The Effect Of Exchange Rate Volatility On Exports

Table of Contents

Abstract

Acknowledgements

Table of Contents

List of Tables

List of Figures

List of Acronyms

Chapter 1 : Introduction

Chapter 2 : Literature Review
2.1 Relationship between Exchange Rate Volatility and Trade
2.2 Effect of Exchange Rate Regimes
2.3 Exchange Rate Target Zones
2.4 Inflation Targeting and Exchange Rate Volatility

Chapter 3 : The Model and Data
3.1 Real Exports
3.2 Gross Domestic Product
3.3 Real Bilateral Exchange Rate
3.4 Volatility
3.5 Predictions of the model
3.6 Data sources

Chapter 4 : Tests
4.1 Unit Root Tests
4.2 Co-integration Tests

Chapter 5 : Empirical Results and Discussion
5.1 Summary Statistics
5.2 GARCH effects in the Real Exchange Rate
5.3 GARCH (1, 1) coefficients
5.4 Time series graphs
5.5 Unit Root Tests
5.6 Lag Order Selection
5.7 Co-integration Test
5.8 The VAR Model

Chapter 6 : Limitations of the study and Improvements

Chapter 7 : Conclusion

References

List of Tables

Table 5.1: Moment summary statistics of row data

Table 5.2: Correlogram test results for Autocorrelations (AC), Partial Autocorrelations (PAC) and Box-Pierce Q-statistics

Table 5.3: GARCH (1, 1) constant, ARCH and GARCH coefficients results

Table 5.4: ADF and Phillips-Perron test results for Unit Root

Table 5.5: Information Criteria results for optimal number of lags

Table 5.6: Trace statistic for Johansen-Juselius test result

Table 5.7: Maximum Eigenvalue statistic for Johansen-Juselius test result

Table 5.8: Lagrange Multiplier (LM) test for residual autocorrelation

Table 5.9: Jarque-Bera normality test results

Table 5.10: VAR model short-run results

Table 5.11: Granger causality Wald test results

Table 5.12: Forecast Error Variance Decomposition of Exports

List of Figures

Figure 2.1: Krugman Model

Figure 5.1: Squared Log difference of the Real Exchange Rate

Figure 5.2: Real Exports

Figure 5.3: Industrial Production

Figure 5.4: Real Exchange Rate

Figure 5.5: Real Exchange Rate Volatility

Figure 5.6: Eigenvalue VAR model stability test results

Figure 5.7: Impulse Response Functions

List of Acronyms

Abbildung in dieser Leseprobe nicht enthalten

Abstract

The purpose of this thesis was to examine the effect of real exchange rate volatility between the Canadian and US dollars on real exports from Canada to US. The study used quarterly data from 1960-2017.

The GARCH (1, 1) was used to model exchange rate volatility. After finding the variables were non-stationary with no co-integration, a VAR model was used to investigate the short-run relationship in the variables using Granger causality, impulse response functions and variance decomposition estimates.

The results revealed that the effect of exchange rate volatility is of mixed signs with coefficients that are not statistically significant.

Acknowledgements

Honour to God for the gift of life and His everlasting wisdom and favour that have brought me this far in life and honour to Him for all the great blessings yet to come my way.

My appreciation is extended to the Irish Government Department of Foreign Affairs through Irish Aid for funding my master’s degree studies.

I am deeply grateful to Professor Gregory Connor, my supervisor, for the valuable advice, assistance and guidance during the preparation of this thesis and the entire period of my master’s degree studies.

I am also grateful to all the staff of the Department of Economics, Finance and Accounting both academic and administrative for all the constructive advice and support during the course of my master’s degree studies.

Finally, my deepest gratitude to my parents, siblings and Yvonne for all the whole-hearted love and support that words cannot express. To you all, I am forever indebted.

