The result and discussions of this research show the outcome of the analysis of this research which includes the graphical analysis, descriptive statistics which gives the features of the series employed in the study, the formal pretest, result obtained from estimation, the robustness of the result and the discussion of the result.
Table 41
Descriptive Statistics of the series
|
Mean
|
Max.
|
Min
|
Std. Dev.
|
Jarque-Bera
|
Prob.
|
Obs.
|
Exports to China
|
5.28
|
6.66
|
2.40
|
0.84
|
43.06
|
0.00
|
117
|
Exports to France
|
5.67
|
6.74
|
3.71
|
0.50
|
5.63
|
0.06
|
117
|
Exports to India
|
7.33
|
8.19
|
5.91
|
0.40
|
2.53
|
0.28
|
117
|
Exports to Netherland
|
5.61
|
6.83
|
3.64
|
0.61
|
6.37
|
0.04
|
117
|
Exports to US
|
5.93
|
7.96
|
3.71
|
1.00
|
1.96
|
0.37
|
117
|
Imports from china
|
6.42
|
7.05
|
5.55
|
0.28
|
6.85
|
0.03
|
117
|
Imports from France
|
3.73
|
5.28
|
2.08
|
0.63
|
0.53
|
0.77
|
117
|
Imports from India
|
4.90
|
5.48
|
3.93
|
0.26
|
8.50
|
0.01
|
117
|
Imports from Netherland
|
4.72
|
5.95
|
3.56
|
0.44
|
0.04
|
0.98
|
117
|
Imports from US
|
5.71
|
8.22
|
-1.15
|
0.82
|
8639.37
|
0.00
|
117
|
Naira-Yuan exchange rate
|
4.22
|
7.02
|
3.42
|
0.59
|
43.83
|
0.00
|
117
|
Naira-Pounds exchange rate (France)
|
6.31
|
8.85
|
5.60
|
0.61
|
13.05
|
0.00
|
117
|
Naira-Rupee exchange rate
|
1.94
|
2.67
|
-0.83
|
0.47
|
372.48
|
0.00
|
117
|
Naira pounds exchange rate (Netherland)
|
6.31
|
8.84
|
5.23
|
0.61
|
13.27
|
0.00
|
117
|
Naira-Dollar exchange rate
|
12.47
|
17.93
|
10.22
|
1.45
|
36.06
|
0.00
|
117
|
GDP of China
|
8.08
|
8.41
|
7.51
|
0.20
|
5.30
|
0.07
|
117
|
GDP France
|
6.50
|
6.61
|
6.35
|
0.07
|
7.43
|
0.02
|
117
|
GDP India
|
6.15
|
6.29
|
5.88
|
0.10
|
19.75
|
0.00
|
117
|
GDP Netherland
|
12.26
|
12.41
|
12.13
|
0.06
|
4.30
|
0.12
|
117
|
GDP of US
|
8.42
|
8.50
|
8.28
|
0.05
|
9.13
|
0.01
|
117
|
GDP of Nigeria
|
10.95
|
11.59
|
10.63
|
0.29
|
24.04
|
0.00
|
117
|
Naira-Yuan exchange rate volatility
|
3.78
|
8.61
|
0.94
|
2.12
|
9.74
|
0.01
|
115
|
Naira-pounds exchange rate volatility (France)
|
6.73
|
12.05
|
6.02
|
1.34
|
181.39
|
0.00
|
115
|
Naira-Dollar exchange rate volatility
|
9.18
|
12.11
|
-3.67
|
1.47
|
12224.55
|
0.00
|
113
|
Naira-Rupee exchange rate volatility
|
4.49
|
14.76
|
0.99
|
2.20
|
274.12
|
0.00
|
115
|
Naira –pounds exchange rate volatility (Netherland)
|
5.48
|
9.26
|
3.06
|
1.07
|
45.55
|
0.00
|
115
|
Note: all series are presented as log of the real values, except volatility series.
The result of the descriptive statistics indicates that average value of Nigeria export to India was the highest while highest imports were from China. Expectedly, real naira-dollar exchange rate posted the highest, followed by naira-pounds and then naira-Yuan. In the case of GDP, the log of gross domestic product of Netherland has the highest value of 12.41 followed by log of GDP of Nigeria. This implies that Nigeria GDP grow faster than that of the US and the Asian countries. In fact, the monthly average growth of GDP between 2011 and 2020 indicates that Nigeria grew faster than all the Asian countries and the US.
