Exposure and the Introduction of the Euro

Exposure and the

Introduction of the Euro

An investigation on the change in Foreign

Exchange Rate Risk of listed European firms

Master Thesis

MSc Business Administration - Finance

by

Daan A. Hendrix

s1554956

Student at the

University of Groningen

Faculty of Economics and Business

Supervisor:

Prof. dr. K.F. Roszbach

Exchange Rate Exposure

and the Euro

An investigation on the change in Foreign

Exchange Rate Risk of listed European firms

Daan A. Hendrix

S1554956

D.A.Hendrix@rug.nl

JEL codes: F10; F31; F37

Key words: Foreign Exchange Rates; Exposure; Firm Value; Euro

Abstract:

I.

Introduction

Section 1.1 Introduction

On the first of January 1999, at midnight, the euro was introduced in the European Union. This experiment in monetary modification is a milestone in European integration. In a short period of time, it transformed the European political and economic landscape. Never before had some of the world’s largest economies agreed to surrender their national currencies and national sovereignty in favour of a common currency and a common central bank. It is seen as one of the most exiting experiments in monetary history. The goal was to form a strong economic and monetary union with the underlying idea to decrease economic dependency on non European countries such as the United States or China. One of the most important arguments in favour of the European currency is that it would reduce foreign exchange rate risk1. In this manner, European companies or firms with a significant investment in Europe would benefit. It is clear that since the introduction of the euro trade between euro-area countries is now more transparent since there is no longer a risk of exchange rate exposure (ECB, 2008). However, it is still unclear whether the degree of exchange rate exposure in the euro-area with non euro-area countries changed due to the introduction of the euro. Therefore, the research question of this paper is:

Is there a significant change in foreign trade weighted exchange rate exposure in the euro-area after the introduction of the euro?

So this paper takes a new look at the exposure puzzle by studying the potential impact of the introduction of the euro. The introduction of the euro as a common currency in 1999 provides an useful experimental setting to investigate the foreign exchange rate

phenomenon. Regarding exchange rate exposure, there is already extensive literature written. However, the change in exchange rate exposure due to the introduction of a

1

common currency is still underexposed. This makes the research question interesting from a practical point of view. Was it wise for countries to participate with the euro, based on the implications and changes in foreign exchange rate risk? Or is their a point in stating that the euro-area countries should again make use of their former currencies? The empirical results have important policy implications as they demonstrate the possible benefits of currency stabilization for companies around the world. With lower foreign exchange rate exposures, firms benefit from increased potential to carry higher business risk or to sustain more financial leverage.

In this paper a firm-level analysis of changes in stock returns related to foreign exchange rate risk is carried out. The set up is as follows; in the following section, more details about the process of starting a monetary union and the fundamental changes in the euro-area are given, this is important for explaining why the degree of exchange rate exposure did or did not change. The third section gives more information about the literature review of exchange rate exposure. The first part of this section is written from a

theoretical point of view on exchange risk exposure. The second part focuses more on the empirical results of research on exchange rate exposure. In the fourth section, the

methodology is further examined, whereas section five gives an overview of the data used. The sixth section discusses the results obtained, it answers the main question whether the degree of trade weighted exchange rate exposure in the euro-area differs significantly after the introduction of the euro. Finally, section seven contains the conclusion.

II.

Exposure and the Euro

Section 2.1 The Introduction of the Monetary Union and the Euro

what exactly changed due to the introduction of a common currency. Was it easy to start a monetary union? Is there more intra trade due to the introduction of a common

currency? And what is the exact role of the European Central Bank (ECB) in the euro-area, now individual countries no longer have the possibility to intervene in their exchange rate? All these question are important to explain why the degree of foreign exchange rate in the euro-area did or did not change and whether it was wise to participate with the euro. The first subsection will give some historical background information about the process of forming a monetary union and the role of exchange rates fluctuations. Section 2.2 provides more details about the difficulties in introducing a common currency and how this affects the model used, while section 2.3 gives more information about the overall impact of the euro on trade. Section 2.4 examines whether the ECB actively influences the euro exchange rate, whereas section 2.5 gives more information about the European crisis and how this affects the decision of keeping a single currency. To begin with the start of the European monetary union.

Creating the European monetary union was an interesting but long process, partly due to difficulties in exchange rate movements. European integration already started in 1960 when six members of the European Economic Community (EEC) decided to cooperate in monetary affairs. One of the decisions, taken by the treaty of Rome, was to set up a free trade area with the underlying thought of starting an economic union at the end of the decade. In 1969 the decision was taken to draw up a plan for an economic and monetary union, leading to the Werner report in 1970 which focussed on the creation of a monetary union in three stages. However, these intentions already failed due to the downfall of the Bretton Woods fixed exchange rate regime in 1971. European countries recognised that untamed exchange rate movements could harm further trade integration. In 1972 another step was taken to stabilise exchange rate currencies among some European currencies. Again, this effort failed due to currency unrest and the international recession that

creation of the European Monetary System (EMS). The main focus of the EMS was on monetary policy coordination and convergence towards price stability in order to maintain stable exchange rates. A few important measurements were taken. Links between central banks were strengthened, capital controls were removed and low inflation policies were introduced in every country. Even more important, exchange rate realignments were made conditional on policy commitments in order to decrease the frequency and impact of exchange rate devaluations (ECB, 2008). Again, not every country kept its promises. An overview of the crisis in the European Monetary System is shown in Table 1.

Table 1. Overview of the crisis in the European Monetary System, during the time span 1991-1993 (Jonung and Drea, 2010)

The crisis in the European Monetary System (1991-1993)

November 14, 1991 Finland, which had maintained a peg to the European Currency Unit (ECU), devalues the markka by 12% due to the collapse of its Soviet trade and a domestic banking crisis.

June 2, 1992 Danish voters narrowly reject the Maastricht Treaty.

August 26, 1992 The pound sterling falls to its Exchange Rate Mechanism (ERM) lower limit.

September 8, 1992 Finland severs the markka’s ECU link.

September 13, 1992 Italy devalues the lira by 7% against other ERM currencies.

