基于趋势平滑和GARCH 的证 券市场预测.pdfVIP

( 2002) . 18 . 1 Journal of Xi an University of T echnolo y Vol No 93 : 1006-47 10( 2002) 01-0093-05 GA RCH 1 1 2 王炳雪, 史忠科 , 阎东明 ( 1. , 7 10072; 2. , 7 10048) : 首先将证券市场运动用局部多项式趋势模型进行平 , 然后分别用AR 模型和 GARCH 模型考虑序列之间自相关性和波动的变化性。参数的条件最大似然估计应用了状 态空间模型的卡尔曼滤子递推和 GARCH 模型的条件方差递推, 模型阶数的选取应用了 Akaike 的最小化信息矩阵方法。计算实例表明了这种组合方法预测能力的优越性。 : 证券市场; 多项式趋势模型; 状态空间模型; 卡尔曼滤子; GARCH 模型 : F830. 9: A The Prediction of Security Market based on Trend Smoothing and GARCH 1 1 2 WANG Bin -xue , SHI Zhon -ke , YA N Don -min ( 1. Northwestern Polytechnical Universit y, Xian 710072, China ; 2. Xian University of T echnolo y , Xian 7 10048, China) : Abstract T he operation of securities market is first smoothed by local polynomial trend model, then AR and GARCH models are used to study the variations in auto-correlation and fluctuation between time series. T he state space Kalman filter recursive computation and GARCH model s conditional variance r ecur sive computation ar e used to estimate the . conditional likelihood of parameters Akaike s minimum AIC procedure is used to select the best model fitted to the data w ithin and betw een the alternative model classes. The calculated real examples indicate that this combination method has distinct superiority in forecast . Key words: secur ities market; polynomial trend mode; state space model; Kalman filter ; GARCH model , , [ 1] , [ 2, 3] , , , , , , Kita aw a G, Gersch W[ 4, 5] Hiroshi T [ 6] , ,