Table 2
Result of unit root tests
Series
|
Augmented DickyFuller
|
Phillip -Perron
|
Level
|
First Diff.
|
I(d)
|
Level
|
First Diff.
|
I(0)
|
Exports to China
|
-6.92***
|
|
I(0)
|
-6.77***
|
|
I(0)
|
Exports to France
|
-14.07***
|
|
I(0)
|
-8.33***
|
|
I(0)
|
Exports to India
|
-4.52***
|
|
I(0)
|
-7.91***
|
|
I(0)
|
Exports to Netherland
|
-4.63***
|
|
I(0)
|
-4.47***
|
|
I(0)
|
Exports to US
|
-10.64***
|
|
I(0)
|
-3.39*
|
|
I(0)
|
Imports from china
|
-3.46**
|
|
I(0)
|
-7.62***
|
|
I(0)
|
Imports from France
|
-6.73***
|
|
I(0)
|
-6.47***
|
|
I(0)
|
Imports from India
|
-7.63***
|
|
I(0)
|
-7.95***
|
|
I(0)
|
Imports from Netherland
|
-5.62***
|
|
I(0)
|
-5.58***
|
|
I(0)
|
Imports from US
|
-6.31***
|
|
I(0)
|
-9.67***
|
|
I(1)
|
Naira-Yuan exchange rate
|
-2.54
|
-10.76***
|
I(1)
|
1.01
|
-3.21*
|
I(0)
|
Naira-Pounds exchange rate (France)
|
-6.13***
|
|
I(0)
|
-6.09***
|
|
I(1)
|
Naira-Rupee exchange rate
|
2.94
|
-10.44***
|
I(1)
|
2.51
|
-10.44***
|
I(0)
|
Naira pounds exchange rate (Netherland)
|
-11.61***
|
|
I(0)
|
-5.00***
|
|
I(1)
|
Naira-Dollar exchange rate
|
1.04
|
-7.18***
|
I(1)
|
1.95
|
-6.01***
|
I(0)
|
GDP of China
|
-2.94
|
-6.88***
|
I(0)
|
2.40
|
-6.88***
|
I(0)
|
GDP France
|
-1.93
|
-8.45***
|
I(0)
|
3.10
|
-8.89***
|
I(0)
|
GDP India
|
-1.06
|
-8.09***
|
I(0)
|
2.28
|
-8.42***
|
I(0)
|
GDP Netherland
|
-4.19***
|
|
I(0)
|
-7.14***
|
|
I(1)
|
GDP of US
|
-2.89
|
-13.23***
|
I(1)
|
2.45
|
-6.42***
|
I(0)
|
GDP of Nigeria
|
-8.31***
|
|
I(0)
|
-8.86***
|
|
I(0)
|
Naira-Yuan vol.
|
-8.72
|
|
I(0)
|
-8.45***
|
|
I(0)
|
Naira-pounds vol. (France)
|
-3.32*
|
|
I(0)
|
-3.46**
|
|
I(0)
|
Naira-Dollar vol.
|
-7.05***
|
|
I(0)
|
-7.03***
|
|
I(0)
|
Naira-Rupee vol.
|
-10.68***
|
|
I(0)
|
-10.68***
|
|
I(0)
|
Naira –pounds vol. (Netherland)
|
-4.79***
|
|
I(0)
|
-4.66***
|
|
I(0)
|
Note: *** =significant at 1%, **Significant at 5%,
This could be traced to quick recovery of the economy from the global economic crisis and fiscal stimulus provided by the government during oil price slump in 2016 coupled with periods of elections that took place in 2011, 2014 and 2019. These periods of elections are always accompanied by large fiscal expansion which then metamorphosed into economic activity and hence increase in GDP. The average real exchange rate volatility of naira-dollar is the highest, posting 9.18. This is followed by naira-pounds (France), posting 673. In terms of the maximum values of real exchange rate volatility, naira-Rupee had the highest maximum value followed by naira-dollar. Generally, it can be observed that real exchange rate of naira per unit of each foreign currency is high and this could have implication for trade flows between Nigeria and each of the partners.
