Hypotheses

Variables 1st Model (Exports of Goods and Services) 2nd Model (Exports of Goods)
Exchange Rate - -
Employment to Population ratio + +
Foreign reserves + +
Inflation - -
GDP per Capita + +
Real Interest Rate + +
Quality of Port Infrastructure + +

Research Type
A quantitative research strategy has been adopted and statistical tools (Pearson correlation and
regression) have been used on a structured data panel including the years 2007-2017. A sample of 10
countries was attained of which 5 were oil exporting and 5 were oil importing.

Independent Variable
The independent variables for this model is Exchange rate, Employment to population, GDP per Capita,
Foreign reserves, Inflation, Interest rate, Manufacturing, Quality of port infrastructure.
The data was used from the official database pertaining to the World Bank and IMF.

Dependent Variable
Exports of Goods and Services (1st Model)
Exports of Goods (2nd Model)
The data source for the dependent variable was used from the World Bank.

TimeLine
10 years of data (2007-2017) has been included.

Statistical Tests
The two tests that have been conducted are Regression Analysis and Correlation test by using EViews.
Data Analysis

-Correlation Matrix (1st Model “Exports of Goods and Services”)

-Correlation Matrix (2nd Model “Exports of Goods”)

Results

Exports and Exchange rate

Exchange rate shows in the first run a value of 0.117. This shows that for every increase in exchange
rate, there is an increase by 0.117 in exports of goods and services.

In the second run exchange rate shows a value of 0.091. This shows that for every increase in exchange
rate, there an increase by 0.091 in exports of goods.

This shows how services are affecting the correlation between exchange rate and exports positively, as
there is a difference by 0.026.

Exports and Employment to population ratio

Employment shows in the first run a value of 0.068. This shows that for every increase in employment,
there is an increase by 0.068 in exports of goods and services.
In the second run employment shows a value of 0.093. This shows that for every increase in
employment, there is an increase by 0.093 in exports of goods.

This shows how services are affecting the correlation between employment and exports negatively, as
there is a difference by 0.025.

Exports and foreign reserves

Foreign reserves shows in the first run a value of 0.911. This shows that for every increase in foreign
reserves, there is an increase by 0.911 in exports of goods and services.

In the second run foreign reserves shows a value of 0.913. This shows that for every increase in foreign
reserves, there is an increase by 0.913 in exports of goods.

This shows how services are affecting slightly the correlation between foreign reserves and exports
negatively, as there is a difference by 0.001 which mean that the exports of goods and services both are
affected by foreign reserves at the same degree which is high as the results shows 0.9 correlation which
mean that there is a strong relationship between foreign reserves and exports.

Exports and GDP per capita

GDP per capita shows in the first run a value of 0.352. This shows that for every increase in GDP per
capita, there is an increase by 0.352 in exports of goods and services.

In the second run GDP per capita shows a value of 0.379. This shows that for every increase in GDP per
capita, there is an increase by 0.379 in exports of goods.

This shows how services are affecting the correlation between GDP per capita and exports negatively, as
there is a difference by 0.027.

Exports and Inflation

Inflation shows in the first run a value of 0.094. This shows that for every increase in Inflation, there is
an increase by 0.094 in exports of goods and services.
In the second run Inflation shows a value of 0.069. This shows that for every increase in Inflation, there
is an increase by 0.069 in exports of goods.

This shows how services are affecting the correlation between Inflation and exports positively, as there
is a difference by 0.025.

Exports and Interest rate

Interest rate shows in the first run a value of -0.124. This shows that for every increase in Interest rate,
there is a decrease by 0.124 in exports of goods and services.

In the second run Interest rate shows a value of -0.103. This shows that for every increase in Interest
rate, there is a decrease by 0.103 in exports of goods.

This shows how services are affecting the correlation between Interest rate and exports negatively, as
there is a difference by -0.021.

Exports and Manufacturing

Manufacturing shows in the first run a value of -0.035. This shows that for every increase in
Manufacturing, there is a decrease by 0.035 in exports of goods and services.

In the second run Manufacturing shows a value of -0.029. This shows that for every increase in
Manufacturing, there is a decrease by 0.029 in exports of goods.

This shows how services are affecting the correlation between Manufacturing and exports negatively, as
there is a difference by -0.006.

Exports and Quality of port infrastructure

Quality of port infrastructure shows in the first run a value of 0.124. This shows that for every increase
in Quality of port infrastructure, there is an increase by 0.124 in exports of goods and services.
In the second run Quality of port infrastructure shows a value of 0.145. This shows that for every
increase in Quality of port infrastructure, there is an increase by 0.145 in exports of goods.

