Johansen Procedure --

[souliaris]

Hi: I am not sure whether this is a bug/problem, but I have not been able to replicate known results for the Johansen procedure using Eviews. EViews files with the data are attached with this post (see the VAR object called "var_issue" in the page "UKM1EM" for the VAR run). Upfront apologies if I am comparing apples with oranges.

GAUSS, OxMetrixs and CATS/RATS give the same answer for the attached data.

Here is the output from CATS, which matches output from GAUSS, OxMetrics[PcGIVE]:

Input commands are given below...

OPEN DATA "C:\Temp\kkk.txt"
CALENDAR(Q) 1963
ALL 1989:02
DATA(FORMAT=FREE,ORG=COLUMNS) 1963:01 1989:02 MP DP II RNET

SOURCE 'C:\CATS2\CATS.src'
@CATS(LAGS=2,DETTREND=DRIFT) 1964:02 1989:02
# MP DP II RNET

Here is the output from CATS...

MODEL SUMMARY

Sample: 1963:04 to 1989:02 (103 observations)
Effective Sample: 1964:02 to 1989:02 (101 observations)
Obs. - No. of variables: 92
System variables: MP DP II RNET
Constant/Trend: Unrestricted Constant
Lags in VAR: 2

I(2) analysis not available for the specI(1)-ANALYSIS
p-r r Eig.Value Trace Trace* Frac95 P-Value P-Value*
4 0 0.543 97.330 90.826 47.707 0.000 0.000
3 1 0.115 18.272 15.517 29.804 0.556 0.750
2 2 0.057 5.932 4.887 15.408 0.706 0.818
1 3 0.000 0.019 0.015 3.841 0.890 0.904


ified model.

The unrestricted estimates:
BETA(transposed)
MP DP II RNET
Beta(1) -7.145 -48.190 7.668 -52.749
Beta(2) -5.060 93.828 0.731 -44.388
Beta(3) 3.300 1.843 -4.803 -6.306
Beta(4) -7.027 -39.842 -0.342 -10.958

ALPHA
Alpha(1) Alpha(2) Alpha(3) Alpha(4)
DMP 0.014 0.001 -0.002 0.000
(8.739) (0.444) (-1.149) (0.050)
DDP -0.001 -0.002 0.001 -0.000
(-1.062) (-2.574) (1.089) (-0.075)
DII 0.002 0.000 0.003 0.000
(1.818) (0.315) (1.987) (0.078)
DRNET -0.001 0.002 0.002 -0.000
(-0.544) (1.254) (1.461) (-0.101)

PI
DP II RNET
DMP -0.112 -0.626 0.119 -0.774
(-5.882) (-3.402) (8.018) (-6.762)
DDP 0.019 -0.139 -0.011 0.123
(2.113) (-1.644) (-1.676) (2.343)
DII -0.011 -0.073 0.006 -0.156
(-0.733) (-0.510) (0.507) (-1.758)
DRNET 0.004 0.194 -0.013 -0.045
(0.263) (1.335) (-1.127) (-0.500)

Log-Likelihood = 1826.499

Here is the output from Eviews (Linear deterministic trend in the data, Option 3): Notice the trace and max tests do not match.

Date: 11/12/09 Time: 10:56
Sample: 1964Q2 1989Q2
Included observations: 101
Trend assumption: Linear deterministic trend
Series: MP D(P) I RNET
Lags interval (in first differences): 1 to 2


Hypothesized Trace 5 Percent 1 Percent
No. of CE(s) Eigenvalue Statistic Critical Value Critical Value

None ** 0.374505 62.29970 47.21 54.46
At most 1 0.085014 14.90935 29.68 35.65
At most 2 0.056713 5.935843 15.41 20.04
At most 3 0.000386 0.038988 3.76 6.65

Trace test indicates 1 cointegrating equation(s) at both 5% and 1% levels
*(**) denotes rejection of the hypothesis at the 5%(1%) level

Here is the data file in ASCII form:

mp dp ii rnet (1963:01 to 1989:02, quarterly)

