Economics and finance fundamentals and applications

Exchange Rate Volatility and International Trade

The foreign exchange rate market offers investors a chance to make a considerably larger return on their investment than any other market in the world. However, along with these potential gains comes a considerable risk as well. Foreign exchange rates are extremely volatile and dependent on many variables. Understanding the factors that influence foreign exchange rates can mean the difference between profit and loss for an investor.

Financial markets are experiencing a greater amount of integration than ever before. This is mainly due to advances in communications, such as the Internet, that allow for the ready exchange of business across borders that was not possible in the past. This movement towards a "global economy" will be certain to have an impact on the foreign exchange markets. International markets used to be the realm of large corporations, but now this is not always true. The Internet has allows small and medium businesses to compete on this level as well. Foreign exchange rates are becoming an important issue for many people who were previously not concerned because it did not directly effect their lives. Now it is important for everyone to have an understanding of what drives foreign exchange rates.

The foreign exchange market is the only market that is open 24 hours a day. The day begins in Japan, then moves to Hong Kong and then to London and the United States. It is difficult to maintain order in such an environment and the central banks sometimes intervene through trading to make sure that global chaos does not erupt. This trading usually takes place using the U.S. dollar. This is partially because the U.S. dollar has lower transaction costs than other currency. A British exporter wishing to purchase Japanese yen would pay a transaction fee to the broker for the transaction. This ultimately drives up the price of the export transaction, or possible makes it less profitable and may serve to limit foreign trade activities. As smaller businesses enter into this market, the volatility caused by these transactions has a larger impact on an individual economy and the global market as a whole. Small businesses may not be able to survive the volatility as well as larger businesses that often have options not available to small businesses to protect themselves from the downside risk of this volatility.

Many economic theorists have constructed models to help predict this volatility. The following research will explain some of these models including the Purchasing Power Parity Model, Monetary Model, and the Portfolio Balance model. These are not the only models; however, they are the most widely accepted among those who play the foreign exchange markets. These models are not perfect and these imperfections will be the subject of further discussion. There have been many academic studies conducted around these theories. The following research will examine some of this work and discuss its impact. It is hoped that this research will help give a well-rounded understanding of how the foreign exchange rate market operate and its effects on foreign trade volumes.

Purchasing Power Parity (PPP)

The real question is how to predict volatility trends and be able to adjust sales and exports in relation to the predicted exchange rates. There are many models that propose to do just that. One of the first models for attempting to predict volatility trends was the Purchasing Power Parity

PPP) model. This model assumes that the exchange rate between two currencies would be equal to the relevant national price levels of those countries. This would result in a common currency rate and the common currency would have the same purchasing power per unit of goods in each country. This is a nice theory, if the economies of the two countries are similar and have similar inflation rates, GDP and other similarities, but this is usually not the case.

The PPP is often discussed in terms of absolute PPP and relative PPP. Relative PPP occurs when the rate of depreciation of one currency relative to another matches the difference in aggregate price inflation between the two countries (Sarno and Taylor, 2002). The real exchange rate is adjusted for relative national price differences between the two countries. If the PPP model is true, then the real exchange rate is constant. This would mean that fluctuations in the real exchange rate would represent a deviation from the PPP. This means that the two are directly related and that the PPP is directly related to the real exchange rate (Sarno and Taylor, 2002).

It has been difficult to prove this model in the long run. It is difficult to prove empirically, although many valiant attempts have been made using different time periods of data. These results are widely mixed One of the most widely accepted and cited studies was conducted by Diebold, Husted and Rush (1991). It supports the ability of the PPP model to predict long-term volatility in foreign exchange markets. These researchers cleared many of the problems associated with previous research by eliminating the unit root for a large sample of exchange rates. This had been a confounding variable in many of the previous tests, especially those that dealt with small sample sizes.

Diebold, Husted and Rush developed a model that worked in the long-term. However, this example is set against many others before them who had failed to produce long-term predictive results.

This work seems to work as long as we are comparing apples to apples, that is industrialized nations to industrialized nations. These previous studies, not only fail to predict long-term results, but do not take into account shocks to the market and their effects on volatility. They seem to have forgot that it is volatility that we wished to measure in the first place and shocks cause volatility, therefore the negligence to account for shocks to the market, make the test results only valid in theory and they do not work in the real world.

