Abstract
The implementation of monetary policy and the final goal will result in the problem of time lag. The purpose of this article is to provide monetary authorities a short-term indicator as an aid to the intermediate goal. First, we choose monthly data from the foreign exchange market, money market, bond market and stock market in financial sector, then construct financial condition index(FCI) using principal component analysis and Deutsche Bank analysis, the most difference between the two approaches is that the former uses a large number of financial variables to extract the covariation or comovement and achieve the purpose of simplifying information while the latter constructs indicators with a linear regression model but consistent with economic intuition and theoretical expectations. Next, using the Granger causality test, VAR impulse response, and out-of-sample forecast to compare which FCI is more relevant to macroeconomic variables such as industrial production index, unemployment rate, house price index, and inflation rate. The empirical results show that FCI using principal component analysis has a leading or feedback relationship with most of the macroeconomic variables in the Granger causality compared with Deutsche Bank's analysis. In addition, the principal component analysis method for the unemployment rate and the inflation rate on out-of-sample forecasts outperform Deutsche Bank's analytical method and AR models, so FCI using principal component analysis that simplifies large amounts of information can indeed serve as a short-term indicator for the central bank's intermediate targets.