This paper considers the dynamic features of housing prices in metropolises that are characterised by a high degree of internationalisation. Using the wavelet coherency procedure, the degree of comovement and causality between housing, stock markets, and macroeconomic uncertainty are investigated. In addition, the existence of volatility spillover across housing markets is assessed in the time-frequency domain using a novel procedure that involves combining the wavelet decomposition with a time varying parameter vector autoregression model. The results highlight that the clustering of global business in a limited number of metropolises that act as "global hubs" leaves the local housing markets exposed to international shocks and volatility spillover. The empirical analysis suggests that the correlations between real estate and stock markets from one side, and real estates and uncertainty on the other side, intensify during the turmoil periods. Causality and co-movement relationships appear predominately in the medium and long run periods. The evidence presented in this paper suggests that policymakers cannot ignore the possibility that international shocks to housing markets may affect the domestic markets. In this respect, macroprudential policy tools may target tapering off the unintended effects of housing market globalisation such as house price shock synchronisation, especially when these shocks take place in cities that are also major financial centres.