金融资产收益非对称性分析

滑动分块自助法(moving-block bootstrap)是一种适合相依时间序列的再抽样方法,与传统的非对称性检验法相比,它能有效地拟合统计量的抽样分布,这是因为传统的非对称性检验均假设收益序列独立同分布,与大多数金融资产收益的统计特征不符;利用滑动分块自助法对中国股市指数进行实证分析可以知道,在“熊市”期间,收益显著右偏,而在“牛市”期间,非对称性不明显。由于金融资产收益的非对称性是影响投资组合、资产定价的重要因素,因此在金融理论与实践中具有重要意义。

An Analysis of Asymmetry for Financial Returns

The moving-block bootstrap is a re-sampling method for dependent time series,which can mirror the behavior of the statistics of interests.Asymmetry of financial returns is important in theory and practice of finance and economics.It is an important factor for investment portfolio and assets pricing.The traditional methods for asymmetry testing are all based on the independence of financial returns,which are inconsistent with the stylized factors.In this paper,we introduce moving-block bootstrap asymmetry testing.Our empirical analysis using China stock indexes shows that it is a positive skew in a 'bear market' and symmetric in a 'bull market'.