Energy Economics 25 (2003) 741–765
0140-9883/03/$- see front matter
2003 Elsevier Science B.V. All rights reserved.
PII: S0140-9883Ž03.00044-6
Exchange rate of the US dollar and the J curve:
the case of oil exporting countries
Ayoub Yousefi , Tony S. Wirjanto * a b,
Department of Economics, Business and Mathematics, King’s College, University of Western Ontario, a
London, Ontario N6A 2M3, Canada
Department of Economics, Department of Statistics and Actuarial Science, University of Waterloo, b
200 University Avenue West, Waterloo, ON, Canada N2L 3G1
Received 3 July 2002; received in revised form 14 November 2002; accepted 2 December 2002
Abstract
This study examines the effects of changes in the exchange rate of the US dollar on the
trade balances of three oil-exporting countries, namely Iran, Venezuela and Saudi Arabia.
An exchange rate pass-through model is applied to allow changes in the exchange rate of
the dollar to affect prices of traded goods. Then, the impact of changes in prices on the
quantities of imports and exports of these economies is estimated. The results suggest a
partial exchange rate pass-through to these countries’ import and export prices in terms of
the US dollar. While the three countries raise the price of their primary export (namely crude
oil) in response to a depreciation of the dollar, Saudi Arabia’s long-run pricing strategy in
securing a larger market share stands in contrast to that of the two other OPEC members.
The sum of the estimated long-run price elasticities of demand for imports and exports is
found to exceed unity for Iran and Venezuela, but less than unity for Saudi Arabia.

2003 Elsevier Science B.V. All rights reserved.
JEL classifications: F31; F32; F14
Keywords: Trade balance; J-curve; Invoicing currency; Exchange rate pass-through; Crude oil
1. Introduction
The adjustment pattern of a nation’s trade balance in response to changes in its
exchange rate has been subjected to much discussion in the international trade
*Corresponding author. Tel.: q1-519-888-4567x5210; fax: q1-519-725-0530.
E-mail address: twirjant@uwaterloo.ca (T.S. Wirjanto).742 A. Yousefi, T.S. Wirjanto / Energy Economics 25 (2003) 741–765
literature. The dramatic swings in the exchange rate of the US dollar and the
sluggish response of the US trade balance during the 1980s sparked a new wave of
interest in this area.1
According to the textbook view, depreciation of the deficit nation’s currency visa-
vis a trade-partner’s currency raises the cost of import contracted in foreign
currency, while export revenue contracted in domestic currency remains unchanged.
The two effects together lead to a deterioration of the trade balance immediately
following a depreciation of domestic currency. In the short run, import and export
prices respond with little or no decline in volume. Assuming a zero initial trade
balance and dominance of the exporter currency in invoicing trade contracts, the
trade balance continues to deteriorate in the medium term. Over time, the relativeprice-
induced volume effect comes to dominate the price effect and the trade
balance improves. This pattern of the trade balance adjustment is commonly referred
to as the ‘J-curve effect’. The three time spans named by Magee (1973) as the
‘currency contract’, ‘pass-through’ and ‘quantity response’ periods, respectively. The
J-curve effect presupposes the Marshall-Lerner condition, which states that price
elasticities of demand for imports and exports sum to greater than unity. In addition
to the elasticity condition, another equally crucial prerequisite for the J-curve effect
is the invoicing currency of imports and exports. For instance, if an exporter or
importer uses a foreign currency in his foreign trade, his revenue or costs in terms
of domestic currency will be immediately affected when the exchange rate changes.
On the contrary, if he uses his own currency, a change in the exchange rate will
leave his revenue or costs unaffected immediately following currency depreciation.
The theoretical underpinning of the choice of invoicing currency was pioneered
by Magee (1973, 1974) and subsequently adopted in the studies by Mckinnon
(1979), Magee and Rao (1980) and Melvin and Sultan (1990), etc. Despite its
pivotal role as a catalyst in the transmission of changes in the exchange rate to the
trade balance, invoicing currency has been subjected to only few empirical
investigations. Information on the invoicing currency has also been very limited. It 2
was not until early in the 1970s that most European country began to collect data
on the invoicing currency. Tables 1 and 2 provide some noteworthy information on
the invoicing currency of international trade. First, for almost all of the countries
listed in Table 1, the share of own currency used in invoicing exports is greater
than the corresponding figure for invoicing imports. Second, the currencies of large
countries are used more than those of small countries in invoicing international
trade. For instance, more than half of the world trade, i.e. 54.8% is invoiced in
During the 1985:I–1988:II period, approximately 3 years, the trade-weighted average value of the 1
dollar measured against the currencies of G-10 countries declined by over 42%. Despite this precipitous
decline in the dollar’s value, the US nominal trade deficit continued to rise from $23.38 billions in
1985:I to $42.47 billions in 1987:III. Numerous empirical studies were launched to examine the causes
of the sluggish response of the US nominal trade deficit to the continuous depreciation of the dollar
since the first quarter of 1985. See for example Deyak et al. (1990), Hooper and Mann (1:1989), Koch
and Rosenswieg (1988), Krugman and Baldwin (1987), Mann (1986), Meade (1988) and Moffett
(1989).