There are many problems associated with assumption in the PPP model. One of the key assumptions is that the same weights are applied to each country. Price index weights will not be equal for all countries. Furthermore these weights change over time. Not every country produces the same goods n and services and therefore cannot be equally paired with another country that produces different goods and services. Some goods and services may be absent altogether from one country. These balances shift over time and the PPP model fails to account for these shifts.

Summers and Heston (1991) solved this delimma by borrowing a technique from estimating the Gross Domestic Product (GDP) of a country. In order to make a fair comparison of the production levels of unequal countries, economists construct a variable using a common basket of goods for each country. That is each country is evaluated on the same goods as every other country regardless of any other product that they may produce. It is an attempt to compare apples to apples. Summer and Heston call their model the International Comparison Programme (ICP) data set.

This would seem like a reasonable solution to the problem of how to compare unequal countries, however, they made several errors in the development of their model that make it impractical to estimate PPP. They used random, large time intervals and these are not necessarily the same time intervals used in PPP equations. The data would be comparing two different economic time periods and therefore is not a valid solution to the problem. For this reason, the price indices from official sources are still the basis for PPP calculations, even though they may not be entirely flawless either.

The modern version of PPP developed in stages. First, there was the original version of the equation in and all of its faults. Then there were the random walk hypothesis, cointegration studies, and long-term studies, panel data studies and finally studies that did not use linear econometric techniques. Absolute PPP draws a correlation between the ratio of two national price levels and the nominal exchange rate. If this were true then changes that effect one would automatically effect the other in the same amount. There are many effects that this would over look. The first thing that is overlooked is that short-term and long-term effects may be different.

PPP was originally supposed to be applied only to long-term analysis. Some early criticisms of the theory indicate that the theory is not supported and does not hold true when applied to short-term analysis. They therefore rejected the theory altogether. Frenkel (1978) applied PPP to long-term data and found it to produce a close unity on long-term data for high inflation countries. This would seem to confirm the posits of the developers of PPP. However, there are several things that Frenkel neglected to account for which tend once again rally support for the rejection of the PPP model. Frenkel neglected to test the stochastic properties of the residuals to determine if they are stationary. This could have the effect of making shocks to the real exchange rate have a lasting effect, which would mean that PPP was not a real estimate of long-term effects.

There are many factors that the PPP model fails to consider and it appears that the PPP model only works in long-term, highly inflationary economies. Even these applications have been criticized. There are many later versions of this theory that consider nonlinear and linear economic models. The PPP theory continues to suffer from a lack of conclusive empirical evidence. Therefore this theory is perhaps the least widely accepted of the economic models for predicting exchange rate volatility.

Further work needs to be done on real exchange rate behavior as opposed to theoretical exchange rate behavior in order to devise a better and more realistic working model of the PPP. In addition the proper circumstances must exist in order to apply the PPP model. It is not suitable for predicting short-term volatility. It may be useful in predicting long-term volatility under the prescribed market conditions. It is a useful tool, under certain circumstances, but has a limited applicability. It may also be a useful tool to confirm data obtained be other methods, as a checks and balances system for long-term predictions.

Monetary Approach to exchange rate

The monetary model of the exchange rate is theory that the exchange rate is the relative price of foreign and domestic money. It bases this determination of the price on a simple supply and demand model for these moneys. This model is simplistic and as such lacks several factors. It makes many assumptions that simply are not based on the real world. The first assumption is that foreign and domestic assets are direct substitutions for one another. Another is that foreign exchange markets work according to economic models.

There has been an overwhelming lack of support for this model. The rejection of this theory may have been due to simple statistical errors, involving small sample sizes and the limited number of time-series observations. It may be that model works it proper statistical methods are employed. The model has been rejected largely on the basis of a lack of cointegration relationships. When these errors are corrected that monetary model does show favorable ability to predict behaviors when measured on a quarterly or annual basis..

When applying proper statistical method, it is found that pooling time series can correct error and effectively makes the monetary model a useful tool (unknown 2). Other studies found it useful to use panel data to achieve a more useful prediction using the monetary model (Levin and Lin, 1992: Pedroni, 1995)

Foreign exchange markets are much more complex than domestic markets and it is this different degree of complexity that makes the monetary inadequate at predicting volatility. The Foreign exchange markets involve more product substitution from more sources. They also involve massive differences in economies of size. This is especially true when it comes to differences between undeveloped and industrialized countries. The foreign markets are not equal and do not represent a truly competitive atmosphere.

