Bahir Dar University: Impacts of Exchange Rate On Export Earning of Ethiopia Prepared by
Bahir Dar University: Impacts of Exchange Rate On Export Earning of Ethiopia Prepared by
Bahir Dar University: Impacts of Exchange Rate On Export Earning of Ethiopia Prepared by
May, 2004 EC
Bahir Dar,
Ethiopia
Executive summary
This study is conducted to give clear understanding on arguments between that exchange
rate variation has significant effect on the export earning of Ethiopia or not. The study has
discussed questions that whether the export sector of Ethiopia is promising or not, and does
exchange rate matter on the country’s export sector. To briefly investigate these questions,
the study has employed a forty years observation of sample periods to analyze the time
series data using OLS Econometrics estimation techniques. And the study justifies that
exchange rate variation has no significant effect on the country’s export sector. Moreover,
though the country’s competitiveness has decreased as the time goes, the export earning
has improved from time to time and possible to say the export of the country is promising
due to the contribution of other factors other than exchange rate variation.
Acknowledgment
First and for most, we would like to thank God for being with us
in completion of this study. We also express our heartfelt
gratitude to our adviser, instructor, Surafel Melak, Dean of
Business and Economics College at Bahir Dar University,
devoting his golden time for us from the beginning up to the
end.
Finally, we say thanks all who supported us by providing the
required data, and by giving us their constructive suggestions
and advice in doing this research. Our appreciation to all is
great!
Table of Contents
ACKNOWLEDGMENT ………………………………………………
ACRONYMS ………………………………………………………………
1. INTRODUCTION……………………………………………………………….
1.6 Methodology…………………………………………………………………………….
2. LITERATURE REVIEW……………………………………………………………………
3. METHODOLOGY………………………………………………………………………
3.1 Model Specification……………………………………………………………………….
3.2 Econometric Tests………………………………………………………………………
4. ANALYSIS OF DATA……………………………………………………………………
4.1 Descriptive analysis………………………………………………………………..
4.2 Empirical analysis…………………………………………………………………….
International trade does not carried out using a single currency just like what
is done in domestic trade, because international trade is conducted at least
between two nations. This mismatch in currency needs to consider price of
one’s currency in terms of other’s currencies what we call it the concept
exchange rate, which is used to compare the price of one country’s currency
with other’s currency. (Salvator, 1990).
Even though exchange rate is not the only means for the price at which
different countries currency units are exchanged, its valuation plays an
important role for the export of a nation. Assuming other things remaining
constant, the change in exchange rate has pro and cons on the export sector
and in turn, it affects the economic growth of a certain country. (Mankiw,
2004)
Since Ethiopia has become more open and one participant of international
trade, its currency has been expressed in terms of other nation’s currency and
the country’s export was affected by the valuation of exchange rate from time to
time. This in turn affects its economic growth as well.
Though there is a decline in net export from the experience of the past few
years due to high increment in import trade component, other things
remaining constant, Ethiopia’s export has shown an increase due to the
devaluation of its currency, it would be worse too much for its export and
further for its economic growth.(Befkadu and Berhanu, 1999/2000)
The year 1992 was a very important period in Ethiopia that has shown a
deliberate exchange rate variation to promote the country’s export sector. This
increase in exchange rate is still continuing. One dollar was exchanged by birr
2.07 before 1991 E.C. and again in 1983 E.C the transitional government
devalued the exchange rate of birr from 2.07 per dollar to birr 5 per dollar.
This value of birr has also decreased to birr 8.85 in 1993 further. Then
currently 17 birr is required to get one American’s dollar. (Befkadu&Berhanu,
1999/2000)
Previously, two different studies have been conducted about Ethiopian export
sector and trade balance of the country. (Yamrot: Jemal, 2005). They have used
macro-economic time series data obtained form NBE. Variables such as real
GDP, real exchange rate, real private consumption, real world GDP, foreign
exchange availability, tariff and real private sector credits are used in the
analysis of both the above studies. Based on the findings of these studies, real
GDP, real private consumption, real exchange rate and foreign exchange
availability can significantly affect the export sector of Ethiopia.