10.90464472 . 11.03814243 0.043133334
10.92184264 0.004742403000000062 11.07714547 0.043966667
10.91887712 0.008016401999999978 11.08023303 0.042066666
10.93555549 0.01380099499999998 11.10122093 0.043533333
10.94450312 0.0007787200000000105 11.11693037 0.048966667
10.93686466 0.01356664899999993 11.13065429 0.05
10.94062184 0.01165606600000002 11.12984788 0.0507
10.92394731 0.00838362199999998 11.15284473 0.067933334
10.92391285 0.01426331000000003 11.14001951 0.0752
10.92656882 0.008316973999999978 11.14682048 0.0677
10.91221629 0.01128687900000003 11.15663616 0.064366666
10.93054693 0.006065643999999981 11.16788261 0.062533334
10.93271154 0.01013566600000004 11.17011116 0.062433333
10.90963515 0.01311834099999998 11.17177229 0.063366667
10.91481566 0.01042365099999998 11.17791416 0.0748
10.88758117 0.00850613400000011 11.17720095 0.073033334
10.91370419 -0.001398604000000026 11.20305232 0.063633334
10.91270187 0.01288765199999986 11.20362472 0.056466667
10.94469819 0.002470684000000167 11.21263085 0.055833333
10.93981707 0.007427964999999981 11.21240128 0.0727
10.92726404 0.01750682100000001 11.25169056 0.080100001
10.92322692 0.01922789899999988 11.23599868 0.082466666
10.91621791 0.01233666500000008 11.2608679 0.07679999999999999
10.91667765 0.01349370799999994 11.26931131 0.07440000099999999
10.88705627 0.01164521799999996 11.25867817 0.084933332
10.85773824 0.007989636000000022 11.26610379 0.0925
10.85412708 0.01121208600000001 11.2747571 0.096066666
10.87153682 0.01510864300000003 11.28519642 0.089699999
10.85396157 0.02337573200000009 11.27115959 0.091066666
10.86963053 0.01587006199999985 11.29891508 0.080466668
10.85764529 0.02102295300000012 11.30367249 0.075100001
10.86147133 0.0185801699999999 11.32210389 0.072
10.9046798 0.02267323599999993 11.30257474 0.075100001
10.88901241 0.02144687600000017 11.32800453 0.0652
10.90254022 0.01919352399999985 11.34278936 0.058033333
10.90013285 0.01620605200000002 11.34839274 0.0479
10.92791611 0.01307961000000013 11.34663473 0.0496
10.9575232 0.0158057359999999 11.36232348 0.056866666
10.96230033 0.0215457160000001 11.35713678 0.077533334
10.9623727 0.02597972399999993 11.40120029 0.080866667
10.94301665 0.02500077799999989 11.44773607 0.100866667
10.99552117 0.007192564000000124 11.44525672 0.089200001
10.94067017 0.03596303999999994 11.45257018 0.127299999
10.89339618 0.05266318000000014 11.44517111 0.1475
10.83846213 0.04226195599999993 11.42994595 0.155633332
10.80126792 0.06548388499999991 11.44338244 0.133166666
10.77980199 0.04381615100000014 11.45279322 0.127733332
10.79030232 0.05272986099999999 11.4400429 0.126033335
10.74909539 0.05976926799999993 11.42875006 0.113766667
10.73213277 0.05898000600000009 11.41645862 0.097700001
10.731809 0.04435236499999995 11.41671188 0.106033335
10.71533714 0.03780933499999994 11.42908723 0.114699999
10.73285945 0.03119784000000003 11.44630478 0.093033333
10.72148144 0.03974155099999999 11.45286756 0.109366668
10.71184397 0.03209804299999996 11.46461718 0.117199999
10.67739148 0.04188227200000006 11.48082574 0.151300001
10.67049774 0.03445651699999996 11.46395578 0.11416667
10.67824784 0.03494864900000006 11.46809545 0.079000001
10.72860485 0.01926233699999991 11.47088398 0.067400001
10.76946059 0.01479099900000003 11.48822285 0.061466667
10.80502994 0.02186104799999999 11.50384436 0.065966668
10.80612248 0.02334708900000004 11.51093348 0.090966667
10.81803772 0.02299246099999997 11.51481368 0.096699998
10.81954801 0.03085274000000004 11.51424459 0.118266665