September 16, 1992 Britain suspends ERM membership. Italy suspends foreign exchange market interventions and allows the lira to float. Spain devalues the peseta by 5%.

September 20, 1992 French voters narrowly approve the Maastricht Treaty. November 19, 1992 Sweden abandons its ECU peg.

December 10, 1992 Norway abandons its unilateral ECU peg. January 30, 1993 Ireland devalues the punt by 10% within the ERM.

May 14, 1993 Spain devalues the peseta by 8%; Portugal devalues the escudo by 6.5%.

Despite the depressions, the discussion about an European and monetary union continued. The single European act, signed in 1986, was one of the first major steps towards a monetary union because it involved the end of exchange controls (Jonung and Drea, 2010). The next major step was the signing of the Maastricht Treaty in February 1992, where member states had to meet strict criteria (De Grauwe, 2007). One of these criteria was that participating countries had to join the semi-pegged exchange-rate mechanism (ERM II) under the European Monetary System for two years. Furthermore, during this period countries were not allowed to devaluate its currency.

One can conclude from the historical background that exchange rate movements are of great influence in creating the European monetary union. The first attempt failed due to the downfall of the Bretton Woods fixed exchange rate regime, whereas the second failed due to currency unrest and an international recession. It is clear that in the process of forming a monetary union, one of the main challenges was to stabilize exchange rates. Therefore, strict measurements were taken to stabilize the domestic exchange rates and decrease the frequency and impact of exchange rate devaluations.

Section 2.2 The Process of Introducing a Common Currency

The process of irrevocably fixing the national currencies with the euro was complex and has several implications for the model used in this paper. The name euro was already officially adopted on the Madrid Council of 15 and 16 December 1995. It was decided that the euro would replace the European Currency Unit (ECU) at a ratio of 1:1 on the first of January 1999. However, the conversion rates of these national currencies (shown in appendix A) must be equal to the closing rates of the market on 31 December 1998. Without the latter condition, jumps in the exchange rate of the national currencies at the start of the EMU would result in enormous capital gains and losses. These two conditions created potential self-fulfilling speculative movements of the exchange rates prior to 31 December 1998. Because the EMU announced that it would use the market rates of the last day prior to the European monetary union, any movement on the last day would be self-validating. To avoid such a situation, the authorities announced in May 1998 the fixed values at which the currencies would be converted at the start of the EMU. The market smoothly drove the exchange rates towards the fixed conversion rates announced (De Grauwe, 2007).

In May 1998 eleven countries met the convergence criteria determined by the Maastricht Treaty. They were Austria, Belgium, Finland, France, Germany, Ireland, Italy,

with fixed rates against each other (Jonung and Drea, 2010). An overview of the major steps towards the euro is given in Table 2.

Table 2. Overview of the major steps towards the Euro during the time span 1989-2002 (Jonung and Drea, 2010)

Major steps towards the Euro (1989-2002)

February 1986 Signing of the Single European Act, advancing economic and political integration within the European Community.

April 1989 The Delors Report calls for Economic and Monetary Union (EMU) leading to a single European currency through three stages. June 1989 The Madrid Summit of the European Council agrees that Stage 1 of

EMU will start July 1, 1990. Stage 1 includes completing the internal market and removing all obstacles to financial integration.

October 1990 The Rome Summit of the European Council agrees that Stage 2 of EMU will begin January 1, 1994.

December 1990

The Dublin Summit of the European Council marks the beginning of intergovernmental conferences on EMU and political union.

February 1992 Signing of the Maastricht Treaty to establish the European Union, the successor to the European Community.

June 1992 Danish voters narrowly reject the Maastricht Treaty.

September 1992 Currency crises force Britain and Italy to abandon the Exchange Rate Mechanism (ERM).

July 1993 Member states agree to widen the “narrow” band in the ERM from 2.25% to 15% around the central rates.

January 1994 Stage 2 of EMU starts. The European Monetary Institute comes into operation and begins the transition from co-ordination of national monetary policies to a common monetary policy. Economic convergence is strengthened through adherence to “convergence criteria” set out in the Maastricht Treaty

May 1995 The European Commission adopts a Green Paper “On the Practical Arrangements for the Introduction of the Single Currency.”(A green paper is a document intended to stimulate discussion and start a process of consultation).

December 1995 The Madrid Summit of the European Council reaffirms January 1, 1999 as the date for the irrevocable locking of exchange rates, thus for the introduction of the euro. The “euro” is officially adopted as the name for the new single currency.

May 1998 Special meeting of the European Council decides that 11 member states satisfy the conditions for adopting the single currency.

June 1998 The European Central Bank and the Eurosystem are set up. January 1999 Stage 3 of EMU begins. The exchange rates of the 11 initial

participating nations are irrevocably fixed and the euro begins to trade on financial markets.

January 2001 Greece adopts the euro

It is important to know that the authorities already announced the fixed values of the euro-area currencies in May 1998. Bartram and Karolyi (2006) state that, because of financial speculation, one should use 1998 as effective year for the introduction of the euro. In this manner, one accommodates the fact that the introduction may have been anticipated to some extent. In addition, Bris et al. (2003) gives two main arguments for using 1998 as benchmark year. First of all, the European Council already decided in May which countries were allowed to enter the final stage of EMU. The second argument is that the forward rates of all euro countries already converged mid 1998. This implies that using 1999 as actual introduction date of the euro would be too late. Bris et al. (2003) uses yearly data and therefore take 1998 as benchmark year. Since this paper examines monthly returns, May 1998 is used as benchmark month for the introduction of the euro.

Secondly, Greece is excluded from the sample since its membership was determined later than other euro countries. Thereby making it difficult to classify whether Greece was an euro or euro country in the period between the introduction of the euro in non-physical form (1999) and until Greece itself actually adopted the euro in January 2001 (Bris et al, 2003). In line with other papers examining the effect of introducing the euro, such as Bris et al. (2003) and Bartram and Karolyi (2006), Greece is not included in the sample. Next, it is discussed how the introduction of the euro changed the degree of trade within the euro-area.

