Lag Length criteria
|
VAR Lag Order Selection Criteria
|
|
|
|
|
|
Endogenous variables: OIL EXCH R RPI LUNEMP GDP
|
|
|
|
|
Exogenous variables: C
|
|
|
|
|
|
Date: 04/10/12 Time: 09:40
|
|
|
|
|
|
Sample: 1980Q1 2011Q4
|
|
|
|
|
|
Included observations: 90
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Lag
|
LogL
|
LR
|
FPE
|
AIC
|
SC
|
HQ
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
0
|
-1297.398
|
NA
|
152838.1
|
28.96439
|
29.13105
|
29.03160
|
|
1
|
-629.2589
|
1232.345
|
0.121419
|
14.91686
|
16.08344
|
15.38730
|
|
2
|
-546.9702
|
140.8050
|
0.043857
|
13.88823
|
16.05473*
|
14.76189*
|
|
3
|
-502.2073
|
70.62590
|
0.037065
|
13.69350
|
16.85992
|
14.97039
|
|
4
|
-469.5335
|
47.19560
|
0.042007
|
13.76741
|
17.93376
|
15.44753
|
|
5
|
-424.0268
|
59.66430
|
0.037071
|
13.55615
|
18.72242
|
15.63950
|
|
6
|
-377.6019
|
54.67826
|
0.033616
|
13.32449
|
19.49068
|
15.81106
|
|
7
|
-312.0832
|
68.43057
|
0.021261
|
12.66852
|
19.83464
|
15.55832
|
|
8
|
-247.3143
|
59.01173*
|
0.014906*
|
12.02921*
|
20.19525
|
15.32223
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* 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
|
|
|
|
|
|
|
|
|
|
|
|
Fig - Table of Lag Order Selection Criteria from Eviews
From Fig it can be seen that the AIC statistic minimises at a lag length of 8. Therefore, according to Akaike, this should be the lag length which is used in the VAR model. However it can also be seen that the SC and HQC both minimise at the lag length 2. As the lag length 2 satisfies more than one criterion, and therefore the majority of criteria assessed in this paper, this length will be chosen.This length should produce the best goodness of fit for the VAR model.