Skip to main content
. 2010 Dec 31;36(2):1134–1142. doi: 10.1016/j.energy.2010.11.036

Table 5.

Results of Vector Error Correction (VEC) Model Selected.

Explanatory variables
Dependent variable
ΔLGDPSA
ΔLELECSA
Co-integrating Eq CointEq1 CointEq2
LGDPSA(−1) 1 1
LELECSA(−1) −1.2085 (0.1431) [−8.4466] −1.2085 (0.1431) [−8.4466]
C in co-integrating Eq 2.6503 2.6503



CointEq −0.0524 (0.1012) [−0.5182] 0.1339 (0.0468) [ 2.8610]**
ΔLGDPSA(−1) −0.1483 (0.1956) [−0.7581] −0.1255 (0.0905) [−1.38739]
ΔLELECSA(−1) 0.3607 (0.3395) [1.0624] −0.1084 (0.1570) [−0.6901]
C 0.0239 (0.0131) [1.8332] 0.0265 (0.0060) [4.3802]



Adjusted R2 −0.0321 0.1329
F-statistic 0.61701 2.8898
Log likelihood 51.7299 81.0313
Akaike AIC −2.5121 −4.0543
Schwarz SC −2.3397 −3.8819



Wald test
Null hypotheses ΔLELECSA does not Granger cause ΔLGDPSA ΔLGDPSA does not Granger cause ΔLELECSA
Chi-square χ2 statistics 1.1288 1.9249
Degree of freedom 1 1
Probability 0.2880 0.1653

Notes: 1. The values in this table are the corresponding coefficient and numbers in parentheses () and brackets [] are the corresponding standard errors and t-values of the parameters, respectively. The parameter symbols of (−1) and (−2) denotes 1 lag value and 2 lag value of their own parameters, respectively.

2. The equivalent equations of ΔLGDPSA and ΔLELECSA are as follows:

ΔLGDPSA = −0.0524*(LGDPSA(−1) − 1.2085*LELECSA(−1) + 2.6502) − 0.1483*ΔLGDPSA(−1) + 0.3607*ΔLELECSA(−1) + 0.0239 ΔLELECSA = 0.1390*(LGDPSA(−1) − 1.2085*LELECSA(−1) + 2.6503) − 0.1255*ΔLGDPSA(−1) − 0.1083*ΔLELECSA(−1) + 0.0264. 3. In Wald test, * denotes the rejection of the null hypothesis at the 5 percent significance value.

4. In VECM, ** denotes the rejection of null hypothesis at the 5 percent significance value in co-integration equations. In VECM equations, the probability of the error correction term in ΔLGDPSA and ΔLELECSA equations by t-test is 0.6077 and 0.0072 respectively. It indicates that there is a long-term causality from LGDPSA to LELECSA (i.e. LGDPSA → LELECSA).