Research On China 's Inflation Forecast Based On Financial Condition Index

With the development of the capital market, the real estate market, the financial system and structure, on the one hand, asset price volatility has brought new challenges to the monetary policy. On the other hand, the ability of the asset price to forecast the macroeconomic level has been improved. In recent years, the price of the house price has been improved, the volatility of the stock price has been increased and the inflation is very high. Thus, creating the FCI according to our own financial situation is very important to our monetary policy.After reviewing relevant literatures, we can find that most of the literatures use the variable of interest rate, real effective exchange rate and real stock price in the aspect of choosing index. In the aspect of calculating weight of index, we have four methods of choosing index. In the aspect of calculating weight of index, we have four methods including reduce aggregate demand equation formula, impulse response analysis, factor analysis and macroeconomic model. This paper referring to the previous acade mic approach to build FCI and based on the above four variables on the above four variables on the increase in the real money supply indicators to construct more in line with the financial conditions index China’s actual conditions. In the financial conditions index weight calculation method described above, we firstly constructs two weights financial conditions index which based on the total demand reduction for mula to calculate the static weight of the financial index(FCI_s) and then use of state of the total demand reduction formula based on spatial model building weight coefficient becomes financial conditions index(FCI_m). In contrast, the judge can better predict inflation(CPI, PPI) changes in the financial condition of the two weighted index which established. In the prediction process, not only to predict the samples also were predicted within sample, the analysis of financial condition index for inflation if the data inside and outside of the sample has a very good prediction.After thoroughly empirical analysis, we can make the conclusions as follows. Firstly, the index in different financial variables significant share of the weight gap, indicating the degree of influence of different variables on inflation was also significantly different. Integrated to reflect a country’s interest rates, exchange rates, the price of the underlying assets and the money supply and other financial variables financial condition index is more reasonable than the financial variables with a single, more comprehensive. Secondly, through the dynamic correlation analysis within the sample, Granger causality test and VAR impulse response analysis, the results of the analysis of the residual sum of squares model shows that monetary conditions index for the CPI, PPI in the sample have a good prediction. And FCI in PPI optimal prediction lag is greater than the CPI optimal lag, there is a difference may be due to 2004, consumer needs of the residents of the emergence of a significant change, thereby driving commodity prices lower upstream product prices change significantly. Third, the effect of FCI forecasting CPI and PPI is not very good out of sample. Through the comparative analysis of RMSE, we analysis whether FCI can improve the prediction accuracy to CPI and PPI. As the lag number is not fixed and prediction accuracy cannot be improved in all FCI lags, we cannot figure out the correct FCI lags to predict the inflation. Thus, there is limit in forecasting the inflation by using the FCI in practice, but can be used as a reference tool in making the policy. Forth, the forecasting effect of dynamic weight of FCI is better than the static weight of FCI in sample. However, we cannot conclude which one is better out of sample.