石油期货收益率的分位数建模及其影响因素分析

石油期货收益率的分位数反映了收益率分布特征和石油市场风险特征,有必要建模考察分位数的变化模式与影响因素。针对现有研究在模型方法和分析角度上的不足,本文考虑分位数受市场冲击影响而产生的非线性自回归特征,提出门限CAViaR模型并用以分析石油期货收益率的分位数及其影响因素。基于1998-2009年布伦特原油期货价格的研究表明,石油期货收益率的分位数具有自回归特征并受前期油价涨跌的不对称影响,且油价下跌的作用更强。左尾分位数受油价涨跌的共同影响,而右尾分位数仅受油价下跌的影响,二者呈现不同特征。此外,本文通过考察分位数的动态变化模式揭示了油价风险特征,具有重要的风险管理作用。

Modeling the Quantile of Oil Futures Return and Analyzing the Dynamics of Quantile

The quantile of oil futures return can reveal the character of distribution and measure extreme riskso, so it is necessary to model the quantile of oil futures return and analyze the dynamics of quantile. Considering the shortage of current research and the asymmetric influences of positive and negative oil return on quantile, threshold CAViaR model is developed and adopted to analyze the quantile of oil futures return. Using daily data of Brent crude oil futures price spanning from 1998 to 2009, the results show the quantile of oil futures return has autoregressive effect and is affected by lagged oil return. Besides, the positive return has stronger influence on quantile than the negative return. Oil price movement will affect the quantile of left tail, but only the decrease of oil price will make the quantile of right tail change. Moreover, the quantile is related to VaR, so the results can reveal the character of oil price risk and have important implications in risk management.