Market integration and price transmission analysis of foreign and Ghanaian beef and mutton markets; the effect of crude oil price and exchange rate fluctuations

Figures & data

Figure 1. Trend of meat importation in Ghana (source: ceicdata.com/MoFA).

Figure 2. Trend plot of monthly prices of beef in Ghana, Argentina, and South Africa.

Figure 3. Trend plot of monthly prices of mutton in Ghana, Belgium, and New Zealand.

Figure 4. Trend of crude oil prices and exchange rate in Ghana.

Table 1. Descriptive statistics of data series

Variables	BEGHA	BEARG	BESA	MUGHA	MUBEL	MUNZ	CRUDOIL	EXG
Mean	5.510	3.665	5.689	5.494	5.138	4.604	55.523	4.751
Median	5.978	3.490	5.790	5.499	5.059	4.809	55.520	4.566
Maximum	6.487	4.610	7.230	6.523	6.041	5.810	80.630	5.747
Minimum	3.837	2.650	4.150	4.597	4.499	3.110	26.630	3.810
Std. Dev.	0.827	0.583	0.720	0.468	0.398	0.712	13.020	0.648
Skewness	−0.517	−0.026	0.049	0.039	0.040	−0.542	−0.108	0.214
Kurtosis	−1.411	−1.371	−0.26	−0.669	−0.878	−0.702	2.287	1.709
Jarque-Bera	7.404	4.620	0.304	1.288	3.018	4.317	1.387	4.624
Probability	0.025	0.099	0.859	0.525	0.221	0.116	0.500	0.099

Source: Authors’ Computation (2021)

Table 2. ADF, PP, and KPSS unit root test results

Variables	ADF		PP		KPSS
	Level	Diff	Level	Diff	Level	Diff
BEGHA	−3.8433	−4.03***	−21.642	−46.5799***	0.0548**	0.0422
BEARG	−1.2193	−5.8111***	−4.6524	−52.3167***	0.217**	0.2281
BESA	−2.3327	−3.814***	−8.6143	−43.0233***	0.2783**	0.109
MUGHA	−3.4817	−3.8645***	−20.085	−49.939***	0.0691**	0.0609
MUBEL	−4.1923	−5.4907***	−21.477	−44.6964***	0.1138**	0.0347
MUNZ	−3.2736	−3.4864***	−11.353	−32.7272***	0.232**	0.059
EXCHA	−3.1173	−5.4424***	−17.872	−58.7922***	0.1258**	0.0661
Crude Oil	−2.185	−5.2391***	−1.1771	−51.8281***	0.2886**	0.4397

Note: The appropriate optimum lag length was determined using the AIC criteria. The critical value for ADF and PP at 1% and 5% are −2.6 and −4.059 for constant and trend specification, for KPSS is 0.146 for 99% confidence interval for constant and trend specifications. *** and ** indicate rejection of the null hypothesis at 99% and 95% confidence interval, respectively.

BEGHA, BEARG, and BESA represents prices of beef in Ghana, Argentina, and South Africa, respectively; MUGHA, MUBEL, and MUNZ represent the prices of mutton in Ghana, Belgium, and New Zealand, respectively; EXCHA = Exchange rate in Ghana, CrudOil = Crude Oil Prices in Ghana

Source: Authors’ Computation (2021)

Table 3. Zivot–Andrews unit root test under single break point

Variables	Constant t-Min(Break Point)	Trend t-Min(Break Point)	Both t-Min(Break Point)
BEGHA	−5.39355** (2019:12)	−5.26872**(2018:10)	−5.33918* (2018:08)
BEARG	−2.97819 (2017:12)	−2.81568 (2019:08)	−3.26255 (2018:04)
BESA	−3.70238 (2017:04)	−4.07171(2017:12)	−4.06368 (2017:08)
MUGHA	−5.27211* (2017:03)	−5.20178** (2018:03)	−5.33664* (2018:08)
MUBEL	−5.04201* (2020:01)	−4.84301* (2019:08)	−5.23792* (2018:11)
MUNZ	−6.37721** (2019:12)	−6.26081** (2018:05)	−6.42009** (2017:08)
Crude Oil	−1.48510 (2020:02)	−4.15026 (2019:12)	−1.48510 (2020:02)
EXGHA	−3.13886 (2019:01)	−2.49785 (2020:04)	−3.02282 (2020:02)

