Research On Financial Market Volatility Based On Mixed Frequency Data

With the continuous changes in global financial markets,volatility has become an important reference index for assessing financial risks.Due to the continuous development of computer technology,it is no longer difficult to obtain intraday highfrequency data with rich information.Therefore,the traditional GARCH model using a single frequency no longer meets the research needs.In order to make full use of the obtained data information,it is necessary to use different frequency data in a model,that is,mixed frequency data model.This paper studies the volatility of the two major financial markets,the foreign exchange market and the stock market,using the RMB exchange rate data against the dollar and the SSE 50 index data as samples.The theoretical innovation of GARCHMIDAS model in this paper lies in : for short-term fluctuations,the GARCH model is upgraded to the GIR-GARCH model;for long-term trends,the single-factor model is upgraded to a two-factor model..In the selection of independent variables,GDP and economic policy uncertainty index(EPU)are innovatively used to fit the two-factor GARCH-MIDAS model based on realized volatility.According to the model parameter estimation results,when some of the selected independent variables change,the volatility of their respective markets will also change significantly.The evaluation results of the GARCH-MIDAS model show that the prediction effect of the GARCH-MIDAS model is better than that of the traditional model,and the addition of variables such as the monetary policy uncertainty index(MPU),the trade policy uncertainty index(TPU),and GDP can improve the GARCHMIDAS model.The predictive ability of the RMB exchange rate fluctuation;the addition of variables such as geopolitical risk index(GPR)and CPI can improve the predictive ability of GARCH-MIDAS model to the volatility of SSE 50 index.