金融计量赞学张成思lecture10-1.pptVIP

图10-7 EViews中ARCH效应检验结果 正态分布、t分布和广义误差分布对应的t时刻的对数似然函数分别为： 表10-3 标普500股票收益率的GARCH（1,1）估计结果－t分布结果 Dependent Variable: SP500RETURN Method: ML - ARCH (Marquardt) - Students t distribution Sample (adjusted): 1/05/1950 4/13/2007 Included observations: 14409 after adjustments Convergence achieved after 15 iterations Presample variance: backcast (parameter = 0.7) GARCH = C(3) + C(4)*RESID(-1)^2 + C(5)*GARCH(-1) Coefficient Std. Error z-Statistic Prob.?? C 0.000472 5.37E-05 8.790214 0.0000 SP500RETURN(-1) 0.110851 0.008388 13.21467 0.0000 Variance Equation C 5.54E-07 8.50E-08 6.518120 0.0000 RESID(-1)^2 0.069440 0.004231 16.41187 0.0000 GARCH(-1) 0.925040 0.004288 215.7174 0.0000 T-DIST. DOF 7.082672 0.333538 21.23496 0.0000 R-squared 0.004526 ????Mean dependent var 0.000349 Adjusted R-squared 0.004180 ????S.D. dependent var 0.008900 S.E. of regression 0.008882 ????Akaike info criterion -6.955589 Sum squared resid 1.136135 ????Schwarz criterion -6.952435 Log likelihood 50117.54 ????Hannan-Quinn criter. -6.954540 F-statistic 13.09545 ????Durbin-Watson stat 2.056719 Prob(F-statistic) 0.000000 表10-4 标普500股票收益率的GARCH(1,1)估计结果—ged分布假设 Dependent Variable: SP500RETURN Method: ML - ARCH (Marquardt) - Generalized error distribution (GED) Included observations: 14409 after adjustments Convergence achieved after 18 iterations Presample variance: backcast (parameter = 0.7) GARCH = C(3) + C(4)*RESID(-1)^2 + C(5)*GARCH(-1) Coefficient Std. Error z-Statistic Prob.?? C 0.000468 5.23E-05 8.952369 0.0000 SP500RETURN(-1) 0.102357 0.008171 12.52644 0.0000 Variance Equation C 6.03E-07 9.13E-08 6.602502 0.0000 RESID(-1)^2 0.072203 0.003671 19.66856 0.0000 GARCH(-1) 0.921922 0.003960 232.8286 0.0000 GED PARAMETER 1.372433 0.013545 101.3214 0.0000 R-squared 0.005049 ????Mean dependent var 0.000349 Adjusted R-squared 0.004704 ????S.D. dependent var 0.008900 S.E. of regression 0.008879 ????Akaike info criterion -6.947426 Sum squared resid 1.135537 ????Schwarz criterion -6.944272 Log likel