House Prices as Indicators of Monetary Policy: Evidence from China

This paper assesses empirically whether China's central bank should react to house prices and if so how. We use three kinds of VAR models including structural VARs with a combination of short-run and long-run restrictions to solve the endogeneity problem of identifying shocks to monetary policy and house prices. Broader money supply (M2) and the one-year lending rate are used as monetary policy proxies according to the distinctive background in China. The interaction between M2 and house prices is much more evident than the effects of M2 and house prices on GDP and CPI. In the three-variable VARs, GDP reacts to the M2 shock more considerably than it does in the four-variable SVARs. In the structural identification model, M2 and interest rates respond to a house price shock more sensitively than they do in the Cholesky identification. Following a tight money shock, under structural identification, output decreases to a lesser extent and CPI falls slowly, which lessens the negative impact on the macroeconomy compared with the Cholesky identification. Relevant policy implications are also discussed.