Research On Gold Price Volatility Based On GARCH Models

Gold is an internationally recognized hard currency, which has the dual attributes of currency and commodity. It is an important tool for investors to preserve and increase the value of assets. In recent years, affected by the European sovereign debt crisis, the situation in the Middle East and other factors, the price of gold to record high of sharp fluctuations uplink has been high and volatile. More and more investors change the currency into gold in value-added, many investors suffered huge losses in the severe shock of gold price. Research on the volatility of the price of gold and the characteristics of gold price fluctuations, effectively forecasting the future trend of the price of gold, has an important theoretical and practical significance.The article first introduces the major international gold market, and analyses various important factors affecting the gold price fluctuations. Then it selects some appropriate data and data preprocessing, and constructs predictive model through comparative analysis to do the model testing and prediction effect contrast.Multiple linear regression model is frequently used in gold price volatility prediction method, but this model does not consider the residual heteroscedasticity. Time series model is a model does not consider the impact of exogenous variables, the response of this model is also not sensitive to the outside world. Therefore, this article selects London Gold Exchange daily fixing price from January2,2002to February22,2013as a sample, and introduces GARCH model to deal with the conditional heteroskedasticity phenomenon in the price of gold sequence the greatest impact on the price of gold dollar index introduced in the mean equation as an exogenous variable, and the eventual establishment of EGARCH-X prediction model with exogenous variables. The empirical results show that EGARCH-X model is not only effective in achieving a black box function of the neural network, greatly improving the prediction accuracy, and can intuitively grasp the changes in volatility, and achieved very good results. The article conclusion that the volatility of the price of gold there are significant ARCH effect, fluctuations in the price of gold has a leverage effect, there is a very strong long memory and volatility.