Publication in refereed journal
香港中文大学研究人员 ( 现职)
| 庄太量教授 (经济学系) |
全文
| 数位物件识别号 (DOI) http://dx.doi.org/10.1080/14697688.2010.481630 |
引用次数
Web of Sciencehttp://aims.cuhk.edu.hk/converis/portal/Publication/4WOS source URL
其它资讯
摘要Most of the existing technical trading rules are linear in nature. This paper investigates the predictability of nonlinear time series model based trading strategies in the U.S. stock market. The performance of the nonlinear trading rule is compared with that of the linear model based rules. It is found that the self-exciting threshold autoregressive (SETAR) model based trading rules perform slightly better than the AR rules for the Dow Jones and Standard and Poor 500, while the AR rules perform slightly better in the NASDAQ market. Both the SETAR and the AR rules outperform the VMA rules. The results are confirmed by bootstrap simulations.
着者Chong TTL, Lam TH
期刊名称Quantitative Finance
出版年份2010
月份1
日期1
卷号10
期次9
出版社Taylor & Francis (Routledge): SSH Titles
页次1067 - 1076
国际标準期刊号1http://aims.cuhk.edu.hk/converis/portal/Publication/469-7688
电子国际标準期刊号1http://aims.cuhk.edu.hk/converis/portal/Publication/469-7696
语言英式英语
关键词Financial markets; Financial time series; Forecasting ability; Forecasting applications
Web of Science 学科类别Business & Economics; Business, Finance; BUSINESS, FINANCE; Economics; ECONOMICS; Mathematical Methods In Social Sciences; Mathematics; Mathematics, Interdisciplinary Applications; MATHEMATICS, INTERDISCIPLINARY APPLICATIONS; Social Sciences, Mathematical Methods; SOCIAL SCIENCES, MATHEMATICAL METHODS