Section 2.3 The Impact of the Euro on Trade

The main benefit of introducing a common currency is the stimulation of trade among euro-area countries. With a common currency there is no longer exchange rate risk within the area, capital costs are reduced, common trading platforms are used and stock

exchanges across borders merge such as Euronext. It is interesting to examine how the euro affected intra-trade as well as trade with the non euro-area, in order to fully

trade of the euro-area with non euro-area countries would increase the degree of exchange rate exposure for firms.

Since the introduction of the euro, the degree of openness for the euro-area with respect to financial markets and external global trade is increasing steadily. Moreover, the degree of openness measured by trade, international assets and liability investments has

increased by 11% since the introduction of the euro. Even more important for this paper is that this is particularly due to an increase in trade with new EU members and China (ECB, 2008). The value of exports and imports of goods within the euro-area was 27% of GDP in 1998 and increased to 33% of GDP in 2007. With respect to the EU15 countries that did not introduce the euro (Sweden, Denmark and the UK), the euro-area showed on average a 3% higher year-on-year growth rate of exports of goods in the period 1998 up to 2008. Also the year-on-year growth rate of imports of goods during the same time span was 2% higher in the euro-area. For more trade statistics, see appendix B, C and D (ECB, 2008). It is clear that the role of the euro in international markets has increased to some extent, however it should also be noted that the speed of change has been gradual (Ottaviano, 2007). Countries in the European Union but outside the Eurozone (like, Sweden, Denmark and the UK) also experienced significant increases in euro-zone trade after 1999. Dominguez (2006) argues that the increase in trade could also be due to the dismantling of trade barriers within Europe and the initial weakening of the euro after its introduction in 1999. Thereby making it difficult to pin down that the move towards a single currency solely contributed to these trade patterns.

no need to exchange currencies, one does not need commissions or insurance against exchange rate fluctuations. In their paper, the effect differs per country because of different institutions and access to technology, research & development. Allington et al. (2005) argue that the existence of a common currency changed pricing behaviour. In their article they show that, after the introduction of the euro, exporters no longer segment markets by applying country specific prices within the euro-area. However, Haskel and Wolf (2001) indicate that there are similar price differentials for exactly the same products sold by the furniture store IKEA across Europe. In addition, Engel and Rogers (2004), analyzed a set of more than 100 identical products in 18 European cities in the Eurozone from 1990 to 2003 and found no evidence for price convergence in the Eurozone.

One can conclude from the trade analysis that there is a high level of international trade. Both intra trade as well as trade with the non euro-area increased significantly. Based on the contradicting articles discussed, one cannot make a decent conclusion regarding price convergence and a possible effect on the degree of exposure. With respect to trade, it is clear that not only intra area trade but also non euro-area increased. From a theoretical point of view, more intra euro-area trade decreases the degree of exchange rate exposure, while more non euro-area trade leads to an increase in exchange rate exposure. This makes it difficult to predict the exact effect of the euro on exchange rate exposure. Finally, the influence of the introduction of the euro differs per country and industry.

Section 2.4 The Role of the ECB and the Euro Exchange Rate

contrast, economic policy (fiscal or structural) remain largely in control of the member states. The main reason for this dichotomy is that monetary policy in a monetary union is indivisible by nature, while economic policies need to deal with differences in

characteristics and institutions between countries (Dominguez, 2006).

The ECB uses the interest rate to keep prices stable. With respect to exchange rate policy, national financial ministers are given authority in article 11 of the Maastricht Treaty to formulate general orientations and formal agreements for the euro if it is not against the ECB’s price stability mandate. However intervention operations, which involve purchase or sales of euros in the foreign exchange market with the goal of relatively influencing its value, are granted to the ECB and national banks in article 23 of the Maastricht Treaty. The ECB did intervene in the currency market on September 22 and on the 3, 6 and 9 of November 2000 to strengthen the euro against the US dollar. In almost a full decade, the euro is seen as a stable currency and widely accepted in the international currency market. One could argue that the ECB managed to keep prices relatively stable since the introduction of the euro, with an average annual HICP inflation rate of slightly above the 2% in the euro-area (De Grauwe, 2007). More important, despite the fact that the ECB is essentially independent of political influence and has the possibility to influence the value of the euro, it is clear that the ECB does not actively intervene in the euro exchange rate.

Section 2.5 The European Crisis

a singly authority, the European Treasury, that is responsible for tax policy oversight and government spending coordination of EU member countries. Despite these

measurements, some still argue that the euro-area should again make use of their old currencies. The IMF already announced that the emergency fund of the Eurozone is insufficient to cope with the financial problems. In addition, Andrew Bosomworth, top manager of worlds’ largest obligation investment company Primco, announced in the German paper ‘Die Welt’ on the 20th of January 2011 that the only possibility for Greece, Portugal and Ireland to pay off its debts is to leave the Eurozone. Only with an own currency they will be able to stimulate export and boost their domestic economy. This illustrates the importance and relevance of this paper. Are their indeed advantages of having a single currency or is the euro-area better of with their former currency? This paper contributes to the question by investigating the change in exchange rate exposure in the euro-area after the introduction of the euro. Before the latter is investigated, one should first know more about previous literature on exchange rate exposure.

III.

Literature review exchange rate exposure

Section 3.1 Theoretical Review

Table 3.The effect of an appreciation of the home currency on the value of a domestic industry (Bodnar and Gentry, 1993).

Activity Sign of effect

Non-traded good producer (+)

Exporter (-) Importer (+) Import competitor (-) User of internationally-priced inputs (+) Foreign investor (-)

Exchange rate exposure of a multinational operating firm is thus, among others, determined by the proportion of export sales, level of foreign competition and the

substitutability between imported and local goods. There are several methods to decrease the degree of exchange rate exposure. High inelastic consumer demand gives the

producer the opportunity to pass price changes due to exposure on to consumers (Bodnar et al., 2002). Dumas (1978) accounts for the firm’s responsiveness to exchange rate changes. He focuses on the unique ability of multinational firms to move production from one country to another and thereby reducing their exchange rate exposure. The article of Allayannis and Ofek (1997) shows that the use of foreign currency derivatives can reduce exposure. Adler and Dumas (1984) show how firms can hedge their exchange rate risk. This is best shown with an example:

all alike and domestic currency risk is zero. However, the French inflation rate and consequently the exchange rate are random. An US investor receives 1000 FF in three months. The question is what his exposure will be on the target date three months away? 1000 FF will exactly represent the sensitivity of the future dollar value to variations in the exchange rate. Assume a forward contract that allows the investor to sell the 1000 francs against the forward rate F. Then the pay off, $1000(F-S), is exactly the value needed to shield the dollar from unanticipated variations in the exchange rate. The dollar value of the hedged position remains constant at $1000F independent of the economic state of nature. For an overview, Table 4 can be used.

