沪铜期货市场波动聚集现象研究

摘要

By applying autocorrelation functions and GARCH model methods, we conduct empirical analysis to logarithmic return time series of copper future contract in Shanghai Future Exchange ( SHFE) when time-scale τ is 1,5,22,66. We also introduce an index through moving-window method to quantitatively measure the volatility clustering degree in these time series and research relative features of volatility clustering. We find that, volatility clustering exist in all of these various time-scales return time series, the degree of volatility clustering increases with the growth of time range and moving-window, volatility clustering of copper contract price in SHFE can account for the slow decay behavior in autocorrelation functions of copper contract absolute return time series. These facts indicate that the price volatility of copper contract price in SHFE shows self similarity and copper future market has fractal and memory feature.%　　应用自相关函数检验法和GARCH模型检验方法，对时间标度τ=1，5，22，66下的沪铜期货收益率序列进行实证研究，通过移动窗口法引入了波动聚集指数度量沪铜期货市场的波动聚集程度，并对其波动聚集的相关特征做了研究。结果发现：不同时间标度的沪铜连续合约收益率序列都存在波动聚集现象，铜期货表现出自相似的特征；随着时间跨度和移动窗口的增大，波动聚集程度也随之变大；沪铜期货价格的波动聚集是沪铜连续合约收益率序列自相关函数慢衰减的原因。从而表明沪铜期货的价格波动表现出较强的自相似性，沪铜期货市场具有分形特征、存在记忆性。