Impact of the Effective Exchange Rate on the Trade Balance of Sri Lanka: Evidence from 2000 to 2013

Abstract

This study investigates the possibility of improving the trade balance of Sri Lanka through currency depreciation. An error correction model (ECM) was used to examine the short-term and long-term effects of nominal and real effective exchange rate depreciation on the trade balance. While there is no long-run relationship between the real effective exchange rate and the trade balance, depreciation of the nominal effective exchange rate will lead to a deterioration of the trade balance in the long term. Further, the results indicate that there is a positive causality relationship from the real GDP to the trade balance. Conversely, growth in money supply will worsen the country’s trade balance in the long run. The empirical findings show that currency devaluation is not an effective policy tool to improve the country’s trade balance.

JEL Classification: F 140

Keywords

Trade Balance Exchange Rate Sri Lanka

1. INTRODUCTION

Within the South Asian region, Sri Lanka was the first country to engage in a continuous and extensive economic liberalisation process, commencing from 1977. Since then, the Sri Lankan economy has gradually integrated with global markets through significant trade and financial market reforms. Over the past three decades, Sri Lankan governments have taken steps to replace the quantitative restrictions on imports with tariffs, to simplify the tariff structure and to attract foreign direct investment, particularly to support the country’s export sector. Currently, applied tariff rates in Sri Lanka are relatively low by developing country standards. In 1977, the country moved from a dual exchange rate regime to a unified exchange rate regime as part of its financial market reforms (Athukorala and Rajapatirana, 2000; Athukorala et al., 2011). These economic liberalisation policies aimed to boost the country’s economic growth through growth in the export sector.

As envisaged, the Sri Lankan economy opened up to a greater extent with economic reforms and its trade-to-GDP ratio rose to 54 per cent in 2013 (Central Bank of Sri Lanka, 2013). At present, the export sector relies largely on garments and few agricultural products, such as, tea, rubber, coconut and some minor agricultural commodities. During 2013, consumer goods, intermediate goods and investment goods accounted for 18, 59 and 23 per cent of total imports, respectively (Central Bank of Sri Lanka, 2013); fuel imports contributed 23 per cent to import costs. Due to the increasing demand for imports and escalating oil and commodity prices, Sri Lanka has been recording negative trade balances consecutively over the past several decades, despite the growth in overall trade. In 2013, its imports and exports relative to GDP stood at 32 and 23 per cent, respectively (Central Bank of Sri Lanka, 2013). Hence, Sri Lankan policymakers are faced with the challenge of identifying feasible policy solutions to improve the country’s trade balance.

The exchange rate is often used to influence trade flows and economic growth of a country. The Central Bank of Sri Lanka (CBSL) computes the nominal effective exchange rate (NEER)¹ and real effective exchange rate (REER)² to assess the value of the Sri Lankan rupee against a group of other currencies. The NEER and REER have been calculated using the exchange rates and price indices of 24 major trading partners of Sri Lanka, and have been defined such that an increase (or decrease) represents an appreciation (or depreciation) of the Sri Lankan rupee against the currencies of its major trading partners (Central Bank of Sri Lanka, 2007).

As Figure 1 indicates, the REER has gradually appreciated over the past decade, despite a depreciation in the NEER. Many economists and Sri Lankan exporters argue that a depreciation of the Sri Lankan rupee would restore the country’s global competitiveness (Bloomberg News, 2011; Lal and Lowinger, 2002). In this context, it is worthwhile investigating the long-run and short-run impact of the effective exchange rate depreciation on the trade balance of the country.

The remainder of the article is structured as follows. Section 2 discusses the theoretical framework and empirical evidence pertaining to the relationship between the trade balance, exchange rate, output and money supply. Section 3 explains the methodology while Section 4 describes the data. Section 5 discusses the empirical results and findings. Section 6 concludes and presents policy implications.

Figure 1

Source: Central Bank of Sri Lanka.