Chapter 1 : Introduction

This thesis examined the effect of real exchange rate volatility on the real exports of Canada to the United States and found that exchange rate volatility does not have any statistically significant effect on Canadian exports. The thesis is divided into 7 chapters; chapter 2 gives an overview of important literature and contributions by researchers over the years specifically covering the relationship between exchange rate volatility and trade, exchange rate regimes, exchange rate target zones and inflation targeting. Chapter 3 presents the model and data used, definitions of the variables and the predictions of the model. Chapter 4 gives a theoretical and econometric overview of the unit root and co-integration tests. Chapter 5 gives the data output of the empirical results and discussions of test results. This output is presented using graphs and tables. Chapter 6 is a presentation of the limitations of the model and possible areas of improvement. Lastly, chapter 7 concludes and gives policy recommendations moving forward. Exchange rates are a key player in any economy that is engaging in international trade. A stable monetary policy system and financial sector play a key role in ensuring the exchange rate stability of the currency of a country. Firms and traders rely on prevailing exchange rates to forecast amounts to produce, import and export; thus are very much affected by the exchange rate volatility. In addition to this, there is a currency conversion cost in international trade. Traders use a number of products in financial markets to hedge against currency fluctuations; these include among others forwards contracts. This is especially true for short-term hedging than long-term hedging. Obstfeld (1995) states that long-term changes in exchange rates impact more on trade than short-term fluctuations. Doroodian (1999) concludes that hedging is both an imperfect and costly method of avoiding exchange rate risk.

A number of scholars have argued that exchange rate volatility has made trade expensive in North America due to hedging costs and this has reduced trade. It is on this basis that some scholars have put forward the proposition of a common currency¹ in North America, the Amero, an equivalent of the Euro in Europe. Scholars have also argued that a common currency or fixed exchange rate could lead to lower transaction costs associated to fluctuations. Bayoumi and Eichengreen (1994) state that about three quarters of both Canadian and Mexican exports go to the US and insofar as exchange rate uncertainty disrupts trade, currency links would be beneficial to Canada and Mexico. It is estimated that adopting a common currency in North America would eliminate transaction costs of up to $3 billion annually. Together, the three countries in this region signed an agreement, NAFTA, which created a trilateral trade bloc in the region. The implementation of NAFTA² inclined policies began on 01-January-1994. According to statistics from the Office of the United States Trade Representative, trade in goods and services between Canada and US totalled to an estimated value of $673.9 billion in 2017. This makes US a major trade partner and destination for Canadian goods. The trade volume between these two countries is very large; this is one of the justifications for this study. This study focussed on evaluating whether exchange rate volatility does indeed significantly reduce trade and thus justify the move towards a common currency in the region. The new evidence and findings will help policy makers make informed decisions moving towards further integration in the region. There is also limited literature on studies focusing on Canadian exports to the US; this thesis will help fill this gap.

Chapter 2 : Literature Review

This chapter presents key literature on the relationship between exchange rate volatility and trade, exchange rate regimes, exchange rate target zones and inflation targeting.

2.1 Relationship between Exchange Rate Volatility and Trade

McKenzie (1999) defined exchange rate volatility as the risk associated with unexpected movements in the exchange rate. Common belief among many is that exchange rate risk tends to reduce trade between countries due to the risk-averse nature of exporters; however, there is also no single measure of exchange rate risk. McKenzie (1999) further argues that there seems to be a general unresolved fundamental ambiguity in this relationship. In addition, exchange rate volatility may affect different markets in ways not the same and He argued against the use of aggregate trade data that has the potential to obscure any relationship. A positive relationship implies that exporters view exchange rate volatility as an opportunity to make profit if the exchange rate shifts in their favour and are thus considered risk lovers. Bahmani-Oskooee and Hegerty (2007) using annual export and import trade data between US and Mexico for 102 industries from 1962-2004, modelling the data using co-integration and error correction techniques found that exchange rate volatility might have positive, negative or even no significant effect on the volume of trade. Sercu and Uppal (2003) using a general equilibrium economy with stochastic endowments constructing a two-country, one-good, complete markets Lucas (1982) model with both trade and exchange rate volatility being endogenous argue that the relationship between exchange rate volatility and trade can be positive or negative depending on the source of increase in exchange rate volatility. The researchers use this model to improve on the weaknesses identified in previous work that are partial equilibrium and assume a linear relationship between trade and exchange rates, which may not be the case.