Figures outside parenthesis are values of test statistic; break points are in brackets. Critical values are as follows: −4.80000 (95%), −5.34000 (99%) for intercept; −4.42000 (95%), −4.93000 (99%) for trend; −5.08000 (95%), −5.57000 (99%) for both intercept and trend. * and ** indicate rejection of the null hypothesis at 95% and 99% confidence interval, respectively.

Source: Authors’ Computation (2021)

Table 4. Johansen cointegration analysis

Markets	Meat product	H0	Trace Statistics	Maxi-Eigen Statistic	P (Normality Test (Shapiro–Wilk)	P (Homoskedasticity) (ARCH test))
Argentina – Ghana	Beef	r=0 r≤1	32.35*** 1.36	31.00*** 1.36	.7287	.529
South Africa – Ghana	Beef	r=0 r≤1	30.49** 4.62	25.87*** 4.62	.7564	.7572
Belgium – Ghana	Mutton	r=0 r≤1	38.82*** 4.51	33.44*** 10.19	.5174	.5828
New Zealand – Ghana	Mutton	r=0 r≤1	43.63*** 10.19	34.30*** 4.51	.3858	.8567
Crude oil price – Ghana	Beef	r=0 r≤1	33.04*** 6.10	26.94*** 6.10	.6026	.7926
Exchange Rate – Ghana	Beef	r=0 r≤1	32.20*** 5.55	26.64*** 5.55	.01419	.5745
Crude oil price – Ghana	Mutton	r=0 r≤1	40.40*** 5.24	36.85*** 6.55	.6141	.2413
Exchange Rate – Ghana	Mutton	r=0 r≤1	43.40*** 6.55	35.16*** 5.24	.03727	.2223

*** and ** indicate rejection of the null hypothesis at 99% and 95% confidence interval, respectively.

Source: Authors’ Computation (2021)

Table 5. Results of consistent TAR and M-TAR model specification ($\tau \ne 0$)

Meat	Beef				Mutton
Market Pairs	Argentina-Ghana		South Africa – Ghana		Belgium – Ghana		New Zealand – Ghana
Parameter	CTAR	CMTAR	CTAR	CMTAR	CTAR	CMTAR	CTAR	CMTAR
$\tau$	−0.096	0.037	0.111	0.026	−0.036	−0.038	−0.128	−0.046
$\rho 1$	−0.015 (0.112)	0.204 (0.142)	−0.051 (0.073)	−0.023 (0.090)	−0.457*** (0.139)	−0.282** (0.108)	−0.071 (0.062)	−0.16*** (0.052)
$\rho 2$	−0.314*** (0.117)	−0.293*** (0.092)	−0.159** (0.063)	−0.148** (0.057)	−0.419 *** (0.143)	−0.91*** (0.181)	−0.17** (0.072)	0.062 (0.099)
$\rho 1=\rho 2=0$	3.603** [0.034]	6.716*** [0.003]	3.366** [0.042]	3.453** [0.039]	8.904*** [0.000]	15.23*** [0.000]	3.42** [0.040]	4.996*** [0.010]
$\rho 1=\rho 2$	3.851* [0.055]	9.729*** [0.003]	1.262 [0.266]	1.420 [0.239]	0.039 [0.844]	9.600*** [0.003]	1.132 [0.292]	3.994** [0.015]
Lags	2	2	2	2	2	2	2	2
LB(4)	0.920	0.841	0.293	0.239	0.220	0.108	0.388	0.363
LB(8)	0.920	0.731	0.426	0.297	0.187	0.188	0.196	0.350
AIC	−149.733	−155.342	−187.148	−187.314	−220.887	−230.177	−175.97	−178.856
SBIC	−139.518	−145.126	−176.933	−177.099	−212.645	−221.935	−167.73	−170.614

Standard errors are in parentheses, brackets are p-values, and *, **, and *** indicate rejection of null hypothesis at 10%, 5%, and 1% probability level.

