'GECO's Weather Forecast' for the U.K. Housing Market: To What Extent Can We Rely on Google ECOnometrics?

This study follows the stream of research identifying sentiment trends by using online search query data. The potential of the Google data series for the U.K. housing market on a disaggregated level is analyzed in a panel VAR framework. The findings confirm existing research based on U.S. samples that Google subcategories, especially "Real Estate Agency", serve as a robust indicator of transaction volume. Beside investigating how to deal with Google series in general and with heterogeneous cross sections specifically, the main contribution of this study to existing research is the detection of contrary dynamics within the Google "Home Financing" sub-category, which to date yields empirically mixed evidence (Hohenstatt, Kaesbauer and Schaefers, 2011). As expected from existing literature, sensitivity analysis yields that transaction volume responds twice as sensitively as house prices due to a standard deviation increase of the stress indicator. Most importantly, the derived stress indicator of housing market (un-)soundness works at least as well as in downturns, as opposed to "Real Estate Agency", which is primarily a suitable indicator during upturns.