For empirical investigations on the invoicing currency of international trade for the UK, Italy and 2
the European Union, see Carse et al. (1980) and Basevi et al. (1987), respectively.743 A. Yousefi, T.S. Wirjanto / Energy Economics 25 (2003) 741–765
Table 1
Currencies used in world trade (in percentage)
Country Year Share of each country’s Share of each country’s
(Region) exports in imports in
Dollars DM Sterling Own Dollars DM Sterling Own
(1) (2) (3) (4) (5) (6) (7) (8)
Canada Est 85.0 0.2 1.0 95.0 1.0 2.0
France 1979 11.6 10.2 3.2 62.4 28.7 14.0 3.8 35.8
Germany 1980 7.2 82.3 1.5 82.3 33.1 43.0 3.1 42.8
UK 1979 17.0 3.0 76.0 76 29.0 9.0 38.0 38.0
US Est 98.0 1.0 1.0 98 85.0 4.1 1.5 85.0
Japan 1980 61.5 1.9 0.9 32.7 93.0 2.0 2.0 2.0
Oil exporters Est 100.0 0 0 50.0 10.0 8.0 4.0
Other developing Est 85.0 0 15.0 72.0 7.0 4.0
Non-OECD Est 85.0 2.0 7.0 52.0 14.0 9.0
Developed world 54.8 14.4 7.5 54.3 14.0 6.9
(1979 weights)
Est indicates an estimated value.
Source: Page (1981).
Table 2
International usage of currencies (in percentage and using 1979 weights)
Country’s share in Currency’s share Cumulative share
world exports in exports of currencies
Dollar 11.7 54.8 54.8
Deutsche Mark 11.1 14.4 69.2
Pound Sterling 5.9 7.5 76.7
French Franc 6.3 6.4(a) 83.1
Netherlands Gui 4.1 3.0(a) 86.1
Belgian Franc 3.6 2.6(a) 88.7
Japanese Yen 6.6 2.3(a) 91.0
Swiss Franc 1.7 2.1(a) 93.1
Italian Lira 4.7 1.9(a) 95.0
Swedish Krona 1.8 1.7(a) 96.7
Schilling 1.0 0.8(a) 97.5
Danish Krone 0.9 0.8(a) 98.3
Irish Pound 0.5 0.3(a) 98.6
Finnish Markka 0.7 0.0(a) 98.6
(a) indicates that the share is used in own trade.
Source: Page (1981).
terms of the US dollar as shown in column 3 of Table 2. Finally, a sizable 85% of
exports from developing countries, traditionally primary commodities, are invoiced
in terms of the US dollar. Given the inconvertibility of the currencies of almost all
developing countries, the use of the third-country currency on imports has also been
predominant. Thus, the invoicing currency of trade makes the primary commodity
exporting LDCs vulnerable to the changes in the exchange rate of the US dollar744 A. Yousefi, T.S. Wirjanto / Energy Economics 25 (2003) 741–765
against other major currencies. Indeed, depending on the particular pattern of trade
and the invoicing currencies, each developing country’s trade balance can be
affected differently by the changes in the exchange rate of the US dollar.
Among primary commodity groups, crude oil sets a prime example for the use
of a third-country currency which is 100% US dollar. This figure is even larger than
that for the exports from the US, which is 98% as shown in Table 1. As for the
share of primary commodity export, the export of crude oil by OPEC members
made up more than 90% of their total exports over the last two decades. The
particular feature of the invoicing currency of trade of these countries makes them
a suitable representative of the primary commodity exporting countries, whose trade
balance responses to changes in the exchange rate of the US dollar merit, an
investigation similar to those carried out for the US economy. To fill this gap in 3
the literature, the present study examines the impact on the trade balances of the oil
exporting countries of the changes in the exchange rate of the US dollar. Faced
with the lack of data of interest, a small group of countries, hereafter, the ‘group
countries’ is chosen. The group countries consist of Iran, Saudi Arabia and
Venezuela, three major oil exporting countries out of the five founding members of
OPEC, with two of them from the Middle East and one from the western hemisphere.
The remaining part of the article is organized as follows. Section 2 develops a
model for the exchange rate pass-through and quantity response of imports and
exports. Section 3 describes the data and explains the choice of the exchange rates
used in this study. Section 4 provides the methodology for empirical testing and
reports the pretesting results. Section 5 presents a summary of empirical results and
Section 6 offer tentative conclusions and policy implications.
2. The model
In dealing with the trade flow analysis, it has been common to ignore the supply
side of imports by making an assumption that the price elasticity of supply is
infinite. In other words, import prices are considered to be exogenous to any given
country. This assumption is adopted, not because a nation’s export supply elasticity
is thought to be infinite, but because of the fact that it is one of many suppliers.
Several arguments have been made to support the exogeneity assumption. For
instance, Murray and Ginman (1976) make the argument based on the premise that
firms typically operate at less than full-employment capacity, implying that the
industry supply curve will be horizontal before the full-capacity production level is
reached. Also, Warner and Kreinin (1983) employ import prices in terms of foreign
currency in order to apply the method of ordinary least squares to estimate a singleequation
demand model.
There are only a handful of studies on the exchange rate pass-through for LDCs, although there are 3
a few studies, which examine the impact on the trade balances of changes in the exchange rates of the
domestic currencies of these countries. See e.g. Bahmani-Oskooee (1986, 1984). A common drawback
of these studies is their failure to recognize that many developing countries’currencies are inconvertible
that makes it impossible for these currencies to be chosen as the invoicing currency. As a result, the
third-country currency features the invoicing currency of trade of these economies.continue