Furthermore, Bacchetta and Van Wincoop (2000) using a simple benchmark monetary model with separable preferences and only monetary shocks clearly state that the effect of exchange rate volatility on trade is rather mixed or uncertain depending on the countries, policies used, and the period³ of study just to mention but a few. Tenreyro (2007) using a panel data analysis of 104 countries and data covering the period 1970-1997 running country pairs with a gravity model found that exchange rate variability has no significant effect on trade. Aristotelous (2001) conducted a study on UK exports to US over the period 1889-1999 using a Bergstrand (1989) generalised gravity model and moving average standard deviation of real effective exchange rate to estimate volatility found that exchange rate volatility does not have any significant effect on UK-US exports. The researcher further concluded that volatility may have an effect on other variables like prices or foreign direct investment flows. Bailey et al. (1986) in their study of exports of the G-7⁴ OECD countries during the period 1973-1984 using the percentage change in the nominal effective exchange rate as a measure of exchange rate volatility indicate that exchange rate variability over the floating exchange rate period had no significant effect on exports. The result is found to hold whether volatility is assumed to have an immediate or lagged impact on exports. With this result, the researchers further argued that the criticisms of the flexible exchange rate regime as a deterrent to trade were unwarranted. Doroodian (1999) using a traditional export demand function and (G)ARCH to model volatility for three developing countries (India, South Korea and Malaysia) over the period 1973-1996 found that exchange rate uncertainty has a negative and significant effect on trade flows.

De Grauwe (1988) in his study that distinguished between the fixed and flexible exchange rate regime periods⁵ using bilateral trade data of ten major industrial countries stated that exchange rate variability has a statistically significant negative effect on volume of international trade. He went ahead to add that about 20% of the reduction in international trade among the industrial countries could be due to exchange rate variability. He further states that this negative relationship could be due to misaligned exchange rates. Hooper and Kohlhagen (1978) using a model of market equilibrium for traded goods that includes both import demand and export supply conducted a study on US-German trade flows between 1965-1975 concluded that an increase in exchange rate risk; that is an increase in volatility, will unambiguously force trade volumes to go down. Arize et al. (2008) examine export data of eight Latin American countries selecting the period 1973-2004 using co-integration and error correction techniques and find that exchange rate volatility has a statistically significant negative effect on export volumes in both the short and long-run periods. Dell’Ariccia (1999) using a gravity model found that exchange rate fluctuations had a significant negative effect on international trade in Western Europe over the 20-year period from 1975-1994. Sauer and Bohara (2001) study panel data representing developed and developing countries over the period 1973-1993; results showed that exports from LDCs are more adversely affected whereas most of those from the DCs are less affected by exchange rate volatilities. A further analysis showed that the most negatively affected nations are in Africa and Latin America while Asia is not affected to a greater extent. In a rather interesting result, a study by Giorgioni and Thompson (2002) examining US wheat exports over the period 1980-1996 using a panel data approach estimating models using OLSQ, fixed and random effects concluded that import volatility and not exchange rate volatility significantly reduces the volume of exports.