2. LITERATURE REVIEW

2.1 Theoretical Framework

2.1.1 Exchange Rate and the Trade Balance

According to the Mundell-Fleming-Dornbusch³ (MFD) model, real depreciation of a domestic currency boosts exports and encourages consumers to substitute imports by domestic goods. Hence, theoretically the real depreciation of domestic currency improves both the trade balance and output of an economy. According to the Marshall–Lerner (Marshall, 1923; Lerner, 1944) condition, devaluation can improve the trade balance in the long run, if the sum of import and export demand elasticities adds up to more than one. However, the trade balance should be nil or small for the Marshall–Lerner condition to hold.

Nevertheless, short-run effects of a currency devaluation may differ from the long-run effects. As Lal and Lowinger (2002) point out, the exchange rate affects the trade balance through prices and volumes. Currency depreciation will increase the prices of imports in domestic currency terms and reduce the prices of exports in foreign currency terms. An increase in import prices and decrease in export prices may eventually lead to a contraction in imports and an expansion in exports. However, the ‘price effect’ of currency depreciation on the trade balance may precede the ‘volume effect’. Hence, the trade balance may deteriorate in the short run in response to exchange rate depreciation but improve in the long run. This phenomenon is known as the ‘J-curve’ effect.

2.1.2 Real Output and the Trade Balance

During the 1950s, authors such as Harberger (1950) and Alexander (1952) introduced the absorption approach to explain the trade balance of a country. This approach assumes all variables are in real terms and disaggregates the economy’s expenditures into four components—consumption, investment, government expenditure and imports—the sum of which is defined as domestic absorption. An economy’s real income, in contrast, has four components, namely, consumption, investment, government expenditure and exports. In the event of an economic expansion, both real income and absorption may grow and the direction of the balance of payments and trade balance would depend on the relative growth rates of the two variables. If the real income of the economy grows at a faster pace than domestic absorption, the trade balance would improve and vice versa.

2.1.3 Money Supply and the Trade Balance

The monetarist approach suggests that the trade balance is essentially a monetary phenomenon (Hanh, 1961; Mundell, 1968; Pearce, 1961). In this approach, the behaviour of the balance of payments is examined from the point of view of the demand and supply of money. If money demand in the economy exceeds money supply by the Central Bank, the excess demand would be satisfied by foreign inflows leading to an improvement in the trade balance. However, if the Central Bank’s money supply is more than the money demand, the excess supply of money is eliminated through outflows, which leads to a worsening of the trade balance.

Also, the monetarist approach suggests that devaluation of the currency will lead to a reduction in real money balances in the short run, because increased import prices will be reflected in domestic prices. This will result in an increase in money demand relative to money supply. Hence, the balance of payments and trade balance would deteriorate in the short run. Subsequently, economic agents would reduce their consumption expenditure and save more to restore their real money balances and financial holdings. This leads to a trade surplus. Yet, the trade surplus will be eliminated as soon as the desired money balances are achieved (Ogbonna, 2010). Essentially, the monetary approach posits that changes in the nominal exchange rate have only a temporary effect on the trade balance and there is no long-run relationship between the real exchange rate and the trade balance.

2.2 Empirical Literature

Lal and Lowinger (2002) examine the short-run and long-run determinants of trade balances in five South Asian countries, including Sri Lanka. They use quarterly data from 1985 to 1998 and confirm the J-curve phenomenon for all South Asian countries under consideration. Further, they claim that the trade balance of Sri Lanka will improve after four quarters. Perera (2009) also assesses the impact of real depreciation of the Sri Lankan rupee on trade balance in the long run and short run, using the auto-regressive distributed lag (ARDL) model employing bilateral trade data between Sri Lanka and the country’s six major trading partners. However, Perera (2009) concludes that the trade balance between Sri Lanka and its major trading partners does not confirm the J-curve phenomenon and there is no specific pattern in response to depreciation in the currency. In addition, Arize (1994) claims that there is no long-term relationship between the trade balance and the real effective exchange rate in Sri Lanka and India, although positive relationships exist for other Asian countries.