Monetary theories work for predicting behavior within a country because the disparities in variables are not as great as they are on a global level. The monetary model may work when trying to predict volatilits between a set of similar countries, as is discussed by the data pooling results. This effect only works when the countries are similar in characteristics.

Like the PPP model, the monetary model is not universally applicable in every situation. It may prove useful in predicting volatility among groups of similar countries and economies, but is be no means applicable to predicting exchange rates on a global basis. Like the PPP model, the limitations of this model must be considered when deciding to use it to as basis for real predictions.

Thus far, we have been unable to find the "magic pill" for predicting foreign exchange volatility on a global basis. This is the "Holy Grail" of global economists and thus far has eluded us due to lack of empirical evidence and a formula that is applicable in every situation. The monetary model has many shortcomings and it is for this reason that it is not widely used in foreign exchange models.

Portfolio Balance Model

The foreign exchange markets can be compared to the stock market, as the investment in foreign currency can be compared to the return on investment that a person expects from investments in the stock market. Many investors who have traditionally in the stock market are now participating in the foreign exchange markets. This is primarily due to the increasing number of companies that are cross-listed on two different markets. In order to invest in these companies it is necessary to understand how foreign currently markets work. These investors have applied some of the techniques learned on the stock market to the foreign exchange markets. The most common technique, borrowed from the stock markets is the portfolio model of investing.

The portfolio balance model was developed as a result of work by Allen (1973), Branson (1975), and Allen and Kenen (1976). This model is the most widely accepted model. It is based on the approach that foreign exchange rates are not determined the flows of buying and selling, but by the perceived value of long-term assets by a pool of foreign investors (Zietz, 1994). This model takes into account that the market is driven by news and that investors use a speculative approach to their decisions regarding foreign investment and future export decisions.

This approach was studied by Sengupta and Sfeir (1996) as it applied to the foreign exchange markets. IN the stock market there is a skewness preference in a bullish market. This phenomenon tends to disturb the normal mean variance relationship. It says that is the average investor is optimistic about the future in respect to positive future returns, then they will ten d to increase their foreign interests. The inverse is true in a bearish market. This study applied stock market techniques to investors on a global scale. This far, the results obtained by the portfolio model are more widely applicable and accurate in predicting volatility than the PPP or the monetary models.

The portfolio model inadvertently accounts for the emotional responses of foreign investors in relation to their known responses to risk aversion and opportunity taking. This model is based on the stock market, which tends to be highly reactive to news and perception of individual investors. The foreign exchange market is moving towards more diversification and is beginning to react much like the stock market.

International traders can be compared to the stock market volumes and currently tends to react in much the same way. The portfolio balance model is the most accurate prediction model to date. However, like a mutual find approach to stock investing, the portfolio balance model is good for short-term trends and as the world moves to a more globalized economy, the portfolio may be subject to the same reactionary factors as the stock market in the long-term.

Literature Review

Thus far we have discussed the various models for predicting foreign exchange rate volatility. Of the three major models presented, one thing stands out, a lack of conclusive empirical evidence to support any of the three discussed. There is no definitive way for companies involved in import and export to know for certain which way the markets will turn. They may choose to avoid risk and only sell their products domestically, or they may decide to participate on an international level and take on the advantages and risks that the foreign markets can provide.

When a company decides to engage in international trade, the negotiated currency for the trade can go up or down between the time of the agreement and the time of the delivery and payment for the product. This means that the company is in theory acting as an investor in the fact that they are making an agreement with the expectation of receiving good return on their investment. The foreign exchange market can go either way, making the final payment more, less or the same as was first negotiated.

Stable foreign exchange markets tend to allow investors to make a predictable return and should theoretically result in an increase in trade. Due to the risk aversion, behaviors of investors, it would be expected that times of high volatility would lead investors to tend to shy away from foreign trade and tend to participate in their own domestic market.

The risk aversion tendencies have been the overwhelming subject of many studies concerning foreign trade volumes in correlation to foreign exchange rate volatility. The majority of these studies have pointed to a decrease in foreign trade volume in times of thigh volatility. This would lead us to classify the company who participates in export activities to be a low risk taker who is afraid to lose money. This following literature will discuss the validity of this profile through a critical examination of the research that led to this profile.

