Fuzzy dual-factor time-series for stock index forecasting

摘要: There is an old Wall Street adage goes, ''It takes volume to make price move''. The contemporaneous relation between trading and stock returns has been studied since markets were first opened. Recent researchers such as Wang Chin [Wang, C. Y., & S. T. (2004). Profitability of return volume-based investment strategies in China's market. Pacific-Basin Finace Journal, 12, 541-564], Hodgson et al. [Hodgson, A., Masih, A. M. M., R. (2006). Futures a determinant prices different momentum phases. International Review Financial Analysis, 15, 68-85], Ting [Ting, J. L. (2003). Causalities the Taiwan Physica A, 324, 285-295] have found correlation markets. To verify this saying, paper, we propose dual-factor modified fuzzy time-series model, which take index forecasting factors predict index. In empirical analysis, employ TAIEX (Taiwan exchange capitalization weighted index) NASDAQ (National Association Securities Dealers Automated Quotations) experimental datasets two multiple-factor models, Chen's [Chen, (2000). Temperature prediction using time-series. IEEE Transactions on Cybernetics, 30 (2), 263-275] Huarng Yu's [Huarng, K. H., Yu, H. (2005). A type 2 model for forecasting. 353, 445-462], comparison models. results indicate that proposed outperforms listing models employed factors, technical indicator, VR(t), are effective