Bahmani-Oskooee (1991) investigates the effect of real exchange rate depreciation on the trade balance of developing countries using the cointegration test method and concludes that the long-run devaluation of currencies improves the trade balance in most developing countries. However, Bahmani-Oskooee (1994) uses the error correction model to analyse the long-run relationship between the real effective exchange rate and trade balance in 19 developed countries and 22 developing countries (including Sri Lanka) and reports a contradictory view. Based on data from 1971Q2 to 1990Q4, only six of the countries analysed display a long-run relationship between the trade balance and the real effective exchange rate. He concludes that devaluation cannot have a significant impact on the trade balance for countries that do not have such a relationship, including Sri Lanka. Rose (1990) too performs a similar analysis with data from 30 developing countries, spanning from 1970 to 1988. His analysis indicates no significant relationship between the real exchange rate and the trade balance in most of these countries, as well as in Sri Lanka.

On the other hand, De Silva and Zhu (2004) examine the effect of currency devaluation on Sri Lanka’s GDP and trade balance using data for the period from 1977 to 1998. They conclude that devaluation of the Sri Lankan rupee has helped improve the trade balance, but has had contractionary effects on output.

Such contradicting empirical results on currency depreciation and the trade balance in Sri Lanka indicates the importance of further investigation of this issue using more recent data. This study investigates the long-run and short-run effects of the nominal and real effective exchange rate on the trade balance of Sri Lanka using data from 2000Q1 to 2013Q4. This time period is particularly significant since the country has adopted a managed floating exchange rate regime from 2001 onwards.

3. METHODOLOGY

Following the procedure used by Ogbonna (2010), the trade balance is specified as a function of the effective exchange rate, output and money supply.

_t(1)a₀a₁_ta₂_ta₃_ta₄_t_t

where, TB is the trade balance, GDP is real GDP, M2 is the M2 money supply and EX is the effective exchange rate. All variables were transformed to natural logarithms. Two models were separately analysed using the REER and NEER as the effective exchange rate. As the NEER (and REER) is the nominal (real) value of the Sri Lankan rupee in terms of foreign currencies (reported as an index where 2010 = 100), an increase in lnEX indicates an appreciation of the Sri Lankan rupee and a decrease in lnEX indicates a depreciation.

Following Lal and Lowinger (2002), the trade balance was defined as the ratio of imports to exports, to allow the log transformation and remove the sensitivity of the trade balance for units of measurement. Since the TB was defined as the import-to-exports ratio, an increase in lnTB indicates a deterioration of the trade balance.

To avoid any spurious regression in the estimation of Equation (1), it is crucial to identify the time series properties of the data sets employed. As suggested by Engle and Granger (1987), augmented Dickey-Fuller (ADF) tests were performed to identify the unit roots of the variables, in levels. If the variable had a unit root in levels, the ADF tests were used to check whether the series was stationary at first difference.

Cointegration tests were carried out if the variables were I(1) variables, using the Johansen cointegration procedure. The basic concept behind cointegration is that two or more variables may form a long-run relationship, although they may drift apart in the short run. Therefore, a regression containing the levels of such variables will have a stationary error term, even if each variable is non-stationary when taken individually. As per the Engle–Granger representation theorem, a dynamic error correction representation of data exists if there is a cointegration relationship between the variables. Hence, in the event of a cointegration relationship in the variables, the following ECM was estimated by incorporating an error correction term:

\begin{matrix} D_{} {lnTB}_{t} i_{0}_{i 1}^{p} i_{li} D_{} {lnGDP}_{t i}_{i 1}^{p} i_{2 i} D_{} {l n M 2}_{t i} \\ _{i 1}^{p} i_{3 i} D_{} {lnTB}_{t i}_{i 1}^{p} i_{4 i} D_{} {lnEX}_{t i} i_{5} {ECT}_{t 1} p_{t} \end{matrix}

(2)

where, D_lnTB, D_lnGDP, D_lnM2 and D_lnEX denote the first difference of lnTB, lnGDP, lnM2 and lnEX, respectively. ECT_t-1 is the one-lagged error correction term generated from the Johansen procedure. Further, disturbance terms are given by ξ_ts. The variance decomposition and impulse responses were obtained from the model (2).