Table 4. Exposure of 1000 Franc to be received in the future, with three economic states of nature. P* stands for the French franc balance, S for the nominal exchange rate and F indicates the forward rate

(Adler and Dumas, 1984).

Future Quantities State 1 State 2 State 3

FF Balance: P* 1000 1000 1000 Exchange rate: S =$/FF 0.25 0.225 0.200 Dollar Value: SP* $250 $225 $200 Proceeds of Forward Sale: $1000(F-S) $1000(F-0.25) $1000(F-0.225) $1000(F-0.20) Dollar Value of Hedged Position $1000F $1000F $1000F

One can see that the amount of exposure is perfectly hedged, so why is the discussion and research on exchange rate exposure still essential in modern economics?

value of the foreign currency on that date. More difficult are assets or liabilities (physical or financial) whose value in future foreign currency is uncertain. The value may also be sensitive or correlated with exchange rate fluctuations and should indeed be considered exposed. This leads to the notion of exposure as a statistical property what makes it interesting to investigate. The degree on which firms are able to hedge the exchange rate exposure of assets or liabilities with an uncertain future foreign currency value is still not clear. Bodnar et al. (1998) show that less than half of the payables and receivables are hedged and that most hedges are short term. Guay and Kothari (2003) even argue that in the case of perfect hedging, derivatives positions held by US non-financial firms are only around 1/15th of the size estimated effect on firm market value. Furthermore, Brown (2001) finds that firms even hedge for many speculative reasons that are not consistent with financial theory. It should be noted that all these studies do not account for

operational hedges, which are probably more important for mitigating exposure. Despite the latter discussion, it is clear that hedging activities decrease the correlation between stock returns and exchange rates. To improve the ability of multinational firms to manage their exchange risk exposure, it is important to understand more about the volatility of stock return due to changes in exchange rate. Previous empirical findings with respect to the relation of stock returns and exchange rate exposure are discussed in the following part.

Section 3.2 Empirical Review

As mentioned in section 3.1 theoretical papers, such as Bodnar et al. (2002), predict a significant degree of exchange rate exposure. However, empirical studies often fail to find a strong relationship between stock prices and the exchange rate. They found some economic evidence, but the economic importance is often small.

exchange rate. Jorion (1990) found evidence for exchange rate exposure. In his article he shows that foreign sales are the main determinants for exchange rate exposure for large U.S. multinational firms. First of all, he found that the exchange rate exposure of US firms without a high amount of foreign operations does not appear to differ across domestic firms. Secondly and even more important, he shows that firms with a high amount of foreign operations have a higher degree of exchange rate exposure.

It should be noted that the results of Jorion (1990) are not very strong. He found that only 15 of the 287 US multinational corporations investigated faced significant exchange rate exposure during the period 1971-1987. In contrast, for the same period, Bartov and Bodnar (1994) did not found significant evidence for exchange rate exposure on U.S. multinationals. Bartov and Bodnar (1994) reexamined the relation of expected changes in the dollar value and equity in the hope of getting more significant results. With respect to previous research, they tried to deal with the problem of sample selection and only included companies with the same expected exchange rate exposures. Secondly, they tried to deal with mispricing using lagged variables and contemporaneous changes in the US dollar and firm value. Despite their extra effort, they failed to find better results. Possible explanations given in the article were complexity of the relationship between currency changes and firm performances, liabilities and assets. Goldberg and Knetter (1997) also conclude that local currency prices of foreign products do not fully respond to exchange rate changes. This argument of lagged reaction is contradicted by the article of Doidge et al. (2006), who found that the economic magnitude of the lagged variables is small and overall generally insignificant.

indicate that most studies use trade weighted exchange rate indices, derived from national trade figures with foreign countries. Thereby assuming that individual firms are

uniformly related to these national figures, which is in their opinion obviously not the case. The third possible reason for the weak relationship could be the fact that firms shield themselves against exchange rate risk.

After the research of Jorion (1990), also non-US firms were investigated. Bodnar and Gentry (1993) compared the US industry with Japan and Canada over the period 1979-1988. They used a model of stock market returns to industry portfolios for Canada, Japan and the USA. Exchange rate changes were examined in relation to industry profitability and value. The results indicate that exposure differs per industry. In addition, they found a statistically significant exchange rate effect on common stock returns of 20% up to 35% of the industries investigated. However, a larger effect was found for Japan and Canada than for the United States. Dominguez and Tesar (2001, 2005) found a relationship between foreign activities and exchange rate exposure for eight non-US listed firms. In addition, they split the data in three separate subperiods and examined the consistency of the coefficient, measuring the exchange rate exposure during these subsamples. In general, they found that the extent of exposure is about the same in the full period as in the three subperiods. So the exposure findings are not driven by a particular sub sample. The article concludes that exposure is correlated with firm size, multinational status, foreign sales, international assets, competitiveness and trade at the international level.

