Every year, the Ieke van den Burg prize recognises outstanding research conducted by young scholars on a topic related to the macroprudential mission of the European Systemic Risk Board. The members of the Advisory Scientific Committee decide on the winner. This year’s award goes to Marcus Mølbak Ingholt from the Danish National bank for his paper “Multiple Credit Constraints and Time-Varying Macroeconomic Dynamics”. It is a great piece that combines theory and empirical work to advance our thinking on the use of different macroprudential tools across the business and credit cycles.
Among the many macroprudential tools that have been the focus of research, two borrower-targeted policies – the loan-to-value (LTV) and the debt-to-income (DTI) ratios – are often interchangeably used. But more likely than not, only one of them is actually binding. Marcus constructs a tractable New Keynesian dynamic stochastic general equilibrium (DSGE) model with long-term fixed-rate mortgage contracts and two occasionally-binding credit constraints: an LTV constraint and a DTI constraint.
Calibrating his model for the U.S. economy between 1984-2019, Marcus finds the LTV constraint is more likely to bind during and after recessions, when house prices are relatively low, while the DTI constraint mostly binds in expansions, when interest rates, which impact debt service, are relatively high. This results in the policy implication, that a countercyclical DTI limit would be effective in curbing increases in mortgage debt, as these increases typically occur in expansions, while a countercyclical LTV limit cannot prevent debt from rising. Countercyclical LTV limits can, however, mitigate the adverse consequences of house price slumps on credit availability by raising credit limits.
Marcus further uses a county-level panel dataset covering 1991-2017 across the US to test two key predictions for homeowners facing both LTV and DTI requirements. The predictions are that income (house price) growth predicts credit growth if homeowners’ housing-wealth-to-income ratio is sufficiently high (low), as they will be DTI (LTV) constrained. His results provide evidence for these hypotheses.
1. Oct, 2020