Further, the probability value of the Jarque-Bera statistics suggests that all the series, except the volatility series are not normally distributed, indicating that there tends to be evidence of outliers in the series. Consequently, ordinary least square cannot be an appropriate estimation method to employ and this is one of the reasons why nonlinear ARDL is considered. The ARDL recognizes and correct for possible volatility of the series and also deal with issues surrounding outliers and omitted variables.
Meanwhile, before it can be guaranteed that that NARDL must be employed, it remains to inspect the nature of the series. To justify the application of NARDL, it must be the case that the series are a combination of being stationary at level and first difference. Table 2 provides information about the nature of the series and as can be observed, some series such as volatility and exports are stationary level because these series do not possess unit root at levels. However, series such as imports, bilateral exchange rate and GDP are observed to be stationary at first difference.
Table 3
Bounds Test Result
EXPORT MODELS
|
|
CHINA
|
FRANCE
|
INDIA
|
US
|
NETHERLAND
|
F-Stat
|
9.23**
|
18.06***
|
4.33**
|
1.75
|
5.09***
|
IMPORT MODELS
|
|
CHINA
|
FRANCE
|
INDIA
|
US
|
NETHERLAND
|
F-Stat
|
8.33***
|
23.18***
|
8.14***
|
1.08
|
12.19***
|
Note: values are F-Statistic; ***Significant at 1%; ** significant at 5%
The F-statistics of the bound test result of the export and import models shows that there is a long run relationship among the variables as the values are higher than the lower and upper bounds critical values of 3.03 and 4.06 at 5% level of significance respectively, except in the case of U.S where the long run convergence cannot be established. What this implies is that exports to and import from US models do not exhibit long run convergence, and hence only short run dynamics of the models will be estimated and discussed. In Tables 1 and 2, real exchange rate volatilities were reported, albeit it was noted earlier that these series are not observable. However, we also argued that ARCH-type method be employed to generate the series on the ground that the volatility can be time-dependent. Owing to the limited space, it is not possible to show full result from where the appropriate ARCH-type model was chosen. However, this can be made available on request.
Table 4
Fitted GARCH (1,1) models for the real bilateral exchange rate
Model
|
US
|
CHINA
|
FRANCE
|
INDIA
|
NETHERLAND
|
AR(1)
|
0.24**
(0.004)
|
-0.08***
(0.000)
|
-0.78***
(0.000)
|
0.01
(0.18)
|
0.96**
(0.004)
|
C
|
-23.74***
(0.00)
|
3.34***
(0.00)
|
3.82***
(0.00)
|
0.03***
(0.00)
|
4.84***
(0.00)
|
RESID(-1)^2
|
0.87***
(0.000)
|
28.57***
(0.000)
|
3.38**
(0.002)
|
-0.08*
(0.07)
|
0.18**
(0.002)
|
RESID(-1)^2*(RESID(-1) < 0)
|
|
|
|
0.21*
(0.07)
|
|
GARCH(1)
|
-0.13**
(0.003)
|
0.01***
(0.000)
|
0.45*
(0.005)
|
0.12**
( 0.004)
|
1.03**
(0.004)
|
Constant
|
64.55***
(0.000)
|
5.68
(0.29)
|
25.780**
(0.002)
|
182.70*
(0.09)
|
-7.05*
(0.05)
|
ARCH-LM TEST (p-values)
|
0.7837
|
0.0927
|
0.8557
|
0.5331
|
0.8931
|
Note: Values in parentheses are probabilities values. Note: values are F-Statistic; ***Significant at 1%; ** significant at 5%
Nevertheless, the result of GACH (1,1) chosen by the model selection criterion (Schwatz Information Criterion) is presented in Table 4. As can be read off, all the bilateral real exchange rate possess ARCH effect, and the ARCH is time varying as shown by the significance of the error term. But when GARCH (1,1) is estimated, the LM test indicates no time-dependence again. Hence, the fitted values of the GARCH (1,1) is recovered and incorporated as volatility series in the exports and imports models.
In what follows, we now present the results of estimation models in equations 9 and 10. Eq. 9 is export models for Nigeria with respect to each trading partner while Eq. 10 is the imports counterparts. Table 5 shows results of both short run dynamic and long run model of exports to China, France, India, Netherland, and US. One period lag of exports to China, France, US and Netherland enhances current exports.
