高频股指期现货市场波动跳跃及跳跃溢出检验——基于集合经验模式分解和小波降噪

发布时间：2018-04-23 22:02

本文选题：EEMD + 小波分析　；参考：《系统科学与数学》2017年06期

【摘要】：选用EEMD+小波软阈值降噪法对金融高频数据进行降噪处理,然后利用仿真数据和实证方法验证了该方法的有效性,最后把降噪后的数据进行跳跃和跳跃溢出检验.研究表明:1)EEMD+小波降噪方法可以较好的降低金融高频数据的噪音.2)大部分情况下CSI300股指期现货市场对信息的反应不充分,期指更容易出现信息反映过度,两市场间有共同跳跃,跳跃溢出显著.3)通常认为噪音的存在使得期指对新信息的冲击更敏感,跳跃频率、幅度更大,会领先于现货市场发生跳跃,降噪后两市场的跳跃特征无明显变化,所以噪音不是股指期现货市场具有这些跳跃特征的唯一因素.
[Abstract]:The EEMD+ wavelet soft threshold denoising method is used to denoise the financial high frequency data. Then the effectiveness of the method is verified by the simulation data and the empirical method. Finally, the data after the noise reduction is tested by jumping and jumping spillover. The research shows that: 1) the EEMD+ wavelet denoising method can reduce the noise.2 of the financial high frequency data better). In some cases, in the CSI300 stock market, the spot market is not fully responsive to information, it is easier to show excessive information, two markets have a joint jump, jump spillover.3.) the noise is usually considered to be more sensitive to the impact of the new information, jumping frequency, and greater amplitude, leading to a jump in the spot market and noise reduction. There is no obvious change in the jump characteristics of the last two markets, so noise is not the only factor that has the characteristics of these jumps in the stock index spot market.

【作者单位】：新疆财经大学金融学院;新疆财经大学应用经济学博士后流动站;中央财经大学金融学院;
【基金】：国家自然科学基金面上项目(71471182);国家自然科学基金项目(71661028,71401188) 中央财经大学青年科研创新团队中国博士后第61批面上项目新疆财经大学校级(2015XYB018) 新疆自治区普通高校人文社科重点研究基地社会经济统计研究中心招标(050315C05)资助课题
【分类号】：F224;F832.51;F724.5

【相似文献】