4. DATA

The quarterly M2 money supply, NEER and REER data series were obtained from the Central Bank of Sri Lanka. Quarterly import and export data were obtained from the International Financial Statistics (IFS) database of the International Monetary Fund to calculate the import–export ratio (trade balance). The quarterly real GDP data series published by the Census and the Statistics Department of Sri Lanka were used as the real output variable.

5. RESULTS AND DISCUSSION

The ADF tests indicate that lnTB, lnGDP, lnM2 and lnEX (i.e., lnREER and lnNEER) are all non-stationary at levels and their first differences are stationary (see Appendix A for the ADF test results). Accordingly, all variables are I(1) variables. Further, the Johansen cointegration test was carried out for the NEER and REER models with three lags. The results of the cointegration tests are given in Appendix B. According to the Trace test, there is one cointegrating vector, significant at the 5 per cent probability level in both models.

5.1 Long-run Effects of the Exchange Rate, GDP and Money Supply on the Trade Balance

The results of the normalised cointegration equations are provided in Table 1.

Given the aim of this study, the cointegrating vector was normalised by lnTB by setting its coefficient equal to one, so it is easy to interpret the other coefficients as the respective long-run elasticities. For example, the coefficient for lnGDP represents the long-run elasticity of lnTB with respect to lnGDP. The coefficients for lnGDP are significant in both models and are less than −1. This indicates that growth in GDP by 1 per cent will enable Sri Lanka to improve its trade balance by more than 1 per cent. This finding is in line with the results of Lal and Lowinger (2002), who reveal a long-run positive relationship between the real GDP and trade balance. According to the absorption approach pertaining to the balance of payments, an increase in real income relative to absorption in the economy would improve the trade balance.

Table 1
Normalised Cointegrating Equation Parameter Estimates

lnTB lnGDP lnM2 lnEX Constant

NEER Model where, lnEX = lnNEER 1.000 −8.1333* 3.4847* 1.6215* 51.177

[1.8681] [0.8716] [0.6708]

(0.000) (0.000) (0.016)

REER Model where, lnEX = lnREER 1.0000 −3.5970* 1.3360* −0.3033 30.2437

[1.1864] [0.4345] [0.3572]

(0.002) (0.002) (0.396)

	lnTB	lnGDP	lnM2	lnEX	Constant
NEER Model where, lnEX = lnNEER	1.000	−8.1333*	3.4847*	1.6215*	51.177
		[1.8681]	[0.8716]	[0.6708]
		(0.000)	(0.000)	(0.016)
REER Model where, lnEX = lnREER	1.0000	−3.5970*	1.3360*	−0.3033	30.2437
		[1.1864]	[0.4345]	[0.3572]
		(0.002)	(0.002)	(0.396)

Source: Author’s estimates using Stata 12.1.

Notes: * indicate parameters significant at the 5 per cent probability level. Figures in [] and () are the standard errors and probabilities, respectively.

The coefficient of lnM2 is positive and significant in both the NEER and the REER models. Hence, growth in money supply would lead to a deterioration in the trade balance in the long run. An increase in money stock tends to increase the level of real money balances; economic agents may perceive this as an increase in their wealth and thus may increase their consumption and imports leading to a worsening of the trade balance (Ogbonna, 2010).