This shows how services are affecting the correlation between Quality of port infrastructure and exports
negatively, as there is a difference by 0.021 which is expected as quality of port should affect only
exports of goods as no need for port in exporting services.
Normal Distribution of Residuals (1st Run “Exports of Goods and Services”)

Normal Distribution of Residuals (2st Run “Exports of Goods”)

As any econometric model should account for the presence of a residual or error term, and the residuals
must show a normal distribution.
To handle the non-normal distribution, Least squares method was used as the method of least squares
(LS) does not require that the data have normally distributed error for its validity.
Coefficient of Confidence Intervals (1st Run “Exports of Goods and Services”)

Coefficient Confidence Intervals

Date: 06/08/19 Time: 21:09
Sample: 2007 2017
Included observations: 110

95% CI
Variable Coefficient Low High

C 2.03E+10 -2.40E+10 6.45E+10

EXCHANGE_RATE 553165.5 -492141.7 1598473.
EMPLOYMENT_TO_POPULATION -5.11E+08 -1.51E+09 4.87E+08
FOREIGN_RESERVES 0.456051 0.418553 0.493549
GDP_PER_CAPITA 1937839. 1020472. 2855205.
INFLATION 2.38E+09 1.31E+09 3.45E+09
INTEREST_RATE 1.41E+09 2.80E+08 2.54E+09
MANUFACTURING -2.73E+08 -3.55E+09 3.00E+09
QUALITY_OF_PORT_INFRASTR -1.59E+09 -1.26E+10 9.41E+09

A Coefficient of Confidence Intervals Table shows the range of values that you can be 95% certain
contains the true mean of the population.

 Exchange rate 95% of sample should be between -492141.7 and 6.45E+10

 Employment to population should be between -1.51E+09 and 4.87E+08
 GDP per Capita should be between 1020472. and 2855205.
 Foreign reserves should be between 0.418553 and 0.493549
 Inflation should be between 1.31E+09 and 3.45E+09
 Interest rate should be between 2.80E+08 and 2.54E+09
 Manufacturing should be between -3.55E+09 and 3.00E+09
 Quality of port infrastructure should be between -1.26E+10 and 9.41E+09
Coefficient of Confidence Intervals (2nd Run “Exports of Goods”)

Coefficient Confidence Intervals

Date: 06/08/19 Time: 21:50
Sample: 2007 2017
Included observations: 110

95% CI
Variable Coefficient Low High

C 4.85E+09 -3.72E+10 4.69E+10

EXCHANGE_RATE 227813.9 -766471.7 1222099.
EMPLOYMENT_TO_POPULATION -6.95E+08 -1.64E+09 2.55E+08
FOREIGN_RESERVES 0.443848 0.408180 0.479516
GDP_PER_CAPITA 2199277. 1326687. 3071867.
INFLATION 2.38E+09 1.36E+09 3.39E+09
INTEREST_RATE 1.63E+09 5.57E+08 2.71E+09
MANUFACTURING -32738870 -3.15E+09 3.08E+09
QUALITY_OF_PORT_INFRASTR 9.36E+08 -9.53E+09 1.14E+10

A Coefficient of Confidence Intervals Table shows the range of values that you can be 95% certain
contains the true mean of the population.

 Exchange rate 95% of sample should be between -766471.7 and 1222099.

 Employment to population should be between -1.64E+09 and 2.55E+08
 GDP per Capita should be between 1326687. and 3071867.
 Foreign reserves should be between 0.408180 and 0.479516
 Inflation should be between 1.36E+09 and 3.39E+09
 Interest rate should be between 5.57E+08 and 2.71E+09
 Manufacturing should be between -3.15E+09 and 3.08E+09
 Quality of port infrastructure should be between -9.53E+09 and 1.14E+10
  Regression (1st Model Exports of Goods and Services)

Dependent Variable: EXPORTS_OF_GOODS_AND_SER

Method: Panel Least Squares
Date: 06/08/19 Time: 21:08
Sample: 2007 2017
Periods included: 11
Cross-sections included: 10
Total panel (balanced) observations: 110

Variable Coefficient Std. Error t-Statistic Prob.

