Housing prices are subject to boom and bust episodes with long-lasting deviation from fundamentals. By considering a present value housing price model under noisy information, I study the macroeconomic implications of movements in housing prices related (news) and not related (noise) to future fundamentals. I provide empirical evidence of the sizable macroeconomic effects of news and noise shocks. Following Forni et al. (2014, 2016), I identify news and noise shocks through a non-standard VAR technique which exploits future information. In the US, news shocks are the main driver of the housing market at low frequencies, but in the short-medium horizon noise shocks explain a large share of the variability in housing prices, residential investment and GDP. Historically, many housing cycles are driven by noise. The empirical findings are consistent with a model à la Iacoviello which features a rental market. In this model, the usual optimal policy exercise concerns an augmented Taylor rule and a pro-cyclical loan-to-value ratio. I propose pro-cyclical property taxes as the most effective policy tool to deal with fluctuations originating from the housing market.
housing market macro-prudential noise non-fundamental VAR property tax