Exchange Rate Volatility and Negative Effects on International Trade

Overwhelming evidence has been presented by some sources that have a wide degree of credibility including Giocanni Dell'aricia of the International Monetary Fund (Dell'ariccia, 1999). This study analyzed the effects for exchange rate volatility on bilateral trade flows. Exchange rate volatility was found to have a negative effect on international trade. Dell'ariccia stressed that this is especially true if the volatility is unexpected and was not previously predicted. The foreign trade market does not like surprises. He did however, fail to mention whether this broad statement included only negative volatility, or if it included positive volatility. He did not clearly define what was meant by volatility. Did he use both negative and positive volatility for a measurement? This would suggest that investors only prefer a stable market and do not even like positive surprises. However, because Dell'arricia failed to define volatility we cannot be certain fi this was the case. The latter situation, a negative reaction to positive news, would tend to go against the laws of human nature.

Dell'arricia focused his research the European Union (EU) largely because the participants are equal in size and exposure to risk. This was an attempt to eliminate the effect that comparing countries of unequal economic size may have. A country with a smaller economy would not be as able to absorb volatility as a larger country and therefore may perceive greater personal risk by participating in a more volatile market.

One of the key factors in forming the EU and the European Monetary Union (EMU) was to try to minimize exchange rate volatility. IT was meant to correct major discrepancy among currencies in Europe. This exchange rate uncertainty had been a major barrier to trade among European nations. It was hoped that the EU would stabilize trade relations and promote trade among the European nations. This was the subject of Dell'arricia's study, whether these measures had been effective and achieved their goals.

The original plan for the formation of the EU was to allow countries who were ready, to enter first and then prepare the others for entrance at a later time. This idea was largely rejected as it was perceived that this would cause a division and be a hindrance to trade for countries who would not enter until the second stage (Dell'arricia, 1999). This would mean that the sample design for the study would have to consist of two panels, one for primary entrance countries and one for secondary entry countries. However, Dell'arricia did not use this approach therefore it is not known if the staged entrance had an effect on foreign trade volumes between the two panel groups.

This study was the basis for policies concerning trade policies in the EU. It is surprising that such a credible source could make such simplistic design flaws. There were many confounding variables that were not considered in this major study. The trade relationship between industrialized and developing countries would have different characteristics. This study made generalizations that may not be applicable to the situations to which they were applied. This study has been discussed in detail because it embodies the majority of results for studies that confirm the existence of a relationship between low foreign trade participation and exchange rate volatility.

Another such study was conducted by Arize and associates (2000). It achieved the same results but attempted to correct at least one major flaw in the other studies. Arize made the point that many of the other studies had concentrated on industrialized countries (Chowdhury (1993), Thursby and Thursby (1987), and Kenen and Rodrik (1986), among others.. Arize wanted to see fi the same results would be obtained when comparing developing countries. The Devaluation of the U.S. dollar in 1973 caused a high degree of foreign exchange rate volatility. Arize examined the effect using 13 of the Least Developed Countries (LDCs) in the world.

De Grauwe (1988) found a correlation between the dominance of income effects over substitution of product effects. De Grauwe surmised that if exports were characteristically risk adverse then export rate volatility would raise the expectation of return and they would increase exports. He did make the rational point that this effect would be subject to the personal amount of risk aversion of the certain exporter. Other studies tended to ignore the differences between various exporters and place them into a general high-risk aversion category. This brings up the idea that there are many psychological and sociological effects that have an effect on the economy. It is a major failure of economic model to exclude these factors. It cannot be forgotten that the actions of people make the markets and that mathematical probabilities cannot possibly account for these variables. De Grauwe at least acknowledged these variables, but still failed to account for them.

Two studies attempted to account for the hysterical factor and its effects on foreign trade (Dixit 1989, p. 206). They called this factor "hysteresis" and found that uncertainty from high volatility in exchange rates did have an effect on foreign trade. However, they did not indicate whether the effect was in a positive or negative reaction. They only measured changes in volume, not direction. Froot and Klemperer (1989, p. 643) confirm these results and demonstrated that exchange-rate uncertainty can affect the price and quantity of trade, either positively or negatively.