The coefficient of lnEX is positive and significant in the NEER model. This indicates a deterioration of the trade balance in the long run due to a depreciation in the nominal exchange rate. As Tennakoon (2010) has argued, Sri Lanka’s import demand for all product categories (i.e., consumer goods, intermediate goods and capital goods) is relatively price inelastic. Hence, import volumes will be less affected by the currency depreciation. Thus, import expenditure in terms of domestic currency will rise due to currency depreciation. Intermediate goods accounted for 59 per cent of the Sri Lanka’s imports in 2013 (Central Bank of Sri Lanka, 2013). Hence, escalated import costs due to currency depreciation will increase the cost of production in the export sector. As a result, the export sector may lose its cost competitiveness in global markets. This is particularly relevant for Sri Lanka’s largest export sector—textiles and garments, which is highly import dependant. The average import dependency of this industry was 45 per cent in 2013 (Central Bank of Sri Lanka, 2013). Therefore, currency depreciation may have a negative impact on Sri Lanka’s exports leading to a worsening of the trade balance.

However, the coefficient of lnEX in the REER model is not statistically significant at the 5 per cent level indicating there is no long-run relationship between the REER and the trade balance. This finding is in line with the monetarist approach and the findings of Bahmani-Oskooee (1994) and Rose (1990). Hence, Sri Lanka’s trade balance cannot be improved in the long run by the depreciation of the REER.

5.2 Short-run Effects of the Exchange Rate, GDP and Money Supply on the Trade Balance

The short-run effects of the exchange rate on the trade balance were also investigated using the ECM. The error correction models with NEER and REER are given in Table 2.

Table 2
Error Correction Model

NEER Model where lnEX = lnNEER REER Model where lnEX = lnREER

ECT_t–1 −0.3453 −0.4486

(0.000) (0.000)

D_lnTB(−1) −0.0912 −0.0904

(0.529) (0.540)

D_lnTB(−2) −0.0843 −0.0491

(0.450) (0.665)

D_lnGDP(−1) −0.7796 0.0582

(0.358) (0.930)

D_lnGDP(−2) −0.6793 −0.2198

(0.306) (0.723)

D_lnM2(−1) 3.6582* 2.6194

(0.045) (0.104)

D_lnM2(−2) 2.6913 3.4502*

(0.162) (0.044)

D_lnEX(−1) 1.4261* 1.1390*

(0.031) (0.035)

D_lnEX(−2) −0.3181 −0.2408

(0.634) (0.660)

Constant −0.0089 −0.0049

(0.899) (0.946)

R² 0.5739 0.5896

Chi² 57.9270 61.76513

P> Chi² 0.0000 0.0000

	NEER Model where lnEX = lnNEER	REER Model where lnEX = lnREER
ECT_t–1	−0.3453	−0.4486
	(0.000)	(0.000)
D_lnTB(−1)	−0.0912	−0.0904
	(0.529)	(0.540)
D_lnTB(−2)	−0.0843	−0.0491
	(0.450)	(0.665)
D_lnGDP(−1)	−0.7796	0.0582
	(0.358)	(0.930)
D_lnGDP(−2)	−0.6793	−0.2198
	(0.306)	(0.723)
D_lnM2(−1)	3.6582*	2.6194
	(0.045)	(0.104)
D_lnM2(−2)	2.6913	3.4502*
	(0.162)	(0.044)
D_lnEX(−1)	1.4261*	1.1390*
	(0.031)	(0.035)
D_lnEX(−2)	−0.3181	−0.2408
	(0.634)	(0.660)
Constant	−0.0089	−0.0049
	(0.899)	(0.946)
R²	0.5739	0.5896
Chi²	57.9270	61.76513
P> Chi²	0.0000	0.0000

Source: Author’s estimates using Stata 12.1.

Notes: *Parameters are significant at the 5 per cent level. Figures in parentheses are probabilities.

