An Unbiased Reexamination of Stock Market Volatility
指出标准波动性检验存在严重统计问题，提出在小样本中无偏且不要求平稳性的新检验，结果仍否定理性预期与恒定权益回报率模型。
Recent work demonstrates serious statistical problems with standard volatility tests. This paper proposes new tests that are unbiased in small samples and that do not require assumptions of stationarity. The new tests continue to find evidence against the model positing rational expectations and a constant required rate of return on equity.