However, these studies are not in-line with the current situations due to the
reason that the studies cover the time from 1991 to 2003 as a sample for the
other reason after this period; there is a huge variation in exchange rate.
In addition to this, coffee, which served as a major export item for a long period
of time, has to be used as an important variable to affect export of Ethiopia.
So, the present paper differs from the previous studies by the following
important unique features.
Firstly, it covers a longer sample period from 1971-2010,
Secondly, it includes world price of coffee as one important variable and will
look its effect on the export sector of Ethiopia.
Thirdly, this paper makes itself unique in looking the effect of the wide
variation of exchange rate on the export of Ethiopia after the previous studies
have been conducted.
-To identify the role at which exchange rate plays in the economy,
-To look at the trends of export earnings of the country.
The paper will bound itself to the assessment of the change in exchange rate
impact only on the export sector of Ethiopian economy. Hopping 40 years time
coverage is sufficient enough to assess the impact of exchange rate
determination on export of Ethiopia, the study to be analyzed extends from
1971 to 2010.
1.6 Methodology
1.6.1 Type and source of data
For the purpose of analyzing this study, macroeconomic time series data will
be used. All the data will be obtained from different publications of annual
report of national bank of Ethiopia. This will include annual report of real GDP,
foreign exchange rate, the world price of coffee, domestic consumption and the
volume of export reported throughout the sample periods of the study.
1.6.2.1Model specification
For the investigation of this study, we use export as a dependent variable and
other five independent variables (GDP, REER, WPC, oppn and dcon) having the
following relationship in logarithm form i.e.
Here one an easily observe that the functional specification is logarithmic since
it is a better way of expressing the change in the probability of export due to a
percentage change instead of one unit change in each explanatory variable.
Real GDP: is a type of GDP which shows what would have happen
to expenditure on output if quantities had changed but price had
not. As domestic income increases, the export level of a certain
country rises.(Mankiw, 2004)
Real exchange rate
It is the relative price of goods and services of one country to other countries.
As the value of a certain country's currency depreciates, the price of goods and
services are relatively low and residents in that country are willing to buy the
domestic product, while the domestic producers are not beneficial since their
products are cheap domestically. As a result they prefer to export their
products abroad where there is a relatively higher price. The vice-versa holds
true for appreciation. Hence, exports are encouraged and imports are
discouraged while the domestic currency began to depreciate.(Mankiw, 2004)
Domestic consumption
It is the municipality use of domestic products by domestic consumers. When
domestic demand for domestic products rise, selling in local market becomes
profitable than selling abroad and such pull of domestic demand erodes the
availability of goods and services for the world market ( Jemal, 2005). This
shows that when domestic consumption increases, export tends to decline.
So that based on the above descriptions, the explanatory variables are expected
to have the following sign.
Table expected sign of explanatory variables
Explanatory variable Expected sign
Real GDP Positive (+)
Real exchange rate Positive (+)
World price of coffee Positive (+)
Openness Positive(+)
Domestic consumption Negative (-)
2 Literature reviews
According to this approach the given BOP of a certain country can improve if
exchange rate variation has change the domestic demand for import elasticity
and foreign demand for export of that nation. This implies that a certain
country has devalue its currency to export more or to increase foreign demand
for exportable goods and services and to decrease domestic demand for
imports , but this might be diverted from the expectation. It may not show any
change in response to price change .At this time, devaluation causes a
decrease in export earnings and an increase of import expense which worsen
the trade balance account and the vice- versa holds true for over valuation.
Moreover, exchange rate variation may not bring any change in trade balance
account when a given variation has bring equal demand for domestic import
and demand for foreign export, though it bring good export earning only in one
perspective.