However, one period lag of exports reduces current exports to India. The positive effect in the case of the former countries suggests that goods exported to those countries are preferred. Also, it could be as a result of migrant network effect where migrants provide information about the preferred products in the host countries and this information facilitates exports, particularly exports of non-oil products. Nigeria has been benefitting from migrant information to increase exportation to the European countries, particularly Netherland. Also, the share of non-oil export in total export to China within the period under review is less than the case of exports to India. This suggests that some industrial and agricultural products tend to attract preferences from China as well as France, and the US. The reason for negative effect of lagged exports on current exports to India can be traced to large percentage of oil in total export the magnitude of effect suggests that one period exports affects current exports to Netherland more than any other country under study.
Discussing the effect of foreign income on Nigeria exports, the Table indicates that Indian income enhances more exports from Nigeria than other countries. Although income of other countries is positive and important, they are not significant. One seeming reason for this insignificant, albeit, positive is that products exported China, France, US and Netherland are inferior, not competitive in the respective foreign country’s markets. Also, trade regulation in form of product standards may have stunt the possible effect that increase in income would have provided. Further, if there is low or acute product diversification or development of new products, exports may not increase following increase in income. Like foreign income, real bilateral exchange rate shows insignificant effect on exports to China, France, and India. However, one lag period of bilateral real exchange rate affects exports to these countries while current naira-dollar real exchange rate significantly affects exports to the US.
In the case of export to Netherland, no real exchange rate effect was observed either current or previous. This implies that while real bilateral exchange rate delays by one period to significantly affect exports to China, France and India, it has immediate effect on the export to the US. In terms of direction of effect, increase in real bilateral exchange rate (real depreciation) of naira in relation to dollar reduces exports to the US while one period lag increase in real bilateral exchange rate (previous depreciation) enhance exports to China, France and India. However, instantaneous (current) increase (depreciation) in naira-dollar exchange rate retards exports to the US. What this implies is that Nigerian products appear not to be strongly competitive in the US market. Real depreciation can strangulate exports to the US if relative price is so high that it makes exports to be relatively expensive and consequently leads to reduction in exports following real depreciation.
Table 5
Results of Short Run and Long Run NARDL model of Nigeria Export to her trading partners
Variables
|
CHINA
|
FRANCE
|
INDIA
|
US
|
NETHERLAND
|
\(\varDelta ln{EXP}_{-1}\)
|
0.13**
(0.002)
|
0.41**
(0.004)
|
-0.28***
(0.00)
|
0.37***
(0.00)
|
0.67*
( 0.07)
|
\(\varDelta {lnY}_{t}^{f}\)
|
0.19
(0.52)
|
0.17
0.52)
|
0.62**
(0.03)
|
0.43
(0.17)
|
0.32
(0.67)
|
\({\varDelta RBEX}_{t}\)
|
0.03
(0.33)
|
-0.09
(0.11)
|
0.04
(0.1)
|
-0.01*
(0.04)
|
0.06
(0.11)
|
\({\varDelta RBEX}_{,-1}\)
|
0.01**
(0.002)
|
0.09**
(0.04)
|
0.06*
(0.24)
|
0.21
(0.15)
|
0.16
(0.27)
|
\({\varDelta POS}_{t}\)
|
0.05
(0.2)
|
0.11**
(0.04)
|
0.00
(0.11)
|
-0.03*
(0.04)
|
0.01*
(0.05)
|
\({\varDelta NEG}_{t}\)
|
0.02*
(0.05)
|
-0.18**
(0.04)
|
0.01*
(0.07)
|
0.01*
(0.04)
|
-0.03*
(0.05)
|
\({\xi }_{t-1}\)
|
-0.16*** (0.000)
|
-0.90***
(0.000)
|
-0.46***
(0.000)
|
|
-0.66**
(0.004)
|
Long Run Coefficients
|
\({c}_{1}\)
|
2.35
(0.18)
|
2.9
(0.74)
|
-0.79
(0.92)
|
|
7.23
(0.86)
|
\({lnY}_{t}^{f}\)
|
0.3
(0.86)
|
0.19
(0.57)
|
0.35*
(0.08)
|
|
0.16
(2.09)
|
\({RBEX}_{t}\)
|
0.02
(0.55)
|
0.33*
(0.18)
|
0.13
(0.24)
|
|
-0.19
(0.35)
|
\({POS}_{t}\)
|
0.86
(0.34)
|
0.11**
(0.04)
|
0.01*
(0.03)
|
|
0.06*
(0.03)
|
\({NEG}_{t}\)
|
-0.08
(0.1)
|
-0.21***
(0.04)
|
-0.02*
(0.03)
|
|
0.11
(0.15)
|
Adjusted R2
|
0.19
|
0.43
|
0.36
|
0.75
|
0.58
|
Diagnostic Tests
|
LM
|
0.42
|
0.12
|
0.96
|
0.76
|
0.3
|
ARCH-LM
|
0.02
|
0.49
|
0.98
|
0.64
|
0.42
|
RESET
|
0.36
|
0.13
|
0.21
|
0.36
|
0.82
|
J-B
|
0.00
|
0.00
|
0.41
|
0.11
|
0.00
|
Where \({\xi }_{t-1}\) \({c}_{1}\), \({lnY}_{t}^{h}\),\({RBEX}_{t}\), \({POS}_{t}\), \({NEG}_{t}\) means long run coinegrating coefficient, constant, log of each trading partner’s real GDP, bilateral real exchange rate, positive real exchange rate volatility and negative real exchange rate volatility respectively, just as defined in Eq. 9
Values in parentheses are probabilities values. Note: values are F-Statistic; ***Significant at 1%; ** significant at 5%, *significant at 10%.