So far we have only discussed studies that determine that exchange rate volatility does in deed have an effect on export volume. A majority of the studies, and there are many more not mentioned here, only conclude that a relationship does exist. This group of studies makes no indication as to whether the effect is positive or negative. These studies are numerous and tend to only confirm one another.

The other studies can be grouped into two camps of thought. One group of studies only looks at downside risk and risk aversion characteristics of the exporter. These studies, when taken as a group paint a picture of the typical foreign investor. This picture is one an investor who has a high aversion to taking risks and id the possibility of losing money exists, they will take the safe route every time and avoid the risks of the foreign exchange market. This paints a rather pessimistic view of foreign traders. Where some see lemons, other may see lemonade. The studies in this group completely ignore the inventor who may see market volatility as an opportunity rather than a risk.

Another group of researchers focuses on foreign investors who would see the foreign exchange rate volatility as an opportunity to make a considerable profit and would engage in more trade, rather than less. Broll and Eckwert (1999) focused on the positive effects of exchange rate volatility and found that will play foreign trade as an option using forward contract. These investors certainly do not fit the profile of being highly risk aversive, but rather seem to be opportunistic in nature. This study found that high volatility increased the rate of foreign trade among these investors.

The two camps argue on this issue and there is no conclusive evidence for either side. The only conclusive evidence is that an effect exists. That is the only point that is not up for debate. Gagnon 1993; Dellas and Zilberfarb 1993 conclude that higher exchange rate volatility has led to a decrease in international trade. Cushman 1988; Giavazzi and Giovannini 1989; and Stein 1991 say that these results are inconclusive. Yet there are others who say that volatility has a positive effect on international trade volume. It would seem that the two sides are irreconcilably divided and the issue is plagues by a plethora of poorly designed studies that fail to take into account the human side of market activity. The large majority of studies fail to make a statistically significant link.

De Gauwe (1992) attempted to clarify the situation by dividing 12 major industrial countries into two groups. One group, primarily consisting of members of the EMU, had relatively stable exchange rates. The other group has high degrees of exchange rate fluctuation. He found that during the 1980s the growth rates of exports were significantly lower in the EMU countries than in non-EMU countries. This would lead us to believe that high volatility means greater opportunity and thus more trade activity.

Foreign Currency Exchange Hedging large body of theoretical work conforms to the idea that a firm can avoid much risk if it avoids export activity in highly volatile markets. This is certainly a safe philosophy and is entirely a personal decision. But we all know that the world is not made of a group of people who are afraid to take risks. Some companies have demonstrated a desire and ability to effectively take advantage of potentially risky situations and expand their production in response to the opportunity for foreign trade. Previous studies considered combined results and did not identify any of these companies or sectors.

One factor that has led to the ability to do this in more recent years is advances in communications and transportation that have enabled companies to react quickly to respond to changing market conditions. Deals can happen faster and do not take weeks or months to arrange and they did in the past. This ability creates a greater potential and presents a lessened risk due to the ability to react to current market conditions. As discussed earlier, there are a number of models to predict volatility trends in foreign exchange markets. These may have some usefulness in predicting long-term trends, but they do not have the ability to predict today's market. They are good in retrospect for the identification of trends, but real money can be made by being able to take advantages of immediate changes in rates.

Companies with large domestic markets can assume more risk than those with smaller markets, the same is true of companies that choose to participate in foreign markets. It would be expected that due to their lessened risk, they would be more willing to attempt to participate in a volatile market in order to take advantage of an opportunity. Cushman (1988) supports this idea.

Exports can be treated as an option when the domestic market return is certain whatever the realized exchange rate turns out to be. In this case the domestic price of the product fi sold in the country of origin acts as a "strike price." There is now a possibility to export when the exchange rates are favorable. In order to take advantage of this strategy, a company must have sufficient stock on hand and be able to be flexible on their timing to export. Not all companies are able to easily do this, for instance companies who produce goods that have a limited shelf life, or companies that do nor have sufficient warehouse space to store inventory in readiness for shipment for an extended period of time. In this way, exchange rate volatility can have a positive effect on international trade volumes.

