高频股指期现货市场波动溢出效应的研究--基于EEMD的小波降噪

摘要

It is always a hotspot to use high-frequent data as the proof for the Volatility Spillover Effect in the spot and future markets in both theoretical and practical fields, but the high-frequent data is susceptible to the pollution of noise. Based on the disa-dvantage of high-frequent data, there are two methods come out with: first, verifying the validity, is to create a new method to reduce noise, like EEMD + Wavelet Soft Threshold Noise Reduction Method; second is applying the high-frequent data, which has already acc-epted the Noise Reduction, to the study of BEK-GARCH Model for the Volatility Spillover Effect. Both spot and future markets contain larger Volatility Agglomeration Effects , after verifying that the reductions of noise influence from microstructure. The spot markets in China influence on the future markets in short and long terms, but future markets only have effects on the volatility of spot markets in short term. Also, the influence of the future markets to spot markets is greater than that of the spot markets on the futures markets in short term. Overall, the impact of spot markets to futures markets from stock index in China is sustainable.%利用高频数据进行股指期现货市场波动溢出的研究一直是理论界和实务界研究的热点，但是高频数据易受噪音污染，基于此，文章在这些方面做了一些研究：一是提出新的降噪方法， EEMD+小波软阈值降噪法，并验证了其有效性。二是把降噪后高频数据应用于BEKK-GARCH模型研究波动溢出。经验证，降低微观结构噪音影响后，两个市场均存在较大的波动率聚集效应，我国现货市场的冲击在短期、长期均会对期货市场产生影响，期货则仅在短期对现货市场波动有影响，但短期内期货市场对现货市场的影响大于现货对期货市场的影响。总体上看，我国股指现货市场对期货市场的影响较持久。