基于MRS-GARCH模型的我国黄金现货价格波动研究

Study on Volatility of China's Gold Price Based on Markov- switching GARCH Model

黄金具有很好的抵御通货膨胀及规避投资风险的能力,近些年由于受美国经济持续低迷及全球量化宽松政策的影响,国际黄金价格波动加剧。我国的黄金市场已逐步放开,在提高我国在国际黄金市场地位的进程中,投资者同样面临着黄金价格波动的风险,因此对我国黄金价格波动的研究就十分必要。本文以上海黄金交易所Au99.99的日收盘价作为样本数据,利用MRS-GARCH模型对我国黄金现货收益波动进行分析,采用混合Gibbs抽样与EM算法的MCEM-MCML法估计模型参数,结果表明:我国黄金现货价格波动除了受到国际金价波动的影响外,其自身价格波动存在显著的高、低波动两种状态,处于低波动状态的持续时间更长,且收益波动更容易由高波动状态向低波动状态转换。依据BIC准则,MRS-GARCH模型的拟合效果优于Gray的模型和GARCH模型,最后给出了黄金投资的相关建议。

Gold can resist inflation and investment risk excellently. In recent years,America＇s economic slump and quantitative easing policy around the world heightened volatility of gold price. With China＇s gold price marketization,investors need to face the risk of gold price volatility. Based on daily closing price of Au99. 99 from Shanghai Gold exchange,we analyze spot gold price volatility with MRS- GARCH models. In order to void ＂path dependence problem＂,MRS- GARCH model is estimated with MCEM- MCML method. Empirical results suggest that there exist significantly two states in the gold market. The regime tends to change from high volatility state to low volatility state.Based on BIC criterion,MRS- GARCH model has the optimal fit. Finally we give some suggestions.