Hello,
I estimated a VECM in Eviews and R using urca package's ca.jo(), cajorl()
and vec2var() functions.
Specifications are 'no trend' in Eviews and 'none' in R (no
theory, just
testing, feel free to make changes).
Results are different, ecm and cointegrating vectors are completely
different.
R code is:
*johcoint=ca.jo(Ydata,type="trace",ecdet=c("none"),K=2,spec="transitory")
summary(johcoint)
vecm.r1=cajorls(johcoint,r=1)
vecm.r1
vecm.l=vec2var(johcoint,r=1)
ll=irf(vecm.l, impulse = "B",response = "A", boot = FALSE)
plot(ll$irf[[1]])*
Data:
A B C D
1 8.646924 3.925155 2.297737 2.764267
2 8.643810 4.048215 2.140731 2.769231
3 8.634732 4.117114 2.063724 2.747604
4 8.603337 3.976002 1.939290 2.741640
5 8.604344 3.924697 1.928255 2.732419
6 8.628887 3.921517 1.878674 2.718437
7 8.653167 3.906076 1.943236 2.693620
8 8.661854 3.940468 2.107718 2.670370
9 8.609839 3.872782 2.003064 2.689212
10 8.614091 3.905839 1.973719 2.679186
11 8.613692 3.890797 1.939311 2.659350
12 8.651488 4.052423 1.961038 2.640751
13 8.654469 4.137534 2.130873 2.622611
14 8.693121 4.074753 2.108427 2.595760
15 8.699435 3.872412 2.091816 2.622049
16 8.808724 3.851373 2.345740 2.646252
17 8.814437 3.806048 2.057104 2.728953
18 8.836529 3.743046 1.825827 2.748266
19 8.826898 3.693897 1.823880 3.027604
20 8.809117 3.673126 2.020016 2.820051
21 8.654972 3.652903 1.523249 2.538225
22 8.515917 3.659592 1.617734 2.523293
23 8.589919 3.655822 1.827645 2.371598
24 8.595193 3.645937 1.825603 2.251557
25 8.615332 3.629201 1.661946 2.254364
26 8.671222 3.609464 1.733073 2.145093
27 8.611882 3.612110 1.794937 1.819291
28 8.688414 3.579205 1.505888 1.654666
29 8.690125 3.554958 1.426589 1.731257
30 8.725932 3.533288 1.522311 1.788969
31 8.743279 3.527591 1.601261 1.760313
32 8.694805 3.531611 1.634085 1.732271
33 8.687983 3.527327 1.601985 1.836593
34 8.716976 3.514645 1.589035 1.745653
35 8.775464 3.492427 1.471562 1.699377
36 8.808898 3.471036 1.460162 1.686131
37 8.842847 3.451130 1.579547 1.670513
38 8.871786 3.428002 1.597216 1.618989
39 8.907425 3.423887 1.626208 1.652055
40 8.924721 3.403657 1.578576 1.509779
41 8.941122 3.357645 1.521236 1.607082
42 9.009112 3.314089 1.506758 1.544039
43 9.029894 3.267795 1.483968 1.518783
44 9.055359 3.240397 1.517348 1.517085
45 9.040278 3.235410 1.590436 1.509334
46 8.993796 3.252374 1.651106 1.431041
47 8.967464 3.236265 1.672098 1.338936
48 8.952859 3.235916 1.662450 1.287055
49 9.121430 3.217599 1.711292 1.263948
50 9.147871 3.205194 1.676641 1.211038
Will anyone please help why this might happen?
Perhaps I am estimating the models incorrectly?
Thank you
--
View this message in context:
http://r.789695.n4.nabble.com/cointegration-and-VECM-urca-package-and-Eviews-tp4709708.html
Sent from the R help mailing list archive at Nabble.com.