10.82508508 0.023783742 11.51036218 0.129199998
10.81361717 0.03134434299999999 11.56920182 0.126900002
10.78960989 0.05109527499999999 11.55258839 0.142000001
10.75581412 0.04395490400000002 11.55743027 0.162433329
10.71222044 0.042582391 11.55025978 0.182133331
10.69500476 0.04182149999999996 11.5274002 0.172533334
10.66016315 0.03033801400000002 11.51986136 0.158800001
10.6561234 0.02869566899999998 11.50225878 0.1536
10.66683307 0.02100982800000001 11.49602343 0.131066666
10.68675329 0.02064835200000001 11.50170272 0.124733334
10.68247924 0.02156572300000001 11.52473545 0.148233331
10.66678659 0.02760928399999998 11.52024859 0.157233334
10.674279 0.013695716 11.52460697 0.140433334
10.67392252 0.01239084900000001 11.53483372 0.1339
10.68797456 0.01594778499999999 11.52830657 0.111700001
10.71528508 0.01894319700000002 11.53530321 0.100466668
10.74067583 0.01667513199999998 11.56030511 0.112733332
10.75889962 0.002879893000000008 11.56339992 0.100233332
10.75703338 0.015425128 11.57940585 0.09836666400000001
10.77490783 0.014761183 11.58593426 0.092900003
10.82046806 0.009351010000000006 11.59621827 0.0921
10.83944783 0.015805792 11.5986168 0.093766667
10.85754308 0.0160646 11.60698031 0.110477132
10.85840129 0.018858355 11.62328546 0.096081286
10.89559172 0.020530432 11.6371146 0.114226311
10.92270978 0.004175763000000001 11.64120434 0.084367967
10.96644621 0.000888888 11.63856203 0.054031542
11.00016163 0.005770921 11.64572409 0.04219608100000001
11.05850197 0.005521400999999999 11.66176779 0.036159186
11.11389332 -0.001868597999999999 11.67359769 0.026583946
11.17006476 0.006252728999999999 11.69037685 0.030444625
11.18244512 0.018216019 11.69980329 0.038833809
11.23683345 0.010203154 11.70220608 0.02918726
11.28624353 0.009610197000000001 11.72436884 0.027721942
11.31111869 0.012609758 11.75053981 0.036885214
11.34251905 0.011674885 11.76055087 0.026500807
11.38719386 0.005579284000000004 11.76859042 0.029816798
11.41776207 0.010416485 11.78487442 0.029199705
11.42196461 0.022333175 11.80332384 0.049366348
11.42381703 0.01486369700000001 11.81118391 0.050716672
11.46046672 0.01057782599999999 11.82651648 0.046366668
11.47954691 0.01915726000000001 11.82610714 0.052600003

[souliaris]

A gentle reminder: Any news on this (non?) issue? Thanks in advance... Sam

[EViews Chris]

The difference that you're seeing is due to a slightly different choice of how to describe the lags in the model.

From what I can tell from your output, it appears that when you say LAGS=2 in CATS, you're telling CATS that you want to use Y, Y(-1) and Y(-2) in levels. This comes from thinking of the cointegration model as a restricted form of the general levels VAR.

In EViews, when we ask you to specify lags, we ask for the pairs of lags to include in the differenced (VECM) model.

At the end of the day, this means that the CATS option LAGS=2 is equivalent to the EViews lag intervals "1 1". The CATS specification says that there are two lags in the levels VAR, the EViews specification says that there is one lagged difference in the differenced VECM. Similarly, the equivalent of the EViews setting lag intervals "1 2" is actually LAGS=3 in CATS.

This may seem a bit confusing, but it's really just two different ways of saying the same thing.

[souliaris]

Hi Chris: I was starting from a VAR (2) object in Eviews -- and didn't notice the switch to the VECM framework when testing for cointegration. Thank you for the explanation.

Regards, Sam