Table IV. ARDL cointegration... | Emerald Publishing

Table IV.

ARDL cointegration test-statistics and exhibit of the unrestricted error correction model

Region	Model	$X_{S C}^{2}$ (1)	$X_{S C}^{2}$ (3)	F-stat	Status at 5(%) level
(a) Statistics for ARDL bounds cointegration test performed separately for each region
GR	ARDL(2,2,1,1)	1.03 (0.31)	1.05 (0.79)	4.92**	Cointegration
FR	ARDL(3,3,1,1)	2.48 (0.12)	5.93 (0.12)	4.57**	Cointegration
DR	ARDL(7,6,1,1)	0.58 (0.45)	1.54 (0.67)	2.47	No cointegration
OV	ARDL(2,2,1,1)	0.84 (0.36)	5.30 (0.15)	4.95**	Cointegration
FL	ARDL(9,9,1,1)	1.16 (0.28)	1.68 (0.64)	1.94	No cointegration
GE	ARDL(3,3,1,1)	2.63 (0.11)	3.92 (0.27)	3.16	No cointegration
UT	ARDL(1,1,1,1)	2.40 (0.12)	3.97 (0.26)	3.84*	Inconclusive
ZH	ARDL(2,1,1,1)	0.05 (0.83)	0.40 (0.94)	6.71***	Cointegration
ZL	ARDL(10,9,1,1)	2.54 (0.11)	3.19 (0.36)	2.12	No cointegration
NB	ARDL(4,4,1,1)	0.54 (0.46)	2.78 (0.43)	1.23	No cointegration
LI	ARDL(2,2,1,1)	0.36 (0.55)	3.49 (0.32)	4.04**	Cointegration
Bound critical values
1%		5%		10%
I(0)	I(1)	I(0)	I(1)	I(0)	I(1)
4.29	5.61	3.23	4.35	2.72	3.77
Independent variable	Estimate	Std. Error	t-value	P-value
(b) Unrestricted error correction model estimate for GR (⁠ $y_{t}^{i}$ ⁠, i = ZH)
Const.	0.638	0.66	0.97	0.34
$Δ y_{t - 1}^{i}$	−0.205	0.10	−2.04	0.04**
$Δ y_{t - 2}^{i}$	−0.444	0.10	−4.49	0.00***
Δx_t₋₁	−0.178	0.07	−2.47	0.02**
Δgdp_t₋₁	2.250	0.47	4.77	0.00***
Δr_t₋₁	0.007	0.00	1.40	0.17
$y_{t - 1}^{i}$	−0.156	0.05	−3.35	0.00***
x_t_-1	0.135	0.03	4.10	0.00***
gdp_t₋₁	−0.031	0.07	−0.43	0.66
r_t₋₁	0.005	0.00	1.69	0.10*

Region	Model	$X_{S C}^{2}$ (1)	$X_{S C}^{2}$ (3)	F-stat	Status at 5(%) level
(a) Statistics for ARDL bounds cointegration test performed separately for each region
GR	ARDL(2,2,1,1)	1.03 (0.31)	1.05 (0.79)	4.92**	Cointegration
FR	ARDL(3,3,1,1)	2.48 (0.12)	5.93 (0.12)	4.57**	Cointegration
DR	ARDL(7,6,1,1)	0.58 (0.45)	1.54 (0.67)	2.47	No cointegration
OV	ARDL(2,2,1,1)	0.84 (0.36)	5.30 (0.15)	4.95**	Cointegration
FL	ARDL(9,9,1,1)	1.16 (0.28)	1.68 (0.64)	1.94	No cointegration
GE	ARDL(3,3,1,1)	2.63 (0.11)	3.92 (0.27)	3.16	No cointegration
UT	ARDL(1,1,1,1)	2.40 (0.12)	3.97 (0.26)	3.84*	Inconclusive
ZH	ARDL(2,1,1,1)	0.05 (0.83)	0.40 (0.94)	6.71***	Cointegration
ZL	ARDL(10,9,1,1)	2.54 (0.11)	3.19 (0.36)	2.12	No cointegration
NB	ARDL(4,4,1,1)	0.54 (0.46)	2.78 (0.43)	1.23	No cointegration
LI	ARDL(2,2,1,1)	0.36 (0.55)	3.49 (0.32)	4.04**	Cointegration
Bound critical values
1%		5%		10%
I(0)	I(1)	I(0)	I(1)	I(0)	I(1)
4.29	5.61	3.23	4.35	2.72	3.77
Independent variable	Estimate	Std. Error	t-value	P-value
(b) Unrestricted error correction model estimate for GR ( $y_{t}^{i}$ , i = ZH)
Const.	0.638	0.66	0.97	0.34
$Δ y_{t - 1}^{i}$	−0.205	0.10	−2.04	0.04**
$Δ y_{t - 2}^{i}$	−0.444	0.10	−4.49	0.00***
Δx_t₋₁	−0.178	0.07	−2.47	0.02**
Δgdp_t₋₁	2.250	0.47	4.77	0.00***
Δr_t₋₁	0.007	0.00	1.40	0.17
$y_{t - 1}^{i}$	−0.156	0.05	−3.35	0.00***
x_t_-1	0.135	0.03	4.10	0.00***
gdp_t₋₁	−0.031	0.07	−0.43	0.66
r_t₋₁	0.005	0.00	1.69	0.10*

Notes:

In (a), the UEC model is estimated with a constant for all regions. The lag order is selected with AIC and further adjustment when necessarily to correct for serial correlation and dynamic stability of autoregressive structure of the UEC model. χ²_SC(m) is the m-order LM residual serial correlation test of the estimated ARDL model. The critical values are taken from Table CI(iii) and and CII(iii) of Pesaran et al. (2001), with k = 3; For the regression estimates in (b), the residual standard error = 0.02, multiple R-squared = 0.46 and the adjusted R-squared = 0.39. Statistical significance is denoted by;

∗, ∗∗ and ∗∗∗ at the 10, 5 and 1% levels, respectively

[ViewLarge]

Sharing Unavailable