论文部分内容阅读
期货交易的高杠杆率意味着期货市场的高风险特征,而能源市场因其特殊的战略意义一直以来备受关注,因而对能源期货市场的风险测度对投资者和监管者都极其重要。本文对上海燃油期货构建了四个反映不同交割期限的连续价格序列,基于不同的金融市场典型事实分别运用GARCH、GJR、FIGARCH三个模型对波动率建模,并假设条件收益分别服从正态、学生t、有偏学生t(skst)分布进行动态风险价值(VaR)测度,然后运用严格的似然比(LR)检验和动态分位数回归(DQR)检验对风险测度的可靠性进行后验分析(Backtesting),尝试从中提取出在风险管理中最有应用价值的典型事实。研究发现:(1)基于skst分布的波动模型的动态风险测度准确性明显优于其他分布下的相同模型;(2)基于杠杆效应的GJR模型和基于长记忆性的FIGARCH模型并没有表现出比普通GARCH模型更高的精度;(3)远期合约的市场平均收益更高,风险测度比近期合约更准确。
High leverage in futures trading implies high-risk characteristics of the futures market, and the energy markets have drawn so much attention for their particular strategic importance that the risk measurement of the energy futures market is of crucial importance to both investors and regulators. In this paper, four consecutive price series reflecting different delivery deadlines are constructed for Shanghai Fuel Futures. Based on the typical facts of different financial markets, GARCH, GJR and FIGARCH models are respectively used to model the volatility. Assuming that the conditional returns obey the normal, Students t, with skst distribution of the dynamic VaR measure, and then use the strict likelihood ratio (LR) test and dynamic quantile regression (DQR) test to test the reliability of the risk measure Backtesting, trying to extract the typical facts that are most valuable in risk management. The results show that: (1) the accuracy of dynamic risk measure of the volatility model based on skst distribution is obviously better than that of other models under the same distribution; (2) the GJR model based on leverage and the FIGARCH model based on long memory do not show the ratio The higher accuracy of the general GARCH model; (3) the average market return of forward contracts is higher, and the risk measure is more accurate than recent contracts.