Studies of Stock Leverage Effect Model Based on Parallel Algorithm
Parallel algorithm was introduced into the estimation and forecasting of stock return model. In the stock return model,discrete event risk was described by random jump-diffusion process and the leverage effect was described by asymmetric Autoregression Conditional Heteroscedasticity model. The samples were classified into five groups according to the existing of day-of-the-week effect. The estimation and the simulation were processed by parallel computing in five groups of samples. The empirical study on index of Shenzhen security market shows that parallel algorithm can not only increase the computing speed largely but also make the accuracy.
Jump-diffusion process Leverage Effect Parallel Algorithm Simulated annealing Stock Market
Yu LIANG Xi-nan ZHAO Quan-bo GE
School of Business Administration, Northeastern University Shenyang, Liaoning, China Dept of Electrical Automation, Shanghai Maritime University, Shanghai, China
国际会议
杭州
英文
82-85
2006-10-12（万方平台首次上网日期，不代表论文的发表时间）