Prediction on Fund Volatility Based on SVRGM-GARCH Model

摘要

GM-GARCH model is a new hybrid volatility model which integrates grey forecasting model (GM (1,1)) into GARCH model. As for the limitation of the parameters estimation algorithm of GM (1,1) model, a SVRGM-GARCH model is established to enhance volatility forecasting performance further. Firstly, support vector machines for regression (SVR) is utilized to estimate the parameters of GM (1,1) model (SVRGM). Then, the SVRGM model is used to modify the random error term sequence of GARCH model. An empirical research is performed on SSE Fund Index and SZSE Fund Index. The result shows that the SVRGM-GARCH model outperforms the GM-GARCH models and GARCH model, which indicates the model proposed in this study is an effective method for volatility forecasting.