If the sum of elasticity of demand for import and export of a certain nation in
absolute term is less than one, better to give up to use devaluation policy
because it would worsen the trade balance deficit or deteriorates export earning
of a nation. Moreover, devaluating countries can increase export earning if a
country is in a position to have a larger exportable surplus and export capacity
otherwise exchange rate variation could bring insignificant change or leads to
an adverse effect on export earnings.
Another model called IS-LM (Mankiw, 2004) stated that exchange rate variation
of a certain country has changes the price of domestic good relative to foreign
goods and in turn affect the net export of that country. This leads to the IS-
curve to shift up ward /inward/.
This implies that when a certain nation depreciates its currency, the price of
that currency will lower the price of goods and services relative to foreign
goods. This increases export earnings of a nation. This incremental in export
earning leads to high output, income and employment opportunity to external
sector. This makes the IS-curve to shift out ward.
A change in exchange rate affects LM-curve to shift, i.e. when exchange rate
rises, it has an impact on the domestic money supply to rise, and in turn shift
interest rate down. So that it leads to depreciation of domestic currency which
enables to rise the nations export (Mishkin, 2004). This implies that an
increase in exchange rate brings further depreciation of the domestic currency
to the motive of rising export. This could be applicable until the IS and LM
curves equate each other.
Has export earning a quick response to the change in exchange rate?
Or can the change in exchange rate quickly affect export earnings?
When exchange rate of a country is devaluated then the price of its export will
fall and the price of imports rises. Initially, one might expect little to happen to
the amount of exports and imports demanded as consumers take time to
change their preference from import goods to domestically produced goods. In
addition, foreign consumers will take time to adjust from domestic goods to
foreign export. If this was the case, the BOP might be expected to worsen as the
value of exports would decrease and the value of imports would increase.
After a while, the situation improves as deficit gets smaller and then moves to
surplus. In the longer time period, once consumer preferences have adjusted to
the changes in import and export prices, then the amount of export and import
will change.
Generally, the response of export earning or effect of exchange rate an export
earning is explained by “J” curve effect.
Surplus
0 time
Deficit
Does exchange rate variation have the same impact on all countries?
A Marshal learner (ML) condition stated that the supply of import and export is
infinitely elastic in export sector of a devaluating country and the elasticity of
demand for import and export should be greater than one in absolute term.
Countries which provide this condition are advisable to use devaluation policy
to affect its export earning via exchange rate. Countries which are governed by
this Marshall Learner condition have flexible supply of import and export
(Manuer, 1990). Flexible supply in a sense mean that imports and exports are
able to be changed easily when price changes. This fact is visible in countries
which broadly depend on industrial products for their export.
However, for countries which export agricultural products, the Marshall learner
condition is not applicable (Soni,1995), because those agricultural products
have inelastic supply curve. This is because of their nature of production
process involved in the occupation. Sowing, waiting and then harvesting or
collecting agricultural products require more or less rigid time schedule. This
makes world supply of agricultural products less sensitive to price changes.
In addition exchange rate variation can be a remedy for balancing the BOP if a
single country has take exchange rate policy measures among its trade
partners. Otherwise, the exchange rate variation (devaluation) has not any
effect in increasing export earning, and in turn, BOP balance. Plus this, even
though devaluation raise export earnings, countries which are highly indebted
are not advisable to devaluate its currency since devaluation make worsen its
foreign debt burden. (Manuer, 1995)
The aim of this section is to review empirical findings of the study on the
correlation between the real exchange rate and export sector.Yamrot (2005)
attempted to investigate the real effect of exchange rate on trade balance of
Ethiopia, using data source of NBE’s various annual report covering the
sample period starting from 1970 – 2001, with econometrics analysis of OLS
estimators.
