Table 3 Vector error-correction model... | Emerald Publishing

Table 3

Vector error-correction model specifications and test statistics

Statistics	Model 1		Model 2
ECT coefficient $(ϕ)$	−0.56^***		−0.73^***
Coefficients for cointegrated variables in ECT $(γ_{i})$	$LOG {({CSP}_{Crude})}_{t - 1}$	−0.76^***	$LOG {({PN}_{SOMS})}_{t - 1}$	2.34^***
	$LOG {({PC}_{SOMS})}_{t - 1}$	2.01^***	$LOG {({PIP}_{NG})}_{t - 1}$	−5.04^***
	$LOG {({PC}_{ALT})}_{t - 1}$	−2.02^***	$LOG {({PRD}_{NG})}_{t - 1}$	10.40^***
	$LOG {({PC}_{ALT})}_{t - 1}$	−2.02^***	$LOG {({FP}_{SG})}_{t - 1}$	−2.54^***
Johansen trace statistic	60.39^***		121.88^***
Johansen maximum eigenvalue statistic	29.43^**		63.54^***
R²	0.54		0.51
Akaike information criterion	−1.91		1.33
Optimal lag length	3		5
Breusch–Godfrey LM statistics	9.62^***		24.82^***
Jarque–Bera statistics	2.64^***		0.92^***
Inverse characteristic polynomial roots = 1	3		4
Wald test statistics	22.16^***		28.65^*

Statistics	Model 1		Model 2
ECT coefficient $(ϕ)$	−0.56^***		−0.73^***
Coefficients for cointegrated variables in ECT $(γ_{i})$	$LOG {({CSP}_{Crude})}_{t - 1}$	−0.76^***	$LOG {({PN}_{SOMS})}_{t - 1}$	2.34^***
	$LOG {({PC}_{SOMS})}_{t - 1}$	2.01^***	$LOG {({PIP}_{NG})}_{t - 1}$	−5.04^***
	$LOG {({PC}_{ALT})}_{t - 1}$	−2.02^***	$LOG {({PRD}_{NG})}_{t - 1}$	10.40^***
	$LOG {({PC}_{ALT})}_{t - 1}$	−2.02^***	$LOG {({FP}_{SG})}_{t - 1}$	−2.54^***
Johansen trace statistic	60.39^***		121.88^***
Johansen maximum eigenvalue statistic	29.43^**		63.54^***
R²	0.54		0.51
Akaike information criterion	−1.91		1.33
Optimal lag length	3		5
Breusch–Godfrey LM statistics	9.62^***		24.82^***
Jarque–Bera statistics	2.64^***		0.92^***
Inverse characteristic polynomial roots = 1	3		4
Wald test statistics	22.16^***		28.65^*

Note(s): ^***Denoting 0.01 significance level; ^**denoting 0.05 significance level; ^*denoting 0.1 significance level