Table 5. Overview of empirical papers dealing with Exchange Rate Exposure. Paper Countries investigated Time Investigat ed Exposure found Yes/No Degree of exposure Weak/moderate/Stro ng Factors of influence on degree exposure

Jorion (1990) USA 1971-1987 -Yes -Weak -Degree of foreign operations (+) Bodnar and Gentry

(1993) -Canada -Japan -US 1979-1988 -Yes -Yes -Yes -Moderate -Moderate -Moderate/Weak -Import/export ratio (+) -Foreign assets to total assets (+)

-degree trade industry (+) Bartov and Bodnar

(1994)

US 1978-1989 -No -Very weak, only small lagged relation found

- Dominguez and Tesar

(2001,2005)

8 Non US firms 1980-1999 Yes -Strong -Firm size (-)

-Multinational status (+) -Foreign sales (+) -International assets (+) -Competitiveness (+) -trade at industry level (+) Griffin and Stulz (2001) -US

-Canada -Uk -France -Germany -Japan

1975-1997 No -Very weak relation, negligible

-Traded good industry(+)

Williamson (2001) -US -Japan

1973-1995 -Yes -Yes

-Strong -Foreign Sales (+) -Degree of competition (+) -Hedging (-)

-Foreign production domestic country (-) De Jong et al. (2002) -Netherlands 1994-1998 -Yes -Strong -Total assets (+)

-Foreign sales ratio(+) -Foreign loans (-) -Foreign operations (-) Doidge et al. (2006) -18 Countries 1975-1999 -Yes -Moderate -Firm size(+)

The main conclusion of the literature review is that most articles do find some relation between exchange rate changes and the stock value of firms. With respect to the United States the results are less convincing, probably due to the fact that the US is a relatively closed economy. Empirical research so far failed to find a strong significant relationship. Important factors that do influence the degree of exchange rate exposure are firm size and the level of international sales. Not all results found are of high economic value, mostly due to lack of data or possible hedging strategies of companies. Interesting to see is that, despite the critique of Doidge et al. (2006), the model of Adler and Dumas first used by Jorion (1990) is still leading in most papers. Section 4 explains why the model of Adler and Dumas (1984) is still important.

IV.

Methodology

Section 4.1 The Model Discussed

Adler and Dumas (1984) show in their article that exposure can be measured best with a statistical regression technique. The amount of exposure is the coefficient of the exchange rate, in a linear regression of an asset’s future domestic currency market price on the (set of) contemporaneous foreign exchange rate(s). As already mentioned, assets or liabilities (physical or financial) which value in future foreign currency is uncertain may be

correlated with exchange rate fluctuations and should indeed be considered exposed. Adler and Dumas (1984) show that the company’s exposure in terms of a certain foreign currency can be measured by the value of the foreign currency future contracts that investors must sell to minimize their variance of a mixed hedging portfolio of stocks and currencies. The exchange rate exposure of investors is the slope of a linear regression, with the value of foreign cash flows in home currency on the percentual change of the relevant exchange rate. In this manner, the regression parameter (or slope) is the units of forward foreign currency that must be sold to hedge exchange rate exposure. They show mathematically that after minimizing the variance of the hedged position, the residual randomness of the hedged position is independent of the exchange rate. The advantage of measuring exposure with a statistical regression technique is that it decomposes the probability distribution of a risky asset’s domestic currency price at a future instant into two parts, namely one that is correlated with the exchange rate(s) and a second that is independent. The first component can be defined as exposed as its variability can be removed by hedging. For the second component, residual variability remains but is not exposed to exchange rate risk. Since its randomness is not correlated with any exchange rate, it cannot be further reduced by hedging instruments such as forward currency transactions (Adler and Dumas, 1984).

exposure with the help of a simple linear regression of stock returns on the exchange rate. More details and modifications to the Adler and Dumas model will be discussed in the following part of this section. In section 4.2, the basic model is discussed while in section 4.3 up to 4.6 the model specifications are further examined.

Section 4.2 The Basic Model

Determining the exposure on industries is a difficult task, since there is still a lot of discussion going on how to measure exchange rate exposure. Previous studies already illustrated that finding strong explanatory results is often complicated and difficult. Not all the data of trade flows and foreign activity are available and there is the difficulty of different exchange rate effects on import and export, firms have the ability of hedging their exposure and so on. These difficulties can possibly lead to weak results. In addition, it is still not clear which variables to include and how to weight these variables. However, it is clear that it all begins with the efficient market hypothesis. The efficient market hypothesis states that changes in the market value of a firm reflects changes in current or expected conditions that are relevant for the profitability of the firm. The change in exchange rate is added to the market model of returns as a measure of a firm’s exchange rate exposure (Bodnar and Gentry, 1993). As mentioned before, Adler and Dumas (1984) theoretically showed that exposure is best measured with a linear regression equation. So exposure can be investigated with the following equation:

it st

it R

R ₀ ₁  ₍₁₎

Where R is the first difference of the natural logarithm of the ith company’s common _it

stock return and R is the percentual change in the trade weighted exchange-rate. The _st

slope coefficient will correctly measure the effect of the unanticipated changes in exchange rates on stock returns (Jorion, 1990). An extended specification to (1), which explicitly controls for market movements is:

it mt i st it R R R ₀ ₁ ₂  ₍₂₎

Were R is the first difference of the natural logarithm of the market index return. _mt

Including the return on the market means that the coefficient  still reflects the change ₁

in returns that can be explained by changes in the exchange rate, however after conditioning on the market return. If the coefficient  is equal to zero this does not ₁

mean that there is no exchange rate exposure, but that the exposure of the firm is equal to the exposure of the market. Exposure is thus marginal since it is measured relatively to the market average.

Section 4.3 Market Return

The advantage of equation (1) with no market return is that it allows measuring the exposure of all firms as a group, rather than individual firms relative to the country average. The disadvantage of excluding the market return is that it does not distinguish between the direct effects of exchange rate changes and the effects of macroeconomic shocks that simultaneously affect firm value and exchange rates (Dominguez and Tesar, 2005). Due to the importance of the latter argument, regression (2) with market return is used for further investigation.

Section 4.4 Seemingly Unrelated Regression

different sets of exogenous explanatory variables. If there is cross-sectional dependence in the residuals, then SUR will account for these correlations. On the other hand, if there is no cross-sectional correlation, the SUR approach is equivalent to an ordinary least square approach (Williamson, 2001).