Franke (1991) argues that whether the expected export volume of a firm grows with exchange rate volatility depends on the optimal export volume being a function of the exchange rate and on optimal adjustment of entry and exit rates. He stated that expected volume of international trade grows with exchange rate volatility for a firm with comparative disadvantage. Doyle (2001) studies Irish exports to the UK from 1979-1992 on both aggregate and sector specific data because exchange rates tend to affect sectors differently, employing a methodology that used a first-order GARCH process, co-integration and error correction techniques. Her findings revealed that both nominal and real exchange rate volatility were important determinants of over 35% of Ireland-UK trade with positive statistically significant effects predominating. The researcher concluded in her study that the positive relationship might be because Irish firms operate in a small open economy with a few options in dealing with increased exchange rate risk and have to keep on trading for the fear of losing market share. A similar sector specific study by Awokuse and Yuan (2006) using a gravity model on panel data for 49 importing nations over two sub-periods found that US poultry exports are negatively affected by exchange rate risk, however, only statistically significant for a method that uses the variance of the spot exchange rate as a measurement of volatility. From this result, it was also concluded that the effect of exchange rate volatility on trade would also depend on the measure of volatility employed in the study. McKenzie and Brooks (1997) study Germany-US bilateral trade data from 1973-1992, observing Germany’s exports to and imports from the US and using ARCH models to estimate exchange rate volatility. The results showed that there is a statistically significant positive effect of volatility on trade. Olayungbo et al. (2011) conducted a study on 40 developing countries in sub-Saharan Africa over the period 1986-2005 using a gravity model of pooled OLS and fixed effects approach to panel GMM; the results revealed a statistically significant positive relationship between exchange rate volatility and aggregate trade. In addition, there was not much difference between the effect of volatility on trade of primary and manufactured products. The authors warn that this result must be interpreted with caution, as the history of exchange rate volatility is still relatively young in developing countries compared to developed nations. Calderon (2004) conducted a study on 79 countries covering the period 1974-2003 using a GMM-IV method for panel data. The findings revealed that real exchange rate fluctuations have a lower impact if the economy is more open to international trade and that trade openness helps attenuate the impact of highly volatile shocks to fundamentals on the volatility of real exchange rate fluctuations. Asteriou et al. (2016) in a four-country study covering the period 1995-2012 use export and import demand functions, GARCH to model volatility, ARDL bound test for long-run relationship and Granger causality to test the short-run relationship. In the short-run, a significant causal relationship from volatility to import and export was detected for Indonesia, Nigeria and Mexico; no causality relationship was detected for Turkey. In the long-run, however, exchange rate volatility had no effect on import and export demand except for Turkey and the magnitude of this effect was small. Lee (2003) using Canadian quarterly data over the period 1971-2000 found no evidence of
co-integration between exports and the explanatory variables including exchange rate volatility that was estimated using the moving average standard deviation (MASD). The researcher went ahead to use an ARDL model and found that exchange rate volatility has adverse short-run effects on exports. Fabiosa (2002) studied the effect of exchange rate volatility on Canadian meat exports to US and Japan over the period 1994-2001 using a standard supply function, an AR(p) model to represent the real exchange rate and a GARCH (p, q) to generate time-varying variance. The effect of exchange rate volatility on exports was found to be of negative sign with most of the volatility parameters not statistically significant. Obstfeld (1995) argues that simultaneity bias may affect cross sectional studies emphasizing that; if two countries have a high level of bilateral trade, relatively small bilateral real exchange rate adjustments suffice to offset asymmetrical shocks and thus the observed relationship between trade and volatility could reflect the effect of trade on volatility and not the effect of volatility on trade. The strategy of devaluing or depreciating a currency to promote exports and reduce imports in order to correct a trade balance (J-curve effect) may work for some countries and not work for others. China has always devalued its currency to increase its exports and this has worked. Rose (2000) states in his study that countries with extensive trade may lower exchange rate volatility deliberately in order to increase trade. Hsing (2008) studied US trade with a number of partner states in Latin America over a 30-year period and found the J-curve effect existed for Chile, Ecuador and Uruguay but did not exist for Argentina, Brazil, Colombia and Peru. Onafowora (2003) examined the short-run and long-run effects of real exchange rate fluctuations on real trade balance; the study covered the bilateral trade between three ASEAN countries⁶ and the US and Japan over the period 1980-2001. Using a VECM framework that treats variables as endogenous, impulse response functions were used to trace out the effects of exchange rate shocks on bilateral trade ratio. The results revealed evidence of co-integration and thus a long-run relationship among real trade balance, real exchange rate, real domestic income and real foreign income. From this study, it was concluded that the Marshall-Lerner condition holds in the long-run with varying degree of J-curve effects in the short-run.

2.2 Effect of Exchange Rate Regimes

The end of the Bretton Woods Gold Standard in 1971 dubbed the Nixon shock meant that countries could no longer redeem gold for their dollars. This gave rise to Fiat money whose value is derived from the confidence investors have in the economy or the monetary system of the country in question. Compared to gold, which is relatively stable in value, this meant that the value of a currency would be subject to substantial fluctuations over the years that followed. According to Frieden et al. (2006), the exchange rate is the most important price in the economy and has a ripple effect on all other prices. McKenzie (1999) in his paper pointed out that a debate developed after the collapse of the Bretton Woods pact; the laissez faire economists embraced this while others argued that risk-averse exporters would reduce their output when faced with exchange rate risk caused by a floating exchange rate. Von Hagen and Zhou (2005) found that the exchange rate regime employed by a country would depend on variables like level of development, inflation, foreign reserves, and financial market development just to mention but a few. The exchange rate regime employed by a country does have an impact on the volume of trade it has with its neighbours and the rest of the world. According to the IMF Annual Report on Exchange Arrangements and Exchange Restrictions (2016), the de facto classification of exchange regimes are as follows.

No Legal tender (Currency union); here the currency of another country is used as the sole legal tender or a country may belong to a monetary union in which a similar currency is used. A good example here could be the European Monetary Union that can be deemed a very successful union.