The available evidence suggests that the relative price of goods in Nigeria and the US is the highest when compared to relative price of goods in Nigeria and each of other major trading partners. Hence, as much as nominal exchange rate depreciation would have made exports cheap, the high relative price dampens the benefit. A cursory look at the magnitude of effect indicates that bilateral real exchange rate has small effect on export to China, followed by France, India and then the US. This indicates that a large swing in bilateral real exchange rate will be accompanied by small effect on exports.
Results of nonlinear bilateral real exchange rate is established in this study, that is, bilateral real exchange rate behaves asymmetrically, in which case, the change with time is important in influencing exports. However, the nature of asymmetry differs across trading partners. Also, the direction of effect is also not uniform. Real bilateral exchange rate reports both positive and negative asymmetry, but they are significant in affecting exports to France, India, US and Netherland. In China, only negative asymmetry change in bilateral real exchange rate affects exports. Specifically, positive change (depreciation) encourages exports to France, India and Netherland but discourages exports to the US. Negative change reduces exports to France and Netherland (European countries) and enhances exports to China, India (Asian countries) and the US (America). Again, it can be observed that the magnitude of effect is small.
Speed of convergence is the highest in exports to France and then export to China. The model for the US indicates no long run effect and so, no speed of adjustment is observed. For any distortion in the export model to France, the model will adjust by 90% so that the rest 10% will take place in the immediate period. Since the data are monthly, it follows that it takes around 33 days for equilibrium to be restored after a distortion. The adjustment in the current period will be around 66% for a distortion to exports to Netherland and the remaining 34% will be restored in the following month, hence it will take around 43 days for the system to adjust fully. In the same vein, 46% and 16% of the distortion to exports to India and China respectively. This implies that it takes correspondingly around 63days and more than 6 months for a distortion to exports India and China to adjust to equilibrium.
In the long run, foreign income does not play any role in influencing Nigerian exports to all the countries under study. While bilateral real exchange rate appears not to play any long run role in exports from Nigeria to its major trading partners there is evidence of asymmetry bilateral real exchange rate behaviour in the long run. In particular, positive change in bilateral real exchange rate of naira to pound enhances exports to France and Netherland in the same vein positive increase in bilateral real exchange rate of naira to Rupee encourages exports to India. In case of exports to China, as there is no positive asymmetry in the short run so also there is none in the long run. In fact, there is no significant negative asymmetry behaviour of naira-Yuan on exports to China. Meanwhile, negative increase in bilateral real naira-pound exchange rate drags exports to France. In the same vein, naira-Rupee bilateral real exchange rate shows significant negative asymmetry. In this regard, negative increase (reduction in depreciation) will dampen exports to India. This result suggests that asymmetry behaviour plays important role in influencing exports of Nigeria to its major trading partners.
Various diagnostic tests upon which these results are explained suggest that the models are reliable and valid. The Jacque-Bera (J-B) probabilities show no presence of outliers in the model and that the series are normally distributed. The hypotheses that the models possess no serial correlation (BP-LM) and that the variance is constant (ARCH-LM) are all upheld. Hence, the respective coefficients are unbiased, efficient and also consistent.