Section 4.5 Fixed effects

The question whether one should include fixed or random effects is important as well. To deal with unobserved heterogeneity, one can use fixed or random effects. The

disadvantage of using random effects is the assumption that the independent variables are uncorrelated with the random effects term. Many argue that this is an unrealistic

assumption to satisfy, since unobserved heterogeneity will almost always be correlated with the independent variables. This controversial assumption often makes the fixed effect superior to the random effect model (Beck, 2001; Wilson & Butler 2007). An even more important argument for using fixed effects comes from Petersen (2009). He shows that with including fixed effects, one deals with possible correlation between

observations in the same time period. When residuals are correlated across observations, it is clear that an OLS regression can be biased and either over or underestimates the true variability of the coefficient estimates. With respect to panel data, widely used in

Finance, the way researchers address possible biases in standard errors varies widely and is in many cases incorrect (Petersen, 2009). Petersen (2009) shows in his article that if the firm effect is fixed, the fixed effect removes correlation between observations within the same time period. In this manner, the threat of autocorrelation is removed. This

adjustment gives us the following model:

V.

Data

Section 5.1 Data Characteristics

As already indicated, previous empirical literature did find a relation, but one that is not very strong. From an academic perspective, it is very interesting to examine why there is only a weak relation. Even more interesting and surprising is that Dominguez and Tesar (2005) is one of the only empirical papers that investigate how the amount of exposure is affected by the use of different variables. This paper also focuses more on the use of different variables and hopes to give a better indication of which variables are relevant for explaining the degree of exchange rate exposure. This first sub section gives more details about the data used in this paper. Section 5.2 and 5.3 explain why the specific market return and exchange rate are used, whereas section 5.4 gives more information about the importance of data frequency. In section 5.5 it is tested whether the data are stationaire, while section 5.6 presents the final model.

This sample includes monthly stock returns of European listed firms and only working days are examined. The European stock returns investigated are all European countries that introduced the euro as currency at the first of January 1999. With respect to the data availability, Thomson’s datastream is used. The time span investigated is the first quarter of 1997 up to the third quarter of 2010. The data of the equally weighted and country specific market return is gathered from BRIC DS price index. The nominal trade

weighted exchange rate of the euro is obtained from Thomson’s datastream with the bank of England as source (1990 =100). In the SUR, the changes in the trade weighted

exchange rate, market and stock returns are determined with the help of the first difference of the natural logarithm using the following simple formula:

The descriptive statistics of the first difference of the natural logarithm in market and stock returns are shown in appendix E and F. First, as expected, one can conclude that the means of the stock and market returns are almost equal. Secondly, the standard deviation of the stock returns is higher than the market returns, indicating a higher variance for the country specific firms relatively to their market return. Finally, the mean of the log change in the trade weighted exchange rate is important. Since this mean is positive, one can conclude that on average the euro depreciated against the foreign basket of currencies included in the trade weighted exchange rate. It is also important to know is why specific data is used, which is further explained in the following two subsections.

Section 5.2 Specification Market Return

The first problem is whether to include an equally weighted or value weighted market return. Empirical tests of the CAPM model typically use the value weighted market index to proxy for the market (Dominguez and Tesar, 2001). Bodnar and Wong (2003) argue that when including the value weighted return it first removes the macroeconomic effects. Secondly, they argue that including the value weighted market return removes the more negative effect of exchange rates on larger companies. The latter companies are more likely to experience a higher negative cash flow reaction to domestic currency

appreciations than other firms due to a higher degree of international activity. Bodnar and Wong (2003) argue that including the value weighted return would likely bias tests toward finding no exposure. Dominguez and Tesar (2001, 2005) show that the difference between using a value or equally weighted market return is negligible. Due to the latter two arguments, an equally weighted market return is used for the remaining analysis. Finally, this paper investigates the change in exchange rate exposure after the

Section 5.3 Specification Exchange Rate

trade, it is not possible to create a firm specific exchange rate. Therefore, the nominal trade weighted exchange rate is obtained from datastream. In the empirical literature it is still not clear whether it is best to include the nominal or the trade weighted exchange rate (mostly used in previous literature). The main question of this paper is whether there is a change in exchange rate exposure due to the introduction of the euro in the euro-area. It is not the goal of this paper to indicate which specific currencies were responsible for a possible change in the degree of exposure. Therefore, to keep it synoptic, this paper uses (in line with most previous research) the trade weighted exchange rate.

This paper uses the trade weighted euro exchange rate of the bank of England, which is also used by the IMF. The trade weighted exchange rate measures the local currency relatively to the basket of foreign currencies of trading partners, using the share of trade with each country as weight for that country. The weights reflect the pattern of trade between the euro-area as a whole and countries outside the euro-area. A higher (lower) index means that the euro depreciated (appreciated) against the foreign basket of

currencies included. Important to note is that the trade of countries within the euro-area is excluded, so the weights are solely based on extra euro-area trade. The sterling has the biggest weight, followed by the US dollar. By weighting together the individual exchange rates for the 12 euro-area currencies against the non euro-area currencies, the index is calculated. Thereby making it an effective index for the 12 euro-area currencies as a group. This permits the index to be calculated prior to 31 December 1998, using

“synthetic” euro exchange rates. By geometrically averaging the bilateral exchange rates of the original 11 euro-area currencies and using internal weights based on the country shares of extra euro-area trade, the Bank of England was able to calculate the “synthetic” euro exchange rate2.

2

Section 5.4 Data Frequency

Next, data frequency is important for determining the degree of exposure in an industry. Several papers use monthly stock returns and show that the amount of exposure is increasing in the return horizon, especially when controlling for macroeconomic and market wide capital-market effects (Bartov and Bodnar, 1994; Allayanis, 1997; Bodnar and Wong, 2003). Following Wan (2006), observations over lower data frequency erase more noise and thus explicitly highlight the basic relations between stock values and exchange rates. So daily data is not suitable for investigating exchange rate exposure since it understates the true extend of exposure. Using semi-annual data one should examine a longer period of time to have an acceptable sample size. However, due to policy changes and structural breaks, using a too long time period makes it more difficult to relate specific developments to the amount of exposure found (Dominguez and Tesar, 2005). In line with previous papers, this investigation examines monthly stock returns.

