Research On The Forecast Of Stock Market Volatility Using The Uncertainty Of Economic Policy

The stock market plays a huge role in functions such as financing,regulation,and signaling.It is an indispensable part of the capital markets of various countries today.Therefore,the research on stock market volatility prediction is of great significance.The EPU index characterizes the uncertainty of economic policies,and its non-linear structural transformation characteristics are important considerations for forecasting volatility.When predicting stock market volatility,it is of great significance to consider the non-linear structural transformation characteristics of the EPU index.GARCH and its extended models are mostly used in stock market volatility prediction research,but such models can only fit a single frequency data.Macro variable data are mostly monthly and quarterly data,which are different from daily stock data.The GARCH-MIDAS model proposed by foreign scholars which can just solve the problem of different frequency data.Firstly,this paper considers the conversion characteristics of EPU index nonlinear structure,and introduces it into the GARCH-MIDAS model,constructing a new type of model: STR-GARCH-MIDAS-EPU model;then give the complete construction process of this new model including the nonlinear test of EPU index on stock market volatility and parameter estimation;finally,the STR-GARCH-MIDAS-EPU model is applied to the research of volatility prediction,and MCS test is used to compare the volatility prediction ability of the model with the GARCH-MIDAS-EPU model and the standard GARCH-MIDAS model.The results show that the STR-GARCH-MIDAS-EPU model has significantly improved the ability to predict volatility based on these two models.The traditional GARCH-MIDAS-EPU model only considers the linear structure when considering the impact of the EPU index on the stock market volatility,but ignores the existence of nonlinear structural conversion characteristics,so it is difficult for this model to accurately describe the impact of the EPU index on volatility.Aiming at this problem,this paper improves the long-term volatility component of the model,considers the nonlinear structural transformation characteristics of the EPU index,and constructs the STR-GARCHMIDAS-EPU model.This model divides the fluctuation of stock market volatility into two parts: one is the long-term fluctuation component considering the influence of non-linear structural changes about the EPU index,and the other is the short-term volatility component that uses the yield to fit the GARCH(1,1)process.This paper considers the non-linear structural transformation characteristics of macro variables such as the EPU index and introduces it into the GARCH-MIDAS model.A new type of extended GARCH-MIDAS model is constructed: STR-GARCH-MIDAS-EPU model.The advantage of this model is that it can characterize the impact of EPU index non-linear structural transformation characteristics on financial market volatility.Applying the model to the study of S&P500 index volatility prediction,the results of goodness-of-fit test and out-of-sample volatility prediction test show that the model can effectively improve the stock market volatility prediction ability.