Currency board arrangements; here a Board is given the authority to fix the value of the domestic currency in terms of a foreign currency. Tsang (1999) adds that Currency Boards issue money with 100% foreign exchange reserves, somewhat similar to the gold standard.

Fixed peg arrangements; a country usually fixes its currency at ±1 percent in relation to another currency. The exchange rate is allowed to fluctuate with narrow margins of less than ±1.

Pegged with horizontal bands; the exchange rate is maintained at a value of more than ±1 around a fixed rate.

Crawling pegs; the value of a currency is changed periodically in small amounts at a predetermined fixed rate in response to certain economic indicators.

Crawling bands; exchange rate fluctuations are maintained within certain margins of at least ±1 around a central rate, these rates are also adjusted periodically depending on the economic situation.

Managed Float; here the monetary authority tries to influence the exchange rate without having a specific path or target.

Independent Float; the exchange rate in this case is determined within the market by forces of demand and supply. Interventions are only aimed at managing macroeconomic variables like inflation but not determining the price of the currency itself.

Generally, all the above deeper classifications can be broadly grouped into fixed and floating exchange rate regimes. Mussa (1986) states that real exchange rates show greater short-term variability under the floating exchange rate regimes than under fixed exchange rate regimes. However, it does not automatically follow that this greater variability is undesirable; the quick responsiveness to economic shocks could be a reason why countries adopt the floating regime. It was first shown by Rose (2000) that countries with a common currency trade more and countries will trade less with each other when their currency exchange rates are volatile. The study also draws a distinction between a currency union and a fixed exchange rate regime; the two have different effects. The findings revealed that countries in a currency union trade 3.35 times more with each other than they do with countries outside the union. He also adds that when Ireland left its long 1:1 parity with the British pound in 1979 to join the European Monetary System, its bilateral trade with the UK fell by fifty percent from 1980 to 1985, and had not attained even its 1975 level before the end of the sample in 1990. This decline occurred despite large increases in both real GDP and real GDP per capita that would ordinarily lead to a substantial increase in trade. Adam and Cobham (2007) confirm that a currency union as an exchange rate regime like the European Monetary Union does indeed increase the amount of trade between nations. Their findings also indicated that other regimes are significantly more pro-trade than the flexible exchange rate regimes that embed more uncertainty and transaction costs. Furthermore, Lopez-Cordova and Meissner (2003) use a gravity model to find out how currency unions and monetary regimes like the gold standard affected globalization in the 19th century. Their results indicated that the presence of a similar monetary regime and currency unions do have a significant positive effect on the volume of trade. From this perspective, it is evident that the fixed exchange rate regime does favour trade because some countries especially in the emerging market economies may not have access to forward markets in order to hedge against exchange rate volatility. Aristotelous (2001) argued that the exchange rate regime does not have an effect on export volumes. Bailey et al. (1986) in their study also concluded that the criticisms of the flexible exchange rate regime as a deterrent to trade were unwarranted. Bacchetta and Van Wincoop (2000) argued that adopting a fixed exchange rate system does not necessarily lead to more trade but rather the volume of trade will depend crucially on how the exchange rate system is implemented. Coes (1981) studied exchange rate uncertainty under the crawling peg regime for Brazil and findings revealed that the reduction of real exchange rate after the adoption of a crawling peg regime in 1968 had positive effects on exports. Kenen and Rodrik (1986)⁷ analysed the effect of short-term volatility in real exchange rates using data from industrial countries over the period 1975-1984; a floating rate period. The study deliberately excluded the fixed rate period in order to avoid specification bias arising from the change in the exchange rate regime. Findings revealed that volatility of real exchange rates had not diminished over time as countries gained experience with the floating exchange rate and more importantly, the volatility of real exchange rate depresses the volume of trade.

2.3 Exchange Rate Target Zones

Svensson (1992) defines exchange rate target zones as fixed exchange rate regimes with bands. According to Driffill (2008), a target zone attempts to limit the movement of an exchange rate avoiding the pitfalls of both pegged and a freely floating rate. Scholars have shown that the announcement of an exchange rate target zone has the advantage of reducing exchange rate volatility. Obstfeld (1995) states that exchange rate target zones are designed to give countries greater autonomy over policy while imposing a stabilizing effect on volatile exchange markets.