The study has used variables like real GDP, tariff, foreign exchange rate
availability and real exchange rate as an explanatory variable to explain the
trade balance of the country. The study gave more emphasis to exchange rate
expecting the rise in exchange rate could increase exportability and decrease
its importable goods and services. Finally, Yamrot has found that real
exchange rate positively affects the trade balance, implying that when real
exchange rate rises, trade balance also increases by decreasing price of export
and it makes goods and services cheaper in foreign market.
The study which was carried out by Tadese (2005) tries to investigate the effect
of exchange rate on trade balance of Ethiopia before and after the reforms. He
has used nominal exchange rate, foreign export price and openness as
explanatory variable to explain trade balance of Ethiopia.
The finding of this paper is that nominal exchange rate is estimated to be 0.03
providing that exchange rate devaluation is slightly sensitive to the change in
export of a county.
The study conducted by Rodrick (2009) for the World Bank report also shown
that real under devaluation has positive effect on economic growth and export
expansion, but this effect is significant only for developing countries with low
per capita income like Ethiopia.
Generally, based on the above empirical evidences, real exchange rate has been
found positively significant in a research conducted by Yamrot (2005), Jemal
(2005) and Rodrick (2009). But, unfortunately, it is found to be slightly
significant to the change in export of Ethiopia based on the findings of Tadese
(2005) and Wondimunegn (nd). They showed a disparity in their findings.
In sum, the findings of those researchers have shown that reer (real effective
exchange rate) has positive impact on export sector coinciding to the
conventional theory of exchange rate. However, based on Marshal Learner
condition these studies imply that the country has elastic supply of import and
export. It is also meant that the country is exporting more industrial products
than agricultural products. However, these all issues are beyond the fact
observed in the soil of developing countries like Ethiopia.
And for those findings indicating real exchange rate has negative impact does
not fit with conventional theory of exchange rate. There is an implication that
the variation in exchange rate (devaluation) brings nothing to export earnings.
Also they have null impact in influencing the nation's income. It is this idea
which conflicts with the IS-curve.
In addition to this, all the above empirically reviewed literatures are conducted
a few years back from now, hence not recent. So, what will be the findings of
this recent particular paper?
Chapter three
3 Methodology
In this study only time series, quantitative macro economic data are used.
These secondary data are its collected from the National Bank of Ethiopia’s
various annual report, from MOFED (Ministry of finance and economic
development), and from Ethiopian Economic Association bulletins. A forty year
observations a sample period sampling from 1990-2010 including the two
interior years has been used for the analysis of one explained and five other
explanatory variables.
The study used economic analysis of OLS estimators using stata- 11 software
to regress the explained variable over the explanatory variables.
This is because the relative price of goods and services are relatively high in
abroad. Thus, exports are encouraged and imports are discouraged and the
vice versa holds true for the case of appreciation or over valuation of domestic
currency.
We need to note from the above model that the very basic reason for
expressing the relationship between dependant and in dependant variables
in the form of logarithmic is simply to see the change in the elasticity. It is
to express what a unit change in the explanatory variables would bring
about an average change in the explained variable would bring about an
average change in the explained variable.
3.2 Econometrics tests
In order to assure the validity of the OLS estimation, checking the validity or
nature of time series data is very important. For this purpose stationary test
conducted. For valid estimations of OLS, the time series data itself has to be
stationary; otherwise, using OLS estimation technique with a non stationary
time series data would be senseless. Thus, to make sure that stationeries
condition has been fulfilled unit rout test is under taken.
Moreover, the study has employed diagnostic tests like multicollinearity test,
hetroskedasticity test and autocorrelation test to show the validity of the model
that the study used.