Section 5.5 Unit Root Test

Furthermore, it is tested whether the first difference of the natural logarithm in market returns and trade weighted exchange rate are stationary, with the help of an

Table 6. Unit Root Test of the Country Specific first difference of a natural logarithm in Market Returns

Market Return Austria Belgium Finland France

Statistic Probability Statistic Probability Statistic Probability Statistic Probability

ADF-Test -10.37808 0.0000 -11.11073 0.0000 -10.55933 0.0000 -12.31047 0.0000

Constant - - - -

Trend - - - -

Market Return Germany Ireland Italy Luxembourg

Statistic Probability Statistic Probability Statistic Probability Statistic Probability

ADF Test -12.03316 0.0000 -10.73313 0.0000 -13.26093 0.0000 -10.51511 0.0000

Constant - - - -

Trend - - - -

Market Return Netherlands Spain Portugal Trade Weighted Exchange Rate

Statistic Probability Statistic Probability Statistic Probability Statistic Probability

ADF-Test -12.19575 0.0000 -12.41194 0.0000 -10.91719 0.0000 -12.99941 0.0000

Constant - - - -

Trend - - - -

Table 6 shows the results of the performed unit root tests with and without a trend and intercept3. The Augmented Dickey Fuller test indicates at a 5% significance level that there is no threat of a unit root for the market returns and trade weighted exchange rate examined. Therefore, one can assume stationaire data. Furthermore, one can conclude that there is no need of including a trend in the regression model. Table 7 shows the results of the unit root test for the stock returns.

3

Table 7. Unit Root Tests of the Country Specific first difference of a natural logarithm in Stock Returns. For an overview of the differences in test characteristics one can consult appendix G.

Stock Return Austria Belgium Finland France

Statistic Probability Statistic Probability Statistic Probability Statistic Probability

Levin, Lin & Chu -81.1603 0.0000 -102.512 0.0000 -97.7679 0.0000 -243.184 0.0000

Im, Pesaran and Shin W-stat

-76.0921 0.0000 -98.7414 0.0000 -94.5417 0.0000 -224.634 0.0000

ADF Fisher Chi-square

2895.56 0.0000 5327.74 0.0000 4831.78 0.0000 25206.0 0.0000

PP Fisher Chi-square4

2961.07 0.0000 5640.50 0.0000 5089.22 0.0000 25863.6 0.0000

Stock Return Germany Ireland Italy Luxembourg

Statistic Probability Statistic Probability Statistic Probability Statistic Probability

Levin, Lin & Chu -212.866 0.0000 -57.0594 0.0000 -123.016 0.0000 -40.7918 0.0000

Im, Pesaran and Shin W-stat

-198.235 0.0000 -54.1053 0.0000 -116.038 0.0000 -39.9654 0.0000

ADF Fisher Chi-square

20677.3 0.0000 1817.84 0.0000 7760.03 0.0000 854.449 0.0000

PP Fisher Chi-square

21904.5 0.0000 1975.01 0.0000 8248.33 0.0000 920.116 0.0000

Stock Return Netherlands Spain Portugal

Statistic Probability Statistic Probability Statistic Probability

Levin, Lin & Chu -113.742 0.0000 -102.050 0.0000 -63.9765 0.0000

Im, Pesaran and Shin W-stat

-108.382 0.0000 -95.9681 0.0000 -58.1823 0.0000

ADF Fisher Chi-square

6607.80 0.0000 5114.28 0.0000 1803.97 0.0000

PP Fisher Chi-square

6954.18 0.0000 5214.61 0.0000 1854.42 0.0000

From Table 7 can be concluded, at a 5% significant level, that there is no threat of an unit root for all the stock returns examined. Again, one can accept the alternative hypothesis of stationaire data. In other words, the data is not adjusted since there is no threat of using non-stationaire data. The following section gives more details about the final model used.

4

Section 5.6 The Final Model

To be able to make a distinction between the period before and after the introduction of the euro dummies are included. The final equation used for investigating the changes in exposure due to the introduction of the euro is:

ijt eurot mjt ij mjt ij eurot st j st j ij ijt R R D R R D R ₀ ₁ ₂ ₃ ₄  ₍₆₎

WhereRijt is the return of the ith company’s common stock in country j and R is the st

percentual change in the trade weighted exchange-rate,Rmjt is the rate of return on the

equally weighted market index of country j and D_eurotis a dummy variable that takes the value 1 after 1/5/1998 and 0 otherwise. In this manner, the difference in exchange rate exposure before and after the introduction of the euro can be compared. Based on the discussion of the relevant literature, the following null-hypothesis and alternative hypothesis are tested:

0

H : There is no change in trade weighted euro exchange rate exposure for European

listed firms in the euro-area after the introduction of the euro

1

H : There is a change in trade weighted euro exchange rate exposure for European listed

firms in the euro-area after the introduction of the euro

VI.

Results

Section 6.1 Results

Table 8 . This table reports the ordinary least square estimates of 1j and 2j from ijt eurot mjt ij mjt ij eurot st j st j ij ijt R R D R R D

R ₀ ₁ ₂ ₃ ₄  defined by the regression of Adler and Dumas (1984). Rijtis the return of the ith company’s common stock in country j and R_st is the percentual change

in the trade weighted exchange-rate,Rmjt is the rate of return on the equally weighted market index of country j and D_eurotis a dummy variable that takes the value 1 after 1/5/1998 and 0 otherwise. The table

reports the degree of exchange rate exposure in the euro-area for the whole period examined, which is the first quarter of 1997 up to the third quarter of 2010 (1j), and the change in exchange rate exposure after

the introduction of the euro (2j).56

Degree of Exchange Rate Exposure in the Euro-Area (1j)

Change in Exchange Rate Exposure after the Introduction

of the Euro (2j) Austria -0.340089 (0.289274) -0.066433 (0.300340) Belgium -0.389619*** (0.221433) 0.022725 (0.227220) Finland -0.744228* (0.236307) 0.970268* (0.245095) France 0.293429** (0.113971) -0.483725* (0.119228) Germany -0.767321* (0.194526) 0.506494** (0.199601) Ireland 0.204532 (0.471455) -0.097774 (0.486409) Italy 0.461890** (0.199975) -0.862807* (0.207396) Luxembourg -0.129399 (0.492604) 0.223337 (0.513624) Netherlands 0.037117 (0.364667) -0.625844*** (0.372652) Spain -1.430752* (0.219831) 1.227293* (0.228918) Portugal 0.556027 (0.458282) -1.133470** (0.474366)

Note: The standard errors are in the parenthesis

*Significant at 1% level **Significant at 5% level ***Significant at 10% level

5

Fixed effects are not reported

6

As mentioned in the introduction, the main reason for writing this paper is to investigate whether the monetary union is indeed influenced less by foreign countries after the introduction of the euro. As previously mentioned, a higher (lower) index means that the euro depreciated (appreciated) against the foreign basket of currencies included. With the measurement of exchange rate exposure, a distinction is made between countries within the euro-area.

