最终结果为 CEFD序列变量(C,T,7)的情况下平稳,R序列在(0,0,0)的情况下平稳
我再使用直接根据软件默认的滞后2阶建立VAR模型
接着在估出结果的窗口view-lag structure-lag length criteria,
在弹出的窗口自己设置一个滞后长度,还是使用默认的滞后长度“8"
输出了AIC,SC,LR等各个准则在各个滞后期下的数值,带*的表示该准则下选择的最佳滞后期,最终结果如下图
VAR Lag Order Selection Criteria |
|
|
|
| ||
Endogenous variables: RT CEFD |
|
|
|
| ||
Exogenous variables: C |
|
|
|
| ||
Date: 04/08/15 Time: 13:36 |
|
|
|
| ||
Sample: 1 1457 |
|
|
|
|
| |
Included observations: 1449 |
|
|
|
| ||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Lag | LogL | LR | FPE | AIC | SC | HQ |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
0 | 5584.187 | NA | 1.54e-06 | -7.704882 | -7.697596 | -7.702163 |
1 | 8190.718 | 5202.268 | 4.25e-08 | -11.29706 | -11.27520 | -11.28890 |
2 | 8219.809 | 57.98262 | 4.11e-08 | -11.33169 | -11.29526 | -11.31809 |
3 | 8247.486 | 55.08691 | 3.98e-08 | -11.36437 | -11.31337 | -11.34534 |
4 | 8294.583 | 93.60913 | 3.75e-08 | -11.42386 | -11.35828* | -11.39939 |
5 | 8304.137 | 18.96172 | 3.72e-08 | -11.43152 | -11.35138 | -11.40161 |
6 | 8305.829 | 3.353286 | 3.73e-08 | -11.42833 | -11.33362 | -11.39299 |
7 | 8320.035 | 28.11764 | 3.68e-08 | -11.44242 | -11.33313 | -11.40164 |
8 | 8334.909 | 29.40081* | 3.62e-08* | -11.45743* | -11.33357 | -11.41121* |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* indicates lag order selected by the criterion |
|
|
| |||
LR: sequential modified LR test statistic (each test at 5% level) |
|
| ||||
FPE: Final prediction error |
|
|
|
| ||
AIC: Akaike information criterion |
|
|
|
| ||
SC: Schwarz information criterion |
|
|
|
| ||
HQ: Hannan-Quinn information criterion |
|
|
| |||
是否最优滞后期为8?还是说我验证的滞后长度为8,还可以继续增加设置的滞后期长度?
我之后再进行格兰杰因果检验,是不是应该用8的滞后期?


雷达卡




京公网安备 11010802022788号







