我国钢材期货市场波动率的GARCH族模型研究

Modeling Steel Futures Volatility Using GARCH Models

钢材是仅次于原油的全球第二大大宗商品,因此钢材及其相关产品价格波动的描述对各类参与者的套期保值及规避风险有重要意义。以上海期货交易所钢材期货价格的15分钟高频数据为对象,利用8类GARCH族模型进行了波动率拟合的实证研究,并运用6类损失函数以及Diebold-Mariano检验方法对各类模型的波动率拟合精度进行了比较。结果表明,能够刻画长记忆特征的HYGARCH模型在刻画我国钢材期货市场的波动率上具有相对优于其他模型的精度,但总的来说,各种模型并未表现出显著差异。

Steel is the world＇s second largest commodities after oil, so it important to describe and fit fluctuations in the price of steel and relevant product, and has important significance to the participants of the hedging and avoid the risk. With high frequency data for 15 minutes of steel futures price in the Shanghai futures exchange, then it is empirical research in the fit of volatility using 8 types of GARCH model, and using 6 types of loss function and Diebold-Mariano examination methods for compare the volatility fitting precision of various models. Results show that HYGARCH model which Can depict long memory characteristics have a relatively superior in fitting accuracy, but overall, the accuracy of all models do not show significant difference.