摘要
引入ＡＲＣＨ模型族以突破传统经济计量学“同方差条件”的限制，和更加精确地动态度量代表金融风险的收益序列的条件“异方差”；针对ＡＲＣＨ传统估计方法的不足提出了利用遗传算法的改进算法；最后利用遗传算法实证建立了上海股市收益的各种ＡＲＣＨ模型；得出了上海股市“杠杆效应”、Ｉ－ＡＲＣＨ现象显著存在，Ａ－ＰＡＲＣＨ是上海股市收益“异方差”最合适的描述形式等有益结论．
In this paper,ARCH model class is introduced to solve the dilemma of “heteroskedasticity”and Genetic Algorithm is developed for some serious disadvantages of traditional estimating methods of ARCH.Finally the ARCH models for stock return in S.S.E.are established empirically and some helpful concerned results are concluded.
出处
《系统工程学报》
CSCD
1999年第1期23-28,共6页
Journal of Systems Engineering
基金
国家自然科学基金"九五"重大项目
关键词
ARCH模型族
异方差
遗传算法
上海
股市收益
ARCH model class,heteroskedasticity,genetic algorithm,stock return in S.S.E.
