ON THE ASYMPTOTIC POWER OF THE VARIANCE RATIO TEST
研究了当差分周期k随样本量n增大时，方差比检验统计量的渐近势函数，发现该检验对多种均值回归备择假设不一致，并通过模拟验证了结果。
The variance-ratio (VR) test statistic, which is based on k -period differences of the data, is commonly used in empirical finance and economics to test the random walk hypothesis. We obtain the asymptotic power function of the VR test statistic when the differencing period k is increasing with the sample size n such that k / n → δ > 0. We show that the test is inconsistent against a variety of mean-reverting alternatives, confirm the result in simulations, and then characterize the functional form of the asymptotic power in terms of δ and these alternatives.