具有结构突变的CAPM的阶段异方差和自相关性的调整LM检验

由于金融市场是动荡不定的,资产定价模型CAPM往往会出现结构突变,异方差,序列相关,因此需要对CAPM的随机误差进行齐性检验。对于具有单个结构突变点的CAPM,本文得到了检验阶段异方差和自相关性的调整LM检验统计量。Monte Carlo模拟的结果显示,该调整LM检验统计量具有比普通LM检验统计量更好的检验功效。最后,我们用一个具体的实例论证了方法的有效性。

Adjusted Testing for Phased Heteroscedasticity and Autocorrelation in CAPM with Structural Change

Due to the volatility and unstability of financial market,the CAPM often suffers from structural change,heteroscedasticity and serial correlation.Thus,it is necessary to check the assumption of homoscedasticity of random errors.As to the CAPM with single change point,this paper devotes to deriving the adiusted LM statistics for phased heteroscedasticity and autocorrelation.Monte Carlo simulation shows that the adjusted LM statistics can obtain better statistical test power than ordinary LM statistics.At last,the properties of these statistics are illustrated through a real example.