Chapter four
4 Analysis of data
4.1 Descriptive Analysis
Real effective exchange rate trends of Ethiopia
Exchange rate is not a recently introduced phenomenon in Ethiopia. It was about sixty years ago
that has been started while birr was used. Looking the trends of the exchange rate in pre and post
1991 is very crucial for better understanding of exchange rate trend in Ethiopia by using graph.
reer
400
350
300
250
Figure 2: real effective exchange rate trend
200
The pre 1991 real effective exchange rate
As it is observed from the above graph, the trends of real effective exchange rate starting from
1970 to 1990, the country’s competitiveness has improved; it range between 200-354. But, by
150
that time no devaluation has been taken rather overvaluation taken after Breton wood has been
started in 1971. Moreover, the competitiveness of the country has reached peak on that period
due100
to the reason that Derg has been under taken policy improvement from socialist to mixed
economy.
As it can be seen clearly, Ethiopian export is dominated by agricultural products. Among this,
the most dominant merchandize export item is coffee. Its share to the total merchandize export
varies from time to time. For example, since 1998/99, there was the persistent decline of the
share of coffee in the total merchandize export though it was the major source of visible export
earnings of the country. In addition, its share to the total volume also declined through time, and
even there was a period that has been registered negative growth rate. This is due to other export
items such as hide and skins, oil seeds, pulse, flower, and other items increased their share in the
total volume and value of the national export.
Among these export items, flower and oil seeds have shown remarkable change in percentage
share of export.
Gold 64.7 6.5 97.8 8.2 78.8 5.4 97.8 6.8 281.4 14.0
Flower 21.8 2.2 97.0 5.4 111.8 7.6 130.7 9.0 170.2 8.5
Others 87.8 8.8 63.6 7.7 106.3 7.2 91.3 6.3 112.5 5.6
Total 1000.3 100 91.8 1,4657 100.0 1,447. 100. 2,003. 100.
9 0 1 0
Table1፡ values of major export items source፡ Ethiopian revenue and
customs authority
From the above discussion, one can easily understood that Ethiopia’s export item is diversified
from time to time. Its export earnings increased through time even though its growth rate has
been fluctuated. This implies that an export earning of Ethiopia is still promising.
.
exp
30000
25000
20000
Moreover, as we can easily observe from the above graph, after 1991, the export earning has
15000
doubled itself, as the time goes on. And around the years 2003 the export earning was not
rewarding because of the political instability of a country related to election, but after that, it
shows an improvement in the export earning of the country.
10000
The openness of the country has also shown improvement, especially after the new reform
(1991) as the data testified. But, it has been fallen down in 2004, due to political instability in the
country. And after that it has shown great improvement up to now. This enables the country to
freely trade with other countries and can increase the export earnings. It has almost the same
trend with export earnings. (See the graph below)
oppn
0.6
0.5
0.4
0.3
Figure 4: openness of the economy
When we observe the trends of world price of coffee (the amount of money gained from export),
it has shown a better improvement though the rate of growth is not similar. (See Appendix 4)
As it is clearly observed from the data, domestic consumption was not stable throughout the
sample periods. It has reached on peak in 1999, the time when export also has fall down due to
the prevalence of drought and Ethio-Eriterean war. (See the graph below)
dcon
100
98
96
94
92
90 dcon
88
86
84
82
0 10 20 30 40 50
The study has been checked several tests like stationery, multicollinearity, autocorrelation and
hetroskedasticity. All the variables are stationed, there is weak relationship between the variables
and can be tolerated, no autocorrelation and hetroskedasticity problem has been solved by
robusting the variables so that the model becomes free from some defects. (See appendix 1, 2, 3,)
for autocorrelation, muticollinarity and stationary tests respectively.
An important question raised at the beginning of this study was whether the change in exchange
rate really affects the export earning of the country or not. Different empirical results have shown
different views. Some empirical finding indicated that the variation of exchange rate has no
effect on export earnings of the country. And some other results have found that the variation of
exchange rate has an effect. The finding of this paper looks like the following.
