【摘要】 比较了改进的遗传算法和BHHH算法对EGARCH模型的估计效果.结果显示,改进的遗传算法优于BH-HH算法.然后分别用在正态分布t、分布和广义误差分布(GED分布)假设下的EGARCH模型对上证综指进行实证分析.分析结果表明,EGARCH模型在厚尾分布假设下对序列的拟合效果比在正态分布假设下的效果好.上证综指收益率序列具有明显的杠杆效应.更多还原

【Abstract】 This paper estimates the parameters of EGARCH-t model by the improved genetic algorithm and the BHHH algorithm,then the results are compared.It shows that the former performs better than the later.Then the series of the return rates of Shanghai stock index is analyzed by EGARCH-n、EGARCH-t and EGARCH-GED models,respectively.The results show that EGARCH model with fat-tailed distribution is better than those with normal distribution, and the series of the return rates of Shanghai stock composite index has obvious leverage effect.更多还原