C 2.03E+10 2.23E+10 0.908898 0.3656

EXCHANGE_RATE 553165.5 526940.0 1.049769 0.2963
EMPLOYMENT_TO_POPULATION -5.11E+08 5.03E+08 -1.015483 0.3123
FOREIGN_RESERVES 0.456051 0.018903 24.12597 0.0000
GDP_PER_CAPITA 1937839. 462445.2 4.190418 0.0001
INFLATION 2.38E+09 5.37E+08 4.428711 0.0000
INTEREST_RATE 1.41E+09 5.71E+08 2.474211 0.0150
MANUFACTURING -2.73E+08 1.65E+09 -0.165180 0.8691
QUALITY_OF_PORT_INFRASTR -1.59E+09 5.54E+09 -0.286731 0.7749

R-squared 0.909583 Mean dependent var 6.82E+10

Adjusted R-squared 0.902421 S.D. dependent var 8.52E+10
S.E. of regression 2.66E+10 Akaike info criterion 50.92528
Sum squared resid 7.15E+22 Schwarz criterion 51.14623
Log likelihood -2791.891 Hannan-Quinn criter. 51.01490
F-statistic 127.0058 Durbin-Watson stat 0.584562
Prob(F-statistic) 0.000000

Full Model: EXP = 2.03E+10 + 553165.5 X1 – 5.11E+08 X2 + 0.456051 X3 + 1937839. X4 +

2.38E+09 X5 + 1.41E+09 X6 - 2.73E+08X7 - 1.59E+09 X8 + u
R-squared (R2) is a statistical measure that represents the proportion of the variance for a dependent
variable that's explained by variables in a regression model.

R2 of the model is 0.1, then approximately 90% of the observed variation can be explained by the
model's inputs.

A P-Value of 0.00000, which is less than Significance level of 0.05, provides enough indication that the
model is significant.

Foreign reserves and inflation turned out to be very significant with P-values of 0.0000. Thus, adding
relevant significance to the model, and GDP per capita was also very significant with a P-value of
0.0001.

Moving on to Quality of port infrastructure and Manufacturing show P-values greater than 0.05
indicating weak significance between them and Exports of Goods and Services.

On the other hand interest rate also show a high significance with a P-value of 0.0150, while Exchange
rate show a value of 0.2963 and Employment to population show a value of 0.3123.

The Durbin-Watson stat shows a value of 0.584, which mean that there is a positive autocorrelation.
 Regression (2nd Model Exports of Goods)

Dependent Variable: EXPORTS_OF_GOODS

Method: Panel Least Squares
Date: 06/08/19 Time: 21:49
Sample: 2007 2017
Periods included: 11
Cross-sections included: 10
Total panel (balanced) observations: 110

Variable Coefficient Std. Error t-Statistic Prob.

C 4.85E+09 2.12E+10 0.228647 0.8196

EXCHANGE_RATE 227813.9 501220.0 0.454519 0.6504
EMPLOYMENT_TO_POPULATION -6.95E+08 4.79E+08 -1.451470 0.1497
FOREIGN_RESERVES 0.443848 0.017980 24.68530 0.0000
GDP_PER_CAPITA 2199277. 439873.1 4.999799 0.0000
INFLATION 2.38E+09 5.11E+08 4.648317 0.0000
INTEREST_RATE 1.63E+09 5.43E+08 3.009429 0.0033
MANUFACTURING -32738870 1.57E+09 -0.020860 0.9834
QUALITY_OF_PORT_INFRASTR 9.36E+08 5.27E+09 0.177462 0.8595

R-squared 0.915337 Mean dependent var 5.77E+10

Adjusted R-squared 0.908631 S.D. dependent var 8.37E+10
S.E. of regression 2.53E+10 Akaike info criterion 50.82520
Sum squared resid 6.47E+22 Schwarz criterion 51.04615
Log likelihood -2786.386 Hannan-Quinn criter. 50.91482
F-statistic 136.4955 Durbin-Watson stat 0.671890
Prob(F-statistic) 0.000000

Full Model: EXP = 4.85E+09 + 227813.9 X1 – 6.95E+08 X2 + 0.443848 X3 + 2199277. X4 +

2.38E+09 X5 + 1.63E+09 X6 – 32738870 X7 + 9.36E+08 X8 + u
R-squared (R2) is a statistical measure that represents the proportion of the variance for a dependent
variable that's explained by variables in a regression model.

R2 of the model is 0.09, then approximately 91% of the observed variation can be explained by the
model's inputs.

A P-Value of 0.00000, which is less than Significance level of 0.05, provides enough indication that the
model is significant.

Foreign reserves, inflation and GDP per capita turned out to be very significant with P-values of 0.0000.
Thus, adding relevant significance to the model.

Moving on to Quality of port infrastructure, manufacturing and exchange rate show P-values greater
than 0.05 indicating weak significance between them and Exports of Good.

On the other hand interest rate also shows a high significance with a P-value of 0.0033, and
Employment to population show a value of 0.1497.

The Durbin-Watson stat shows a value of 0.671, which mean that there is a positive autocorrelation.