The coefficient terms of the ECM reflect the short-run relationship between lnTB and the other variables. The error correction term (ECT_t–1) represents the single-period response of the dependant variable to its deviation from equilibrium (Lal and Lowinger, 2002). Generally, the error correction term should be within 0 and −1 to ensure that the long-run equilibrium can be reached and the series is not explosive. Since the error correction terms of both models satisfy this condition, long-run equilibrium is achievable. If the trade balance deviates from its long-run equilibrium in the previous period, 35 per cent of that deviation will be corrected in the current period as per the NEER model, and 45 per cent of the error will be corrected under the REER model.

Figure 2

Source: Author’s estimates using Stata 12.1.

Most of the coefficients of the ECM are not significant except for the first lag of lnM2 in the NEER model and the first lag of lnEX in both the NEER and REER models. Therefore, impulse response functions were examined to identify the short-run dynamics of the trade balance and exchange rate. Figures 2 and 3 show the responses of the trade balance to the NEER and REER, respectively. The impulse responses show that the trade balance deteriorates in response to a depreciation in both the NEER and REER, but improves thereafter for a very short period.

Figure 3

Source: Author’s estimates using Stata 2.1.

The role of real output, money supply and exchange rate in explaining the volatility of the trade balance was also examined. Figure 4 shows the variance decomposition of lnTB with respect to lnM2, lnTB, lnGDP and lnNEER.

The variance decomposition of lnTB with respect to lnM2, lnGDP and lnREER is given in Figure 5 (from the REER model). Both models indicate that the exchange rate plays a minor role in explaining the variance in the trade balance, although the GDP and M2 are major sources of trade balance volatility.

Figure 4

Source: Author’s estimates using Stata 12.1.

Figure 5

Source: Author’s estimates using Stata 12.1.

6. CONCLUSION AND POLICY IMPLICATIONS

This article examines the empirical evidence on long-run and short-run interactions of the exchange rate (NEER and REER), trade balance, real GDP and money supply of the Sri Lankan economy, using an error correction model and quarterly data from 2000Q1 to 2013Q4. The study focuses on the NEER and REER based on the new 24-currency basket and a sample period which has not been covered by past literature. The findings show that the growth in real output tends to improve the trade balance and that the growth in money supply worsens the trade balance in the long run. The results of the study indicate a deterioration of the trade balance in the longer term due to nominal effective exchange rate depreciation, though there is no long-run relationship between the trade balance and the real effective exchange rate. Hence, it can be concluded that currency depreciation is not an effective policy measure to improve the trade balance of Sri Lanka. This is in line with the findings of Sinha (2010) that Sri Lanka’s trade balance is unlikely to improve by currency devaluation as the economy does not satisfy the Marshall–Lerner condition.

This has several implications for policymakers. First, Sri Lanka should focus on improving productivity of its export sector, and diversifying its exports from traditional agricultural commodities to value-added products. The government may implement vocational training programmes targeting the export sector and invest in research and development activities to enhance the productivity of export industries. In addition to product exports, Sri Lanka could also consider exporting services. The government could identify potentially exportable services (such as information technology) and encourage the expansion of a skilled labour force through public or private universities. If successfully implemented, such policies will be more effective in improving Sri Lanka’s trade balance than depreciation of its currency.

Footnotes

ADF Test

H_o: series has a unit root

H_a: series is stationary

APPENDIX B

Table B2

Johansen Test for Cointegration—REER Model (3 Lags, Unrestricted Constant, Linear Time Trend in Level Data)

Null Hypothesis	Alternative Hypothesis	Trace Statistic	5% Critical Value
r = 0	r ≥ 1	49.2937	47.21
r≤1	r = 2	23.1077*	29.68
r≤1	r = 2	8.6686	15.41
r≤1	r = 2	0.2706	3.76

Source: Author’s estimates using Stata 12.1.

Note: *fails to reject the null hypothesis at the 0.05 level.

Notes

References

Alexander

S. S.

(1952). Effects of a devaluation on a trade balance. International Monetary Fund Staff Papers, 2(2), 263–78.

Arize

A. C.

(1994). Cointegration test of a long-run relation between the real effective exchange rate and the trade balance. International Economic Journal, 8(3), 1–9.

