使用的是winrats9,电脑是64位的。如果用winrats Pro(64-bit)估计出来的结果与winrats Pro(32-bit)的结果不同。而且在用64位的估计BEKK模型时,使用同一个数据、同一个模型,有时会给出两个不同的结果,如下所示:
MV-GARCH, BEKK - Estimation by BFGS
Convergence in 110 Iterations. Final criterion was 0.0000075 <= 0.0000100
15. C(1,1) 0.001188722 0.000316781 3.75250 0.00017508
16. C(2,1) 0.000828615 0.000731441 1.13285 0.25727561
17. C(2,2) 0.001208965 0.000196066 6.16610 0.00000000
18. A(1,1) 0.111911767 0.080993066 1.38175 0.16705000
19. A(1,2) -0.317766416 0.096058509 -3.30805 0.00093948
20. A(2,1) 0.099933958 0.073740391 1.35521 0.17534961
21. A(2,2) 0.511859283 0.088845777 5.76121 0.00000001
22. B(1,1) 1.011360096 0.027466636 36.82140 0.00000000
23. B(1,2) 0.119050857 0.042069967 2.82983 0.00465727
24. B(2,1) -0.035337393 0.027221747 -1.29813 0.19424233
25. B(2,2) 0.858085793 0.040569884 21.15081 0.00000000
MV-GARCH, BEKK - Estimation by BFGS
Convergence in 98 Iterations. Final criterion was 0.0000000 <= 0.0000100
15. C(1,1) 0.001191292 0.000363351 3.27863 0.00104313
16. C(2,1) 0.000828215 0.000728118 1.13747 0.25534055
17. C(2,2) 0.001209950 0.000192787 6.27610 0.00000000
18. A(1,1) 0.112674945 0.081496001 1.38258 0.16679292
19. A(1,2) -0.317427971 0.097288724 -3.26274 0.00110340
20. A(2,1) 0.099279881 0.071489006 1.38874 0.16491081
21. A(2,2) 0.511645572 0.086414976 5.92080 0.00000000
22. B(1,1) 1.011068414 0.027028108 37.40803 0.00000000
23. B(1,2) 0.118964545 0.040162980 2.96204 0.00305603
24. B(2,1) -0.035058031 0.026164738 -1.33990 0.18027911
25. B(2,2) 0.858158894 0.038258006 22.43083 0.00000000
请问一下为什么会这样,是我哪里弄错了吗?该怎么办呢