Low repayment rates on mortgages have probably contributed to increased macroeconomic imbalances for several countries in recent years. The objective of this paper is to analyze how the introduction of interest-only-mortgages in 2003 affected the Danish housing market. Using quarterly data between 2001 and 2013, we analyze the long- and short-run relationships between mortgage repayments and three dependent variables: real house prices; the number of property sales; and mortgage debts in relation to disposable income. We find that including mortgage repayments in nominal housing payments improves the econometric fit when predicting house prices and mortgages in relation to disposable incomes in both the long run and the short run, but not when modeling sales. Also, the conventional user cost variable either turn out insignificant or with the “wrong” (i.e. positive) sign, while nominal housing payment, including mortgage repayments, are significantly negative. However, some regression results are sensitive to specification and there are some indications of reversed causality issues. We therefore perform complementary event studies. The results of the event studies suggest that real house prices and the mortgage debt ratio increased considerably faster than otherwise expected between the introduction of interest-only-mortgages and the start of the financial crisis in 2007. Results are similar when using the results from our long- and short-run regressions to simulate the effects of introducing interest-only-mortgages. Our most conservative estimates suggest that the introduction of interest-only-mortgages raised real house prices and the mortgage debt ratio by about twenty and ten percent between 2003 and 2013, respectively. Our results raise questions of whether households base their decisions on real or nominal variables, actual costs, or cash flows, and if interest-only-mortgages have promoted housing affordability or have rather contributed to increased house prices and household debts.
Read full abstract