The Krugman Model in a Perfectly Credible Target Zone

Figure 2 . 1 : Krugman Model

Abbildung in dieser Leseprobe nicht enthalten

Source : Svensson 1992

Krugman (1991)⁸ developed a fundamental model that became a point of reference for a great deal of research in target zone modelling. His model proposes that the behaviour of the exchange rate within the band depends on aggregate fundamental and its expected rate of change. The model predicts the S-shape non-linear relationship between the exchange rate and its fundamental determinants as shown by the curve TT. The line FF represents the equilibrium exchange rate in the free-floating regime. He assumes the exchange rate depends linearly on macroeconomic fundamental and the expected future value of a currency. Within the fundamental, there are two components that is velocity and domestic money supply where velocity is exogenous and stochastic while the money supply is changed or altered by the central bank from time to time to control and manage the exchange rate. As long as the exchange rate lies within the band, the money supply remains unchanged. The stochastic process is assumed to follow a Brownian motion without drift.⁹ The main results from the Krugman model are the honeymoon effect and smooth pasting.¹⁰ As revealed in literature by Svensson (1992), if the exchange rate is higher and closer to the upper edge of the exchange rate band, the probability that it will reach the upper edge is higher. Thus, the probability that there will be future intervention to reduce money supply and strengthen the currency is higher. The target zone exchange rate is less than the free-float exchange rate for a certain level of the fundamental. He further adds that the slope of the target zone exchange rate function is zero at the edges of the band thus the exchange rate at this point is insensitive to changes in the fundamental; this is smooth pasting. Duarte et al. (2012) in their study emphasise that the honeymoon effect implies that a perfectly credible target zone has the stabilization effect; interventions by the monetary authorities to stabilize the exchange rate within the band make the exchange rate more stable than the underlying fundamental.

Time varying re-alignment risk occurs when the exchange rate band is allowed to shift over a period. Bertola and Svensson (1993) were the first to present an exchange rate target zone model with time varying re-alignment risk. The introduction of time varying re-alignment risk changes the process by which the interest rate differentials are determined and the interpretation of interest rate differentials against exchange rate plots. The interest rate differential is now equal to the sum of the expected rate of currency depreciation within the band and the expected rate of re-alignment. Hurley et al. (1993) in their study of the appropriate level of reserves required to defend an exchange rate target zone found that; for the case of Ireland, reserves were approximately optimal for most of the 1980s but significantly below optimal during 1989 and 1992. Furthermore, the authors concluded that foreign exchange reserves should at least be kept above 25% of domestic credit. Frenkel et al. (1989) argue that monetary policy coordination among the major industrial countries aimed at stabilizing the exchange rate would minimize the adverse effects of exchange rate variability and uncertainty on the volume of trade. This coordination would include among others establishing credible exchange rate target zones.

