural model; quasi-natural experiment
JEL codes G11; G21; G52

Working paper no. 824

824 - Residual Mortgage Debt, Insurance, and Defaults in the Netherlands

868KB PDF
Download 824 - Residual Mortgage Debt, Insurance, and Defaults in the Netherlands

Research Highlights:

  • We show a positive causal effect of the residual debt insurance (NHG) or reducing defaults on mortgages.
  • We show that the discontinuity around the qualification threshold is linked to borrowers’ income, due to loan-to-income caps.
  • We also find that the insurance leads to more defaults for borrowers who separate from their partners, possibly indicating moral hazard, as separation allows qualification for the insurance.
  • We combine a quasi-natural experiment with the estimation of a structural model, more precisely an optimal stopping model. Our findings reveal that the (utility from) future value of home equity negatively influences the likelihood of default.

Discover related articles