The bilateral J-curve between Ghana and her key trading partners

Figures & data

Figure 1. Trade balance between Ghana and key trading partners in the period of 2013 to 2018.Source: Author, computed based on the data from (Wits.worldbank.org, 2021).

Figure 2. The J curve after depreciation/devaluation source: Levi (2005).

Figure 3. Plots of cumulative sum of recursive residuals (CUSUM) and cumulative sum of squares of recursive residuals (CUSUMSQ).Panel A: China Panel B: Swiss

Figure 3. (Continued.)

Table 1. Test for unit root

Variables	ADF Test		I(n)
	Level	1st-diff
lnTBGH-CN	−0.450		I(0)
lnTBGH-SW	−2.311*		I(0)
lnRERGH-CN	−1.075	−1.688*	I(1)
lnRERGH-SW	−3.154*		I(0)
lnYGH	−1.859		I(0)
LnYCN	−2.848*		I(0)
LnYSW	−2.594		I(0)

Note: The statistical significance is denoted as *, **, and *** for 10%, 5%, and 1% respectively. Additionally, ADF test is Augmented Dickey–Fuller test, Level defines level form, and the first difference form is defined as 1st-diff. GH is Ghana, SW is Switzerland and CN is China.

Table 2. Bounds test

Trading partner	Dependent variables ΔlnTBt	F-statistics	Critical Value Bonds of the F-statistics						Outcome
			95%		97.5%		99%
			I(0)	I(1)	I(0)	I(1)	I(0)	I(1)
China	ARDL (4, 5, 3, 4)	8.039	2.79	3.67	3.15	4.08	3.65	4.66	Co-integration
Swiss	ARDL (1, 1, 3, 3)	9.024							Co-integration

Table 3. ARDL model results

Panel A. China
Variable	Coefficient	Std. Error	t-Statistic	Prob.
Dependent variable: ΔlnTBt
Long-run elasticities
C	−40.282***	7.972	−5.052	0.000
lnRER(−1)	3.235***	0.647	4.999	0.000
lnYGH(−1)	−0.037	0.093	−0.402	0.689
lnYCN(−1)	1.045***	0.267	3.903	0.000
Short-run elasticities
C	0.172***	0.063	2.710	0.010
ΔlnTB(−1)	−0.364***	0.145	−2.497	0.017
ΔlnTB(−2)	0.195*	0.134	1.450	0.155
ΔlnTB(−3)	0.311**	0.166	1.867	0.070
ΔlnTB(−4)	−0.357**	0.160	−2.229	0.032
ΔlnRER	−1.338	1.090	−1.227	0.227
ΔlnRER(−1)	−0.306	1.306	−0.234	0.815
ΔlnRER(−2)	−1.052	1.353	−0.777	0.442
ΔlnRER(−3)	−0.360	1.301	0.276	0.783
ΔlnRER(−4)	0.469	1.348	0.348	0.729
ΔlnRER(−5)	2.762**	1.212	2.277	0.028
ΔlnYGH	0.348***	0.114	3.054	0.004
ΔlnYGH(−1)	0.391***	0.115	3.406	0.001
ΔlnYGH(−2)	0.332***	0.108	3.054	0.004
ΔlnYGH(−3)	0.351***	0.102	3.423	0.001
ΔlnYCN	1.460	1.335	1.092	0.281
ΔlnYCN(−1)	−2.103***	0.785	−2.677	0.011
ΔlnYCN(−2)	−2.573***	0.785	−3.275	0.002
ΔlnYCN(−3)	−2.137***	0.767	−2.783	0.008
ΔlnYCN(−4)	−3.623**	1.770	−2.046	0.048
ECM(−1)	−0.142**	0.079	−1.797	0.080
Panel B. Swiss
Variable	Coefficient	Std. Error	t-Statistic	Prob.
Long-run elasticities
C	−26.574***	7.031	−3.779	0.000
lnRER(−1)	1.594***	0.285	5.589	0.000
lnYGH(−1)	0.014	0.055	0.271	0.787
lnYSW(−1)	1.498***	0.592	2.529	0.014
Dependent variable: ΔlnTBt
Short-run elasticities
C	−0.031*	0.023	−1.363	0.179
ΔlnTB(−1)	−0.065	0.151	−0.434	0.665
ΔlnRER	0.689	0.854	0.806	0.424
ΔlnRER(−1)	−0.674	0.872	−0.773	0.443
ΔlnYGH	0.042	0.086	0.488	0.627
ΔlnYGH(−1)	0.104	0.084	1.234	0.223
ΔlnYGH(−2)	0.025	0.081	0.315	0.753
ΔlnYGH(−3)	0.076	0.080	0.950	0.346
ΔlnYSW	−0.637	0.975	−0.652	0.517
ΔlnYSW(−1)	−3.031	2.442	−1.240	0.221
ΔlnYSW(−2)	5.324**	2.483	2.144	0.037
ΔlnYSW(−3)	3.592*	2.523	1.423	0.161
ECM(−1)	−0.194**	0.086	−2.232	0.030

Note: ***, **, * represents 1%, 5% & 10% significance level, respectively

Table 4. Diagnostic tests results

Test	China	Swiss
Serial Correlation LM Test	0.6420	0.9160
Ramsey RESET	0.9621	0.3448
CUSUM	ARDL model is stable	ARDL model is stable
CUSUMSQ	ARDL model is stable	ARDL model is stable