Determinants of Housing Prices in Germany: Assessing Linear and Nonlinear Impacts of Selected Factors.

Abstract

This thesis examines the macroeconomic and financial determinants of real house prices in Germany, with an emphasis on how financing conditions shape housing-market dynamics in both linear and nonlinear ways. The empirical design follows a two-stage strategy that separates long-run equilibrium relationships from short-run shock transmission. In Stage 1, ARDL models are used to test for cointegration between real house prices and key fundamentals, including rents, mortgage credit, inflation and unemployment, while controlling for financing costs measured by the 10-year government bond yield. The baseline results show a stable rent–price linkage and a robust negative financing-cost channel, with mortgage credit affecting house prices mainly through lagged adjustments. A set of robustness checks strengthens these findings. Replacing the long-term yield proxy with the ECB mortgage lending rate for house purchase (Mortgage Rate Index, MRI) preserves the negative effect of borrowing costs. Adding a household debt stock measure (debt-to-GDP) does not materially alter the core relationships and provides limited incremental explanatory power at the quarterly frequency. Stage 1 further applies a NARDL specification to test for asymmetries. The evidence indicates that interest-rate tightening has a markedly stronger impact on real house prices than loosening, while nonlinearities in the credit channel are weaker. In Stage 2, a TVP-VAR framework tracks how the short-run effects of financial shocks evolve over time. Interest-rate shocks reduce housing-price growth, and the transmission becomes substantially stronger in recent tightening periods. Stock-price shocks display a more time-varying pattern, with larger positive responses after 2020. Overall, the results suggest that German house prices are shaped by a stable long-run rent–price relationship and a robust financing-cost channel, but the strength of financial transmission is nonlinear and regime-dependent. These findings are relevant for policy discussions on housing-market resilience under financial tightening when direct macroprudential indices are not readily available at quarterly frequency.