Variable Coefficients t-valve p-value
Cons 0.078 1.60 0.120
dlnwpc 0.233 2.98 1.005
dlnrgdp 0.680 2.14 0.040
dlndcon -3.749 -2.74 0.010
dlnoppn 0.710 1.49 0.147
dlnreer 0.122 0.13 0.899
R2 =0.5320 f(5,33)=14.88(0.0000)
Table 2: OLS estimation result
As we observe from the above table, all the explanatory variables of the model can explain the
explained variable by 53%. All the variables except openness and real effective exchange rate
are statically significant. As world price of coffee increased by 1%, the export earnings would
increase by 23% on average. This is due to the reason that coffee plays a dominant role in total
receipt of Ethiopia. Real GDP is also significant and as it is increased by1%, the export earnings
would rise by 6% on average. This is because as income of residents (producers) increased, the
capability of the producers to produce exportable commodities also rises and can produce as well
as export to foreign markets.
Domestic consumption also fits with the hypothesis proposed in the proposal. It negatively
affects export earnings. In contrary, this study proved that real effective exchange rate is
statically insignificant. This result coincides with low income countries such as Malawi, Ghana,
Zambia and Nigeria as stated in the empirical finding and also coincided with empirical study
conducted by wondimunegn (nd) and Tadese (2005).
The very basic reason for this may be because of the ineffectiveness of the Marshal Learner
condition in Ethiopia and perhaps because of the commitment of other partner countries to
devalue their own currency while Ethiopia devaluates its own.
Chapter Five
5 Conclusion and recommendation
5.1 Conclusion
Ethiopia has shown that a visible spiracle of development among the most banana republic
nations. This is specially happening after the previous decades. In the sense of development it is
very important to upgrade all of the economic sectors of the nation. Export is one of the very
important economic sectors of Ethiopia which is predominantly lead by supply of agricultural
products.
The findings of this study have proved that around 53% of Ethiopia’s export sector is influenced
by economic variables of real GDP of the country, its world price of coffee, the real exchange
rate, the domestic consumption and its openness.
Except openness and real effective exchange rate, all the above macro economics variables can
significantly affect export sector of Ethiopia. This implies that the supply side matters more than
the demand side. So, the real effective exchange rate is ineffective to influence export earning of
Ethiopia.
5.2 Recommendation
-This study recommends that the government or other policy designers should not stick with the
exchange rate variation for the motive of influencing export earnings of Ethiopia. To be effective
for this, the country has to transform its export items from agricultural to that of the industrial
products to go in line with the Marshal Learner condition.
-Exchange rate policy has employed to have large impact in overall economy, if the country’s
economy is developed. But here in Ethiopia, it has no effect since the country’s economy is not
developed. So it is advisable to critically look devaluation policy. Rather, the government has to
rapidly boost the growth of the country as it is observed in the previous years.
-The Ethiopian government has changed exchange rate variation for the last two decades to
increase the competiveness of the nation in the international market. But this study advises the
responsible body to revise again this exchange rate policy, since further devaluation would
worsen or exacerbates the existing inflation so that its impact would hurt domestic residents due
to high price of goods and services.
-Ethiopian producers [peasants] are less responsive to exchange rate policy changes and its
impacts. So that, producers are almost inelastic for the change in exchange rate. Plus this, their
land hold is constant and their product per annum is probably the same whatever the exchange
rate moves ups and downs for influencing them. So that devaluation policy becomes nonsense.
For the betterment of this, the government has to work hard in improving the technology and in
providing inputs for the agricultural sector to raise its productivity and via this to boost up export
earnings of the country.
Hence one must note that the export sector of Ethiopia is most influenced by the supply side
factors or variables like real GDP, domestic consumption and world price of coffee but not
mostly affected by the demand side variables like exchange rate policy.
APENDEX
F(3,30)=0.60
Prob > F =0.6200
dlnexp -6.744
dlnwpc -11.557
dlnrgdp -5.426
dlncon -7.739
dlnoppn -5.995
dlnreer -7.891
Mankiw,2004:”Macro Economics”,5thed,Worthpublisher,India,