Source: Authors’ Computation (2021)

Table 6. Results of asymmetric vector error correction model (VCEM) of mutton markets

Meat	Beef		Mutton
Market Pairs	$\mathrm{\Delta }\mathit{A}\mathit{R}\mathit{G}$ $\mathrm{\Delta }\mathit{G}\mathit{H}\mathit{A}$		$\mathrm{\Delta }\mathit{B}\mathit{E}\mathit{L}$ $\mathrm{\Delta }\mathit{G}\mathit{H}\mathit{A}$		$\mathrm{\Delta }\mathit{N}\mathit{Z}\mathit{M}$ $\mathrm{\Delta }\mathit{G}\mathit{H}\mathit{A}$
$\mathit{E}\mathit{C}{\mathit{T}}_{\mathit{P}\mathit{O}\mathit{S}-1}$	0.0389 {0.254}	0.00560 {0.936}	0.0185 {0.428}	0.0796 {0.256}	0.04370 {0.3141}	0.08632 {0.2302}
$\mathit{E}\mathit{C}{\mathit{T}}_{\mathit{N}\mathit{E}\mathit{G}-1}$	−0.0898* {0.014}	−0.15642* {0.037}	−0.0336* {0.0706}	−0.1208** {0.0308}	0.01292 {0.8165}	−0.12773 {0.1694}
Constant	−0.02612 (0.1718)	0.96654*** (0.26110)	1.0649*** (0.21486)	0.292209 (0.83386)	0.541585 (0.1263)	0.182136 (0.2529)
$\mathrm{\Delta }\mathit{G}\mathit{H}\mathit{M}-1$	−0.0031 (0.0761)	0.329265** (0.11573)	−0.0681* (0.03838)	0.185609 (0.14895)	0.228** (0.0876)	0.013487 (0.1755)
$\mathrm{\Delta }\mathit{G}\mathit{H}\mathit{M}-2$	0.055411 (0.0799)	0.41857*** (0.12155)	−0.002894 (0.03944)	0.148899 (0.15308)	0.1572* (0.0855)	0.159577 (0.1712)
$\mathrm{\Delta }\mathit{R}\mathit{M}-1$	−0.0417 (0.1404)	−0.133518 (0.21344)	0.406*** (0.12073)	0.059856 (0.46854)	0.119018 (0.1318)	0.47824* (0.2619)
$\mathrm{\Delta }\mathit{R}\mathit{M}-2$	−0.1955 (0.1419)	−0.132341 (0.21569)	0.166218 (0.12329)	0.309758 (0.47849)	0.174402 (0.1339)	0.5256* (0.268)
Excha R.	0.071019 (0.1032)	−0.6553*** (0.15695)	−0.654*** (0.12694)	−0.155587 (0.49266)	−0.35*** (0.072)	−0.13218 (0.1443)
Crude Oil	−0.0225 (0.0145)	0.016352 (0.02200)	−0.011280 (0.00805)	−0.008392 (0.03122)	−0.00117 (0.0116)	0.009275 (0.0232)
R-squared	0.107059	0.419270	0.498935	0.077353	0.09622	0.243108
AIC	-2.77198	−1.934896	−4.06474	−1.352582	0.10809	−3.37147
Schwarz SC	−2.48524	−1.648152	−3.778	−1.065837	−0.3948	−3.08473
SSR	0.157629	0.364065	0.043271	0.651740	2.808249	0.086553
Loglik	87.00015	63.14453	123.8451	46.54857	4.919474	104.0869

Standard errors are in parentheses, brackets are p-values, and *, **, and *** indicate rejection of null hypothesis at 10%, 5%, and 1% probability level. RM = Reference market.

