EViews.com

I am attaching results of the johanson cointegration test that I obtain. I want to know on the long run elasticity coefficients how do I calculate their significance ? All the test provides is the standard errors (eviews 4). Also in order to run a Vector Error Correction Model (besides cointegration) do I have to have integration of order one I(1). If I have integration order of 2 I cant?
Date: 02/20/12 Time: 17:54
Sample(adjusted): 1976:1 2009:1
Included observations: 133 after adjusting endpoints
Trend assumption: Linear deterministic trend
Series: LOG(VEX) LOG(GDP) LOG(P) V2
Lags interval (in first differences): 1 to 2

Unrestricted Cointegration Rank Test

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

None ** 0.319493 83.45672 47.21 54.46
At most 1 * 0.155039 32.26273 29.68 35.65
At most 2 0.071327 9.856945 15.41 20.04
At most 3 0.000114 0.015108 3.76 6.65

*(**) denotes rejection of the hypothesis at the 5%(1%) level
Trace test indicates 2 cointegrating equation(s) at the 5% level
Trace test indicates 1 cointegrating equation(s) at the 1% level


Hypothesized Max-Eigen 5 Percent 1 Percent
No. of CE(s) Eigenvalue Statistic Critical Value Critical Value

None ** 0.319493 51.19398 27.07 32.24
At most 1 * 0.155039 22.40579 20.97 25.52
At most 2 0.071327 9.841837 14.07 18.63
At most 3 0.000114 0.015108 3.76 6.65

*(**) denotes rejection of the hypothesis at the 5%(1%) level
Max-eigenvalue test indicates 2 cointegrating equation(s) at the 5% level
Max-eigenvalue test indicates 1 cointegrating equation(s) at the 1% level

Unrestricted Cointegrating Coefficients (normalized by b'*S11*b=I):

LOG(VEX) LOG(GDP) LOG(P) V2
-2.255929 7.876386 0.719964 251.7133
-12.99478 26.01501 -2.328684 -86.77336
2.723118 -16.22617 -9.629216 50.57427
1.417454 -0.304146 -1.318929 13.98320


Unrestricted Adjustment Coefficients (alpha):

D(LOG(VEX)) 0.017372 0.017435 -0.003040 2.12E-05
D(LOG(GDP)) -0.002693 -0.000105 0.001982 -0.000193
D(LOG(P)) -0.000545 0.003535 0.009458 0.000169
D(V2) -0.004152 0.000919 -0.000149 7.93E-06


1 Cointegrating Equation(s): Log likelihood 1260.183

Normalized cointegrating coefficients (std.err. in parentheses)
LOG(VEX) LOG(GDP) LOG(P) V2
1.000000 -3.491416 -0.319143 -111.5786
(0.65929) (0.58449) (15.5252)

Adjustment coefficients (std.err. in parentheses)
D(LOG(VEX)) -0.039191
(0.01064)
D(LOG(GDP)) 0.006076
(0.00406)
D(LOG(P)) 0.001229
(0.00810)
D(V2) 0.009367
(0.00133)


2 Cointegrating Equation(s): Log likelihood 1271.386

Normalized cointegrating coefficients (std.err. in parentheses)
LOG(VEX) LOG(GDP) LOG(P) V2
1.000000 0.000000 0.849019 165.6240
(0.36409) (22.9049)
0.000000 1.000000 0.334581 79.39545
(0.16988) (10.6869)

Adjustment coefficients (std.err. in parentheses)
D(LOG(VEX)) -0.265751 0.590396
(0.05866) (0.12089)
D(LOG(GDP)) 0.007439 -0.023942
(0.02373) (0.04891)
D(LOG(P)) -0.044704 0.087665
(0.04718) (0.09722)
D(V2) -0.002570 -0.008807
(0.00772) (0.01592)


3 Cointegrating Equation(s): Log likelihood 1276.307

Normalized cointegrating coefficients (std.err. in parentheses)
LOG(VEX) LOG(GDP) LOG(P) V2
1.000000 0.000000 0.000000 281.3752
(34.6086)
0.000000 1.000000 0.000000 125.0106
(15.1631)
0.000000 0.000000 1.000000 -136.3352
(19.3336)

Adjustment coefficients (std.err. in parentheses)
D(LOG(VEX)) -0.274029 0.639724 0.001181
(0.05978) (0.14052) (0.04409)
D(LOG(GDP)) 0.012837 -0.056109 -0.020784
(0.02411) (0.05668) (0.01778)
D(LOG(P)) -0.018948 -0.065807 -0.099700
(0.04678) (0.10996) (0.03450)
D(V2) -0.002974 -0.006397 -0.003698
(0.00788) (0.01853) (0.00582)

"do I have to have integration of order one I(1). If I have integration order of 2 I cant?"

if some of your variables are I(2) and some I(1) then for integration to be possible the I(2) variables should be at least 2.
i.e. if you have only 1 variable I(2) and the rest of the variables are I(1) then they cannot be co-integrated.


"I am attaching results of the johanson cointegration test that I obtain"

why you state so many results? you should have chose ONLY ONE model and use only that for the co-integration tests.
now you have many and conflicting results. how should anyone help you with that? are you familiar with model specification procedures?


"I want to know on the long run elasticity coefficients"

that implies ONLY ONE co-integration vectors. HOWEVER, if you 2 (as you state) then you should use BOTH in your analysis and try to explain that (if they are close in value that is good).


"how do I calculate their significance ?"

E-Views VECM estimation output automatically presents significance tests for the co-integrating vector coefficients.