Earlier more pessimistic studies made several fatal assumptions. They failed to account for the ability to be flexible when it comes to foreign trade. These models assumed that the company sells all of its products on the foreign market. They also assumed that the companies sold their entire product at once. In reality the companies ship in increments spaced over a period of time. In this way, a stock technique also comes into play called "dollar/cost. Averaging." Dollar cost averaging means buying stock at regular intervals in order to avoid risks of sudden fluctuations. The cost and risk averages over a period of time.

In addition to hedging and dollar/cost averaging techniques, companies use diversity to protect against risk. The company usually has products that they sell domestically and foreign trade only accounts for a portion of their income and therefore only accounts for a portion of their risks as well. Both the economic models and the studies discussed in this section would have the reader believe that foreign trade is an all or nothing, do or die type of situation. Companies have many techniques at their disposal to help hedge against, or take advantage of a volatile foreign exchange market.

In the real world production decision must be made prior to shipment, therefor they are subject to future volatility. Companies use many techniques to guard against the risks and seldom do they take these risks foolishly. Seldom are they reactive as reactive to volatility as previous studies would have one believe.

Doyle, (2001) explored a real-world example that centered on the effects that exchange rate volatility had on exports between the United Kingdom between 1979 and 1992. The economies of Ireland and the UK have been linked for many years and the instability of the economy of the UK had a profound effect in the economy of Ireland. Irish authorities were enthusiastic about joining the EMU in 1979 because of the stability that it would bring (Doyle, 2001). Ireland had relied on the UK market of its primary source of export income. When the UK began to experience problems, it became apparent the Ireland needed to diversify its foreign export opportunities and take measures on if own to stabilize its own economy. They have attempted to d this but still export a considerable amount to the UK (Dellas and Zilberfarb, 1993). In this case foreign exchange rate volatility did cause Ireland to seek other avenues to distribute their products. This leads us to conclude that volatility did in deed effect the willingness of Ireland to trade with the UK and seek a more stable relationship.

This was an isolated case and only involved two countries. It cannot be applied to bilateral trade between every country in the world as there are simply too many variables to be accounted for. However, it is still a useful example as it gave us one trading pair that had long trading history and had a one way exclusive relationship, as far as Ireland was concerned. It gave us the opportunity to measure a definitive reaction in a stable market that became unstable.

In this case Ireland did reduce bilateral trade in response to uncertainty within this bilateral agreement. It served as a wake up call and caused Ireland to wish to employ risk reduction strategies by diversifying. Because this study only included one sample, it does not qualify as truly empirical case, but taken as an example it doe support the theories of exchange rate volatility having a negative effect on international trade. It would greatly add to our ability to draw a conclusion if more bilateral cases were to be explored and compared as this one has dome.

Thus far there have been a wealth of studies exploring the effects of exchange rate volatility the majority of these studies give inconclusive results, either taken individually or when grouped together as a whole. They did find some effects of volatility on international trade volume, but were unable to draw conclusion as to whether this effect was positive or negative. This issue has divided researcher sin to three camps, those who conclude the volatility has a negative effect on international trade, those who feel that exchange rate volatility has a positive effect on international trade and those who can identify an effect, but who do not necessarily know whether it id good or bad. This leaves us only able to make one general conclusion. Exchange rate volatility does have an effect of the volume of international trade.

The key failures of the models to predict the direction that the volatility will take and the failure of the studies to accurately predict the direction of trader reaction coincide with one another. The direction of trade volume would e expected to have an effect in relation to the direction of exchange rate volatility would be expected to have a direct correlation. This aspect was ignored by previous studies and could help to make better predictions to both exchange rates and future trade volumes. This subject should be considered as a topic for future studies.

Another failure of past studies has been the failure to recognize that traders are humans have personalities that effect their decisions. These decisions are base on individual preferences and risk aversion levels. Study results may have had an unidentified confounding variable of choosing a sample population of wither high-risk aversion managers or low risk aversion managers. If the sample were skewed in one direction, it would effect the final results. Further studies need to be mad that take into account risk aversion attitudes of management.

Modeling Analysis

The following modeling analysis is based on a study conducted by Abdur Chowdhury (Chowdhury, 1992). The objective of the study was to determine if exchange rate volatility depresses trade flows. It was similar to other studies previously reviewed in the literature review. The study used an error-correction model to draw its conclusions. Conclusions drawn were consistent with previous studies indicating that exchange rate volatility does indeed have an impact on foreign trade, due to risk-aversion behaviors of the companies involved.