2.4 Inflation Targeting and Exchange Rate Volatility

The collapse of the Bretton Woods also led to a rise in inflation rates and the more industrialised countries moved from exchange rate targeting to inflation targeting. According to Bernanke and Mishkin (1997), inflation targeting refers to the announcement of official target ranges for inflation rates at one or more horizons and the inflation rate stability is the overall objective of the monetary policy. Bank of Canada was among the pioneer central banks to adopt inflation targeting policies along with the Reserve Bank of New Zealand and Bank of England in the 1990s. Developing countries are also adopting inflation targeting as a good monetary policy framework for economic and general price stability including foreign currency price stability. Inflation rates do have an effect on the exchange rate, high inflation rates tend to cause currency price volatility and stability in inflation is likely to cause exchange rate stability. Gali and Monacelli (2002) use a small open economy of the Calvo sticky price model to investigate monetary policy and exchange rate volatility. Findings show that inflation targeting which simultaneously achieves stabilization of domestic prices and output gap may pose a substantially larger volatility of nominal exchange rate and the Terms of Trade. Pontines (2013) using a treatment effect regression technique points out that nominal and real effective exchange rate volatility are lower in countries practising inflation targeting than in countries not practising this policy. The researcher also adds that developing countries that target inflation experience lower exchange rate volatility than non-inflation targeting developing nations. Cabral et al. (2016) investigate the relevance of exchange rate on the reaction function of central banks of 24 emerging market economies over the period 2000-2015 employing panel data fixed effects OLS and GMM. The findings showed that countries whose central banks do not practice inflation targeting experience more exchange rate volatility. Edwards (2006) in a study covering the period 1988-2005 using GARCH models to estimate volatility and data from 2 advanced and 5 emerging economies with a panel data analysis and country specific approach found that there was no evidence that adopting an inflation targeting policy increases nominal or real exchange rate volatility. In addition, the researcher’s findings also revealed that the adoption of a floating exchange rate regime increased the degree of exchange rate volatility in some of the countries. The researcher concluded that there is some evidence that some countries take into account explicit developments in the nominal exchange rate when conducting monetary policy. Aizenman et al. (2011) in their study conducted among emerging markets found that the countries are not actually practising pure inflation targeting. Central banks are employing a mixture of inflation targeting; responding to both inflation and exchange rate by setting interest rates. They further reveal that emerging economies are simultaneously targeting inflation and exchange rates. Bernanke and Mishkin (1997) state that exchange rate targets are intermediate while inflation targets are over longer-term horizons hence inflation targeting policy always takes precedence. Ebeke and Fouejieu (2015) state that the higher the net imports of a country, the stronger the pass-through of imported inflation pressure and thus inflation targeting countries with greater exchange rate pass-through (higher imported inflation) may be more inclined to controlling exchange rate fluctuations and will employ less flexible exchange rate regimes in the process. Mishkin (2004) points out that weak fiscal and monetary institutions make emerging market countries vulnerable to high inflation and currency crises. Even those that are trying to practise inflation targeting may not necessarily neglect the need to respond to exchange rate fluctuations as and when the need arises. The central banks may focus too much attention on limiting exchange rate fluctuations and at times revise their inflation targets as a result. In conclusion, the literature review revealed that there are mixed sign effects of exchange rate volatility on trade; at times even no effect. Furthermore, the literature looked at some monetary policy economic aspects that relate to exchange rate volatility. The exchange rate regime employed, the target zones and inflation targeting practises may all play a role in exchange rate volatility and thus may affect the volume of trade directly or indirectly. It is worth noting that the inflation targeting policy is still relatively young and has registered a lot of monetary policy success in a number of countries especially the developed ones. The policy has also become of great interest in the emerging market and developing economies mainly because of its success in developed economies.

¹ Many have argued that if a common currency is to be adopted in North America, it will be the US dollar. Canada and Mexico will have to give up the Canadian dollar and Mexican peso respectively. This is because the US dollar is an internationally recognised reserve currency and thus the US cannot give it up.

² NAFTA superseded the Canada-United States Free Trade Agreement and included Mexico thus the bloc comprises of the United States, Canada and Mexico. Canada and United States remain very important strategic trade partners to each other.

³ The period of study may be characterised by exchange rate regime shifts, economic booms and depressions, currency and financial crises and political instability all of which may affect the results of the study thus a need for sub-period analysis to control for these economic shifts.

⁴ Canada, France, Germany, Italy, Japan, UK and US.

⁵ Fixed regime: 1960-1969, transitional period: 1970-1972 and flexible regime: 1973-1984.

⁶ Thailand, Malaysia and Indonesia.

⁷ The study covered developed countries using three measures to model risk aversion; first is the standard deviation of the monthly percentage change in the real exchange rate. The second measure is the standard deviation of the real exchange rate obtained from a log-linear trend equation and finally the standard deviation of the real exchange rate obtained from a first-order auto-regressive equation. The quadratic measures of volatility are more consistent with the hypothesis of risk-averse behaviour as identified in their paper.

⁸ S(t)=f(t)+αE_t [ds(t)]/dt; Where s(t) is the log of the nominal exchange rate at time t, f(t) is the fundamental at time t, α is the absolute value of the semi-elasticity of the exchange rate with respect to its expected rate of change and E_t is the conditional expectations operator on the available information at time t.

⁹ Svensson (1992) stresses that for a Brownian motion without drift, the two things to note are; realised sample paths are continuous over time and do not include jumps; and changes in the variable over any fixed time interval are distributed as a normal random variable with a zero mean and a variance that is proportional to the time interval’s length. He further notes that this assumption is convenient because it means that the free-float exchange rate will also be a Brownian motion, which is consistent with observations that free-float exchange rates follow random walks.

¹⁰ A detailed interpretation of the honeymoon and smooth pasting effects in the Krugman model can be found in the paper; Lars E. O. Svensson, An Interpretation of Recent Research on Exchange Rate Target Zones, 1992.