【摘要】 本文采用上证50 ETF及其期权交易数据,运用SVCJ模型、MCMC及傅里叶变换等方法,从P测度及Q测度中提取波动率风险溢价,并分析了其时变特征及影响因素。实证研究表明:SVCJ模型相较于SV模型及SVJ模型具有更好的市场拟合优度;傅里叶变换法能提高波动率风险溢价的估计效率;波动率风险溢价具有时变特征,在市场急剧动荡时期,波动率风险溢价基本为负,投资者厌恶波动风险,购买期权对冲波动风险的意愿较高;在市场非急剧动荡时期,波动率风险溢价基本为正,投资者偏好波动风险,购买期权对冲波动风险的意愿较低;市场收益率、波动率、换手率及投资者情绪对波动率风险溢价具有显著的影响。更多还原

【Abstract】 This paper uses the daily frequency data of Shanghai 50 ETF and ETF options. We use SVCJ model, MCMC and Fourier transform method to extract volatility risk premium from P measure and Q measure. The research shows that the SVCJ model has better market fitting goodness than the SV model and the SVJ model, and the Fourier transform method can improve the estimation efficiency. In the period of market turbulence, volatility risk premium is negative generally, and investors are risk-averse, who have a higher willingness to buy options to hedge volatility risk. In non-violent turbulence period, the volatility risk premium is positive generally, and investors generally act risk seeking. Market returns, volatility, turnover rate, and investor sentiment have a significant impact on volatility risk premium.更多还原