A Nonparametric Distribution-Free Test for Serial Independence in Stock Returns: A Correction
指出Ashley和Patterson（1986）提出的序列独立性检验存在构造错误，导致显著性概率被高估；作者给出了修正版本，并证明修正后的检验统计量具有与Kolmogorov-Smirnov检验相同的极限分布。
A fundamental statistical test of serial independence developed by Ashley and Patterson (1986) to examine a possible form of serial dependence in daily stock returns is shown to be improperly constructed. As a consequence, the significance probabilities that they ob? tain are overstated. This paper presents a corrected version of their test. The test statistic obtained after correction is shown to possess the same limiting distribution as the Kolmogorov-Smirnov test statistic. Applying the corrected test procedure to data identical to that used by Ashley and Patterson, we find that their original null hypothesis can no longer be rejected at conventional significance levels.