Source: Author’s Computation (2021)

Table 7. Results of threshold vector error correction model

Beef Market	Regime 1		Regime 2
Variable	Ghana	South Africa	Ghana	South Africa
Constant	0.0721 (0.2179)	0.0227 (0.4535)	−0.0347 (0.1708)	−0.0102 (0.4337)
$\mathit{E}\mathit{C}{\mathit{T}}_{-1}$	0.0672 (0.5247)	0.0307 (0.5767)	0.1298* (0.0778)	0.0647* (0.0902)
$\mathrm{\Delta }\mathit{G}\mathit{H}\mathit{A}$−1	−0.0071 (0.9726)	0.0330 (0.7581)	0.6057**(0.0391)	−0.3732 (0.4109)
$\mathrm{\Delta }\mathit{S}\mathit{A}-1$	−0.0571 (0.9174)	0.0867 (0.7623)	0.0944 (0.5283)	0.2580 (0.2755)
Threshold Value = −0.1943055, AIC = −609.7106, BIC = −574.6831, SSR = 0.6805009 Cointegrating Vector = (1, − 2.55307), Loglikelihood = 102.8714; Percentage of observation in each regime 51.7% 48.3% Standard errors in parentheses (). * and ** indicate significance at 90% and 95% confidence levels, respectively. Test for Asymmetry of Beef Markets
Variable		$\mathrm{\Delta }\mathit{G}\mathit{H}\mathit{A}$		$\mathrm{\Delta }\mathit{S}\mathit{A}$
$\mathit{E}\mathit{C}\mathit{T}\mathrm{\_}\mathit{P}\mathit{o}{\mathit{s}}_{\mathit{t}-1}$		0.013 [0.794]		0.005 [0.834]
$\mathit{E}\mathit{C}\mathit{T}\mathrm{\_}\mathit{N}\mathit{e}{\mathit{g}}_{\mathit{t}-1}$		−0.299*** [0.001]		−0.098** [0.036]
${\mathit{H}}_{01}:{\mathit{\alpha }}_1^{+}={\mathit{\alpha }}_1^{-}$ For all lags		1.147 [0.346]		2.488* [0.056]
${\mathit{H}}_{02}:{\mathit{\beta }}_1^{+}={\mathit{\beta }}_1^{-}$ For all lags		4.199*** [0.006]		3.781***[0.010]
${\mathit{H}}_{03}:{\mathit{\alpha }}_1^{+}={\mathit{\alpha }}_1^{-}$		0.008 [0.830]		0.046 [0.930]
${\mathit{H}}_{03}:{\mathit{\alpha }}_2^{+}={\mathit{\alpha }}_2^{-}$		2.365 [0.131]		2.835* [0.099]
${\mathit{H}}_{04}:{\mathit{\beta }}_1^{+}={\mathit{\beta }}_1^{-}$		0.744 [0.393]		0.236 [0.630]
${\mathit{H}}_{04}:{\mathit{\beta }}_2^{+}={\mathit{\beta }}_2^{-}$		2.032 [0.161]		4.533** [0.039]
${\mathit{H}}_{05}:\sum _{\mathit{i}=1}^2{\mathit{\alpha }}_1^{+}=\sum _{\mathit{i}=1}^2{\mathit{\alpha }}_1^{+}$		1.169 [0.285]		0.923 [0.342]
${\mathit{H}}_{06}:\sum _{\mathit{i}=1}^2{\mathit{\beta }}_1^{+}=\sum _{\mathit{i}=1}^2{\mathit{\beta }}_1^{+}$		1.760 [0.191]		2.244 [0.141]
${\mathit{H}}_{07}:{\mathit{\delta }}_1^{+}={\mathit{\delta }}_1^{-}$		8.475*** [0.006]		3.394* [0.072]
R2 LB(4) LB(8) AIC BIC		0.33 0.89 0.15 −94.04 −69.52		0.33 0.87 0.71 −167.06 −142.54

${\mathit{\alpha }}_1^{+}\hspace{0.17em}\mathrm{\&}{\mathit{\alpha }}_1^{-}$ are the positive and negative lags for beef price in Ghana while ${\mathit{\beta }}_1^{+}\hspace{0.17em}\mathrm{\&}{\mathit{\beta }}_1^{-}$ are also the positive and negative lags for the beef price of South Africa. The estimates in the curly bracket are the probability values of their corresponding estimated coefficient.Source: Author’s Computation (2021)