Table 6 shows the result of short run dynamic and long run of bilateral import demand model of Nigeria in relation to its five major trading partners, that is, China, India, France, Netherland and the US. In the short run, it can be observed that the one period lag of Nigeria imports has a positive and significant effect on contemporaneous import from all the countries. Specifically, a one percent increase in the lag value of Nigeria import will lead to 0.43%, 0.44% and 0.33% increase in current import from China, France, India, US and Netherland respectively. This indicates that preferences for imported products are still high. The high preference may not be unconnected with technological dynamics taking place in the advanced countries and the emerging Asian countries. These technological advancements, through active and effective Research and Development (R&D) allows more goods of diverse varieties to be produced and hence attracts interest among Nigerians.
This result suggests that Nigerians’ preference for imported products from advanced countries is increasing instead of decreasing. However, a cursory look at the magnitude of effect indicates that preferences are more for products coming from Asian countries (India and China in that order) and less in America (US). This s not surprising because these countries (China and India) tend to produce in large quantity, light manufactured products that are affordable among middle and low income earners in Nigeria. Examples of such products are mobile phones, tablets, medical equipment, and medicines among others. Apart from preference driving importation from these countries, ability to pay also facilitates it. The result indicates that increase in Nigeria real income (measured by real GDP) engenders more import demands from China, India, US and Netherland. It is however surprising that the case is different for importation from France, where negative effect is observed. But such effect is not significant, suggesting that Nigeria income appears not to play effective role in importation from France.
Table 6
Results Short Run and Long Run NARDL model of Nigeria import to her major trading partners
Variables
|
CHINA
|
FRANCE
|
INDIA
|
US
|
NETHERLAND
|
\(\varDelta ln{IMP}_{-1}\)
|
0.43**
(0.000)
|
0.13**
(0.002)
|
0.44***
(0.000)
|
0.11**
(0.004)
|
0.33***
(0.000)
|
\(\varDelta {lnY}_{t}^{h}\)
|
0.08***
(0.004)
|
-0.33
(0.21)
|
0.14*
(0.09)
|
0.24**
(0.002)
|
0.11**
(0.009)
|
\({\varDelta RBEX}_{t}\)
|
0.06*
(0.09)
|
-0.033*
(0.04)
|
-0.10*
(0.06)
|
0.12*
(0.07)
|
0.02
(0.12)
|
\({\varDelta POS}_{t}\)
|
-0.12**
(0.004)
|
-0.01*
(0.03)
|
0.01*
(0.01)
|
0.04*
(0.05)
|
-0.11*
(0.06)
|
\({\varDelta NEG}_{t}\)
|
-0.03**
(0.002)
|
-0.02**
(0.003)
|
0.01*
(0.01)
|
-0.02**
(0.003)
|
-0.08*
(0.05)
|
\({\xi }_{t-1}\)
|
-0.57***
(0.000)
|
-0.03***
(0.000)
|
-0.56***
(0.000)
|
==
|
-0.66**
(0.000)
|
Long Run Coefficients
|
\({g}_{1}\)
|
4.12**
(0.001)
|
9.40***
(0.000)
|
2.95***
(0.000)
|
|
2.81***
(0.000)
|
\({lnY}_{t}^{h}\)
|
0.13**
(0.006)
|
-0.33
(0.21)
|
0.24**
(0.003)
|
|
0.17
(0.29)
|
\({RBEX}_{t}\)
|
0.11*
(0.07)
|
-0.24
(0.13)
|
0.02*
(0.07)
|
|
0.01
(0.18)
|
\({POS}_{t}\)
|
-0.21**
(1.10)
|
-0.01*
(0.03)
|
0.04*
(0.05)
|
|
-0.18**
(0.009)
|
\({NEG}_{t}\)
|
-0.06**
(0.03)
|
-0.00*
(0.04)
|
-0.02**
(0.003)
|
|
-0.11*
(0.08)
|
Adjusted R2
|
0.46
|
0.45
|
0.46
|
0.21
|
0.16
|
Diagnostic (post estimation) test
|
LM
|
0.56
|
0.65
|
0.06
|
|
0.18
|
ARCH-LM
|
0.43
|
0.44
|
0.69
|
|
0.51
|
RESET
|
0.52
|
0.11
|
0.57
|
|
0.36
|
J-B
|
0.43
|
0.92
|
0.34
|
|
0.29
|
Where \({\xi }_{t-1}\) \({g}_{1}\), \({lnY}_{t}^{h}\),\({RBEX}_{t}\), \({POS}_{t}\), \({NEG}_{t}\) means long run cointegrating coefficient, constant, log of Nigeria real GDP, bilateral real exchange rate, positive real exchange rate volatility and negative real exchange rate volatility respectively, just as defined in Eq. 10
Values in parentheses are probabilities values. Note: values are F-Statistic; ***Significant at 1%; ** significant at 5%, *significant at 10%.