Athukorala

Rajapatirana

(2000). Liberalization and industrial transformation: Lessons from the Sri Lankan experience. Economic Development and Cultural Change, 48(3), 543–72.

Athukorala

Bandara

J. S.

Kelegama

(2011). Trade liberalization and poverty in South Asia—Reforms, stylised facts and preview. In Athukorala

Bandara

J. S.

Kelegama

(Eds), Trade liberalization and poverty in South Asia (pp. 1–16). London & New York: Routledge, Taylor & Francis Group.

Bahmani-Oskooee

(1991). Is there a long-run relation between the trade balance and the real effective exchange rate of LDCs? Economics Letters, 36(4), 403–407.

Bahmani-Oskooee

(1994). Short-run verses long-run effects of devaluation: Error correction modeling and cointegration. Eastern Economic Journal, 20(4), 453–64.

Bloomberg News. (2011). IMF says Sri Lanka Rupee devaluation to spur overseas sales. Bloomberg. Retrieved from http://www.bloomberg.com/news/2011-11-23/imf-says-sri-lanka-devaluation-to-spur-exports-protect-reserves.html

Central Bank of Sri Lanka (2007). Exchange Rate, Central Bank of Sri Lanka Pamphlet Series, No. 3, Central Bank of Sri Lanka Website. Retrieved from http://www.cbsl.gov.lk/pics_n_docs/10_pub/_docs/pa/pamphlet/pl_3.pdf

Central Bank of Sri Lanka (2013). Annual report. Colombo, Sri Lanka: Central Bank of Sri Lanka.

10.

De Silva

Zhu

(2004). Sri Lanka’s experiment with devaluation: VAR and ECM analysis of the exchange rate effects on trade balance and GDP. The International Trade Journal, 18(4), 269–301.

11.

Dornbusch

(1980). Open economy macroeconomics. New York: Basic Books.

12.

Engle

R. F.

Granger

C. W. J.

(1987). Co-integration and error correction: Representation, estimation, and testing. Econometrica, 55(2), 251–76.

13.

Fleming

J. M.

(1962). Domestic financial policies under fixed and under floating exchange rates. Staff Papers—International Monetary Fund, 9(3), 369–79.

14.

Hanh

F. H.

(1959). The balance of payments in a monetary economy. Review of Economic Studies, 26(2), 110–25.

15.

Harberger

A. C.

(1950). Currency depreciation, income and the balance of trade. Journal of Political Economy, 58(1), 47–60.

16.

Lal

A. K.

Lowinger

T. C.

(2002). Nominal effective exchange rate and trade balance adjustment in South Asian countries. Journal of Asian Economics, 13(3), 371–83.

17.

Lerner

(1944). The economics of control. New York: Macmillan.

18.

Marshall

(1923). Money credit and commerce. London: Macmillan.

19.

Mundell

R. A.

(1968). International economics. New York: Macmillan.

20.

Ogbonna

B. C.

(2010). Trade balance effect of exchange rate devaluation in Benin Republic: The empirical evidence. International Journal of Social Science, 2(5), 138–51.

21.

Pearce

I. F.

(1961). The problem of the balance of payments. International Economic Review, 2(1), 1–28.

22.

Perera

W. T. K.

(2009). Bilateral J-curve between Sri Lanka and its major trading partners. Staff Studies—Central Bank of Sri Lanka, 39(1–2), 69–85.

23.

Rose

A. K.

(1990). Exchange rates and trade balance: Some evidence from developing countries. Economics Letters, 34(1), 271–75.

24.

Sinha

(2010). A note on trade elasticities in Asian countries. The International Trade Journal, 15(2), 221–37.

25.

Tennakoon

T. M. U.

(2010). Price and income elasticities of disaggregated import demand in Sri Lanka. Staff Studies—Central Bank of Sri Lanka, 40(1–2), 59–77.