This paper investigates the housing-output growth nexus in Turkey by using quarterly data on real gross domestic product, real house prices and real gross fixed capital formation. The main aim of this paper is to identify indicators, for example, output and real capital formation, which potentially has strong causal relationship with housing prices in Turkey. The data set for this study spans the period; 1987:Q1 to 2016:Q2. Using full sample bootstrap Granger causality test and parameter

more »

... ty test, we examine the stability of estimated Vector Autoregressive Regressions (VAR), and thus determined whether the full-sample Granger causality inference might be invalid. When the stability of the VAR model is rejected, the Granger non-causality tests are not valid. For this reason, the bootstrap rolling window estimation technique was used to evaluate the Granger causality between the housing variables and the output growth rate. The bootstrap rolling window estimation results show that there is no Granger causality relation in full sample. However, in some sub-periods, these variables have predictive ability for each other.