Guido Wolswijk

Determinants of Mortgage Debt Growth in EU Countries

Abstract This paper presents an analysis of mortgage debt growth in the EU-15 countries. Mortgage debt has risen quickly in many countries in recent years, reaching historically very high levels that increasingly attract the attention from both researchers and policymakers. While there are some papers on factors behind national mortgage debt developments, studies taking a broader approach, both regarding the number of countries included and the potential determinants of mortgage debt growth, are still lacking. This study aims at filling this gap and presents estimates for real mortgage debt growth for the EU-15 countries, using pooled regressions for the period 1982–2003. Special attention is paid to properly defining mortgage costs, for which an after-tax measure is constructed. The regressions indicate that real mortgage debt is affected by after-tax mortgage interest costs, by house prices, financial deregulation, and stock markets, while the effects of household disposable income and inflation are less evident. These results are shown to be relatively robust using alternative time periods, estimation techniques, and groups of countries, only Greece being somewhat of an exception. Finally, we pay attention to tax instruments for influencing housing market developments, notably the deductibility of mortgage interest payments from personal income tax. The paper concludes that tax instruments have the potential to contribute to more stable mortgage debt developments.

Have Euro Area Government Bond Risk Premia Converged To Their Common State

We derive a model in which a standard international capital asset pricing (ICAPM) model is nested within an ICAPM model with market imperfections. In the latter model an idiosyncratic stochastic factor affects the return of risky assets (over a risk-free rate) on top of the systematic component that is common to all countries (and that is interacted with a timevarying idiosyncratic “beta”). We introduce asymptotic convergence from the full ICAPM model with imperfections to the standard model by multiplying the idiosyncratic factor by convergence operators. The model is then estimated using the weekly 10 year government bond spreads of Belgium, France, Italy, and the Netherlands versus Germany over the period 1991-2006. We find that the idiosyncratic components have converged towards zero for all countries after the introduction of the euro implying that the efficiency of the euro area government bond markets under consideration has increased. Full convergence has not yet occurred however.