Short run dynamic of bilateral real exchange rates positively and significantly affects imports from China and the US, negatively and significantly affects imports from France and India while no significant effect is observed in the case of imports from Netherland. This implies that real depreciation of naira with respect to Yuan and dollar enhances importation from the respective country. These countries (China and US) are well known for their strong market competition in the world in general and in Nigeria in particular. The strong competition may have driven relative price down such that even in the presence of nominal depreciation that may make the goods expensive, the influence of low relative price still motive economic agents to purchase more imports even in the face of real depreciation. Albeit, the magnitude of effect tends to be generally mild as it will post 0.6% and 1.2% increase in importation from China and the US following a 10% real depreciation of naira to Yuan and dollar respectively.
Clearly, real bilateral exchange rate effect on imports differs across trading partners, having more effect on the import from the US. The negative effect of real bilateral exchange rate depreciation of naira in relation to Pounds (for France) and Rupee (for India) implies that imports from these countries are expensive in real term. In this case, relative price is not as low as to motivate agents to increase import demand. Of course lack of massive product varieties which can engender low price can also be a reason. For instance, major imported products coming from India are medical equipment and drugs, while those from France are heavy duty machines, unlike China and the US where a combination of high and heavy manufactured products alongside array of varieties of other products are produced and available for demand. Nevertheless, this line of argument is not full proof and so, further investigation is required.
The investigation should focus on import demand at product level, say, 3-digit product classification. The insignificant effect of bilateral exchange rate on imports from Netherland can be attributed to the nature of products imported from the country. Descriptive statistics suggest that in relative term, Nigeria imports more of agricultural products from Netherlands than others within the period under review. Given the role of agricultural products, it is not surprising that the influence of bilateral exchange rate on imports from Netherland is not recognized.
Do bilateral exchange rate volatilities behave asymmetrically in the Nigeria export and import models? The answer is yes and supported by the result provided in Table 6. It is also important to note, as revealed in the Table that the asymmetric behaviour can be positive, negative or both as the case may be. However, the direction of effect differs across country. Imports from China, Franc and Netherland are negatively affected by both positive and negative asymmetric behaviour of real bilateral exchange rate volatility. In particular, positive change in bilateral real exchange rate volatility (real depreciation) reduces imports from China, France and Netherland, albeit the magnitude is mostly pronounced for import from China.
The indication therefrom is that most importers importing goods from these countries are risk averse, such that during positive change in exchanger rate volatility, then reduce importation. A negative change in bilateral real exchange rate also reports negative effect on imports from China. This still buttress the earlier point that importers in this regard are risk averse. Unlike the case of import from China, France and Netherland, positive change in naira-Rupee and naira-dollar exchange rate volatility engenders more imports from the respective country. Hence, importers patronizing India and US products tend to be risk neutral.
The speed of convergence indicates that each model where long run is established converges to long run equilibrium. The speed of adjustment is mostly pronounced in the import from Netherland, followed by imports from China, India and then France. Convergence in the case of import from France is the lowest because only 3% adjustment will be made in the current month following a disturbance in the model. Consequently, it will take 10 months from the month when the disturbance occurs for the system to settle at a new equilibrium. This is unlike the case of imports from Netherland where 66% adjustment will be made for any disturbance in the import model with respect to that country. Hence, it takes less than 2 months for the new equilibrium to be established.
In the long run, Nigeria income positively affects imports from China and India. Although Nigeria income is also important to imports from France and Netherland, it is not significant in each case. This suggests that in the long run, increase in Nigeria income may not influence demand for imports from these countries. Bilateral real exchange rate has positive long run effect imports from China, India and Netherland. There is no long run significant effect of bilateral real exchange rate on imports from France. Like the case of short run dynamic, there is also long run bilateral real exchange rate asymmetric behaviour. Virtually all the trading partners’ exchange rate behaves asymmetrically in the long run. Positive change in bilateral real exchange rate volatility inhibits imports from China, France and Netherland in the long run. Conversely, positive change in bilateral real exchange rate volatility enhances importation from India. Therefore, it can be conjectured that importers react to asymmetric behaviour of bilateral real exchange rate volatility both in the short and long run. Also, it is also clear that not all importers are risk averse to dynamics of real exchange rate. While importers patronizing China and France markets are risk averse, those patronizing India appears to be risk neutral.
In summary, bilateral real exchange rate volatility possess asymmetry behaviour in the export and import models of Nigeria in relation to the major trading partners considered in this study. However, the direction and magnitude of effect differ across trading partners. All the diagnostic tests for the all the import models report satisfactory values, indicating that there is no evidence of serial correlation or time-varying variance also, there is no evidence of omission of important variables in the models. Further, the normality condition of the model is satisfied.
The gross domestic product has the expected positive coefficients but has no significant effect on the value of Nigeria import. The real bilateral exchange rate in all cases has no significant effect on the value of Nigeria import. The positive volatility measure of exchange rate has a negative significant effect on the value of Nigeria import value in some cases (China, Netherland). Specifically, a one percent increase in positive volatility of naira-yuan in India will lead to -0.12% increase in the value of Nigeria import at 5% significance level. Also, a one percent increase in the positive volatility of naira-euro in euro will lead to -0.11% increase in the value of Nigeria import at 1% significance level. The negative volatility measure of exchange rate has a negative significant effect on the value of Nigeria import in the case of China alone. Specifically, a one percent increase in the value of naira-yuan in India will lead to -0.03% increase in the value of Nigeria import at 5% significance level.
The short run effect lasted into the long run in some cases (China, U.S. and Netherland). The one period lag of Nigeria import value has a negative significant effect on the value of Nigeria import in U.S. alone. Specifically, a one percent increase in the one period lag value of import will lead to -0.96% increase in the value of Nigeria import at 10% significance level. The gross domestic product of Nigeria is in line with the apiori expectation in all cases but not significant. There is no considerable difference in the real bilateral exchange rate effect on the value of Nigeria import in all cases. The positive volatility measure of exchange rate has a negative significant effect on the value of Nigeria import in few cases (China, Netherland). Precisely, a one percent increase in the positive volatility of naira-yuan in China will lead to -0.21% increase in the value of Nigeria import at 5% significance level. Also, a one percent increase in the positive volatility of naira-euro in Netherland will lead to -0.18% increase in the value of Nigeria import under the considered period. The negative volatility measure of exchange rate has a negative significant effect on the value of Nigeria import under the considered period (2011M4-2020M12) in China alone. Precisely, a one percent decrease in the negative volatility of naira-Yuan in China will lead to -0.06% increase in the value of Nigeria import under the considered period.
The adjusted R2 shows that 46% of the explanatory variables explain the variation in the dependent variable in the case of China, 45% in the case of France, 46% in India, 21% in U.S. and 16% in Netherland. It can be observed that the adjusted R2 is low in the case of U.S. and Netherland. This shows that there are other factors that influence the Nigeria import value in the case of U.S. and Netherland.
The result of import models shows significant asymmetry effect of exchange rate volatility on Nigeria import in China and Netherland. This effect is seen in the short run which lasted into the long run. Specifically, positive volatility of naira-yuan has a significant negative effect on the value of Nigeria import while negative volatility of naira-yuan has a negative significant effect on the value of Nigeria import under the considered period ( 2011M4-2020M12) in the short and long run. Also, the positive volatility of naira-euro in Netherland has a significant negative effect on the value of Nigeria import from Netherland in the short run which lasted into the long run.
From table 8 presented which shows the diagnostic result of the export and import model. It can be observed that there is high speed of adjustment from the short run to the long run should there be a shock or a disequilibrium in the system in all the export and import models. The null hypothesis of serial correlation from the L.M. test is rejected in all the export and import models estimated. The null hypothesis of heteroskedasticity from the result of the ARCH-LM test is rejected in all case except export model of China. The RESET test result shows that the null hypothesis of misspecification in all the models is rejected indicating that all models are properly specified. The J-B statistics (Jarque-Bera Statistics) probability value rejects the null hypothesis of normal distribution in some cases in the export and import models. This includes France, China, U.S. and Netherland in the export models and China, France, India, US in the import models.