Section 6.2 Exchange Rate Exposure in Euro-Area Countries

It is now known that foreign exchange rate exposures are traditionally estimated by regressing foreign exchange rate variables on stock returns while controlling for general market movements (Jorion, 1990; Bodnar and Wong, 2003; Bartov and Bodnar, 1994). In the spirit of Adler and Dumas (1984), the incremental or residual effect measured by the exchange rate coefficient is then interpreted as exchange rate exposure. It is important to clearly understand the coefficients of the equation (6) before the results are discussed. The overall degree of exchange rate exposure is given by 1j and the change in exchange

rate exposure after the introduction of the euro is denoted by2j. Table 8 reports the

results of the foreign exchange rate coefficients in the examined period as a whole (1j)

together with their corresponding post-Euro changes in the coefficients (2j), given that

exposures to foreign exchange rate risk have different signs.

research, (Jorion, 1990; Griffin and Stulz, 2001) who show that in most cases not all sample firms or countries show significant levels of foreign exchange rate exposure. In addition, the results are in line with Ottaviano et al. (2007) and (Nardis et al, 2007), who indicate that economic results still differ per country after the introduction of the euro. Remarkable are the findings of negative exposure for Finland, Germany and Spain, which implies that firms experience a decline in firm value (relative to the market) when the euro depreciates. A reduction in firm value after a depreciation of the local currency is against economic intuition. It is, however, consistent with results reported in previous literature. For instance, Jorion (1990) shows that the foreign exchange rate exposures of his sample of 287 US multinational firms are negative. In particular, using the same exchange rate definition, Jorion reports average exposures of -0,234 (1971-1975), -0,079 (1976-1980) and -0,078 (1981-1987). In addition, Bodnar and Wong (2003) found

positive mean and median exposure for all estimation horizon with the inversed exchange rate index.

Even more interesting is to investigate whether the degree of exchange rate exposure changed after the introduction of the euro. From Table 8, one can see that the change in exchange rate exposure following the euro (2j) is of the opposite sign as the degree of

exchange rate exposure beforehand (1j), indicating that the change in exposure is

mean-reverting toward zero. This induces that after the introduction of the euro, there is a net reduction in foreign exchange rate exposure within the euro-area. This is in line with the economic intuition and theory previously discussed. Moreover, it is shown that for all the euro-area countries that have a significant degree of exchange rate exposure at a 5% significance level, the amount of exposure decreased after the introduction of the euro. Furthermore, there is no significant difference between weak EMU countries (Spain, Finland, Ireland, Italy and Portugal) and strong EMU countries (Austria, Belgium, France, Germany, Luxembourg and the Netherlands)7.

7

Finally, several caveats are in order to put the empirical results into perspective. It is argued by Bartram and Karolyi (2006) that derivatives markets continue to grow rapidly, which greatly increases the availability of hedging instruments for firms. In addition, it can be argued that the globalization of today does not lead to larger exposures due to more foreign operations, but that firms may increasingly have the possibility of establishing their production facilities abroad. In other words, low significance and reductions in exchange rate exposure could also be due to firms successfully hedging against exchange rate risk with financial and operating risk management strategies. In fact, Allayannis and Ofek (2001) show that firms using derivatives for hedging have lower levels of foreign exchange rate exposure. Nevertheless, there is little information available about changes in actual hedging activities over time. Moreover, Bartram and Karolyi (2006) argue that there is no information about whether corporate hedging

VII.

Conclusion

Section 7.1 Conclusion

This investigation distinguishes itself from previous research in several ways. First, the paper gives a better overview on how exchange rate exposure is best measured. Since the exact methodology is still not clear, important progress can be made. While financial theory predicts that foreign exchange rate changes affect the value of non financial firms in general, the empirical evidence of foreign exchange rate exposure documented by a sizeable number of empirical studies has been surprisingly weak (Bartram and Karolyi, 2006). Previous research, with the exception of Dominguez and Tesar (2001, 2005), is not aware of the importance of the variables used. For instance, it is argued that if one uses a value weighted market return the macroeconomic effects are removed. In this manner, results are biased towards finding no exposure. The latter could be one of the reasons why previous empirical literature found such weak results. Secondly, this paper investigates exchange rate exposure related to the introduction of the euro. Since relatively little empirical investigations deal with the change in exchange rate exposure due to the euro, it gives new and refreshing insights. One of the only papers that did investigate the change in exchange rate exposure due to the euro, the paper of Bartram and Karolyi (2006), does not deal with several key limitations. First, it only investigates a short time-horizon after the introduction of the euro (up to August 2001). Secondly, Bartram and Karolyi use weekly stock returns. It is already shown by Wan (2006) that observations over lower data frequency erase more noise and thus explicitly highlight the basic relations between stock values and exchange rates. Using weekly data understates the true extend of exposure. This paper investigates the influence of the euro on exchange rate exposure from the first quarter of 1997 till the third quarter of 2010 and uses monthly stock returns.

cross-border mergers, improving price transparency, increasing competition and

eliminating exchange rate risk (Dominguez, 2006). The process of forming an European Union was a difficult task, where stabilizing the different exchange rates within the union was one of the main challenges. It is shown that countries were not always pleased to accept the exchange rate measurements needed to join the monetary union. However, by losing the ability to devaluate the national currency, countries joining the euro would be part of a stronger economic block. In this manner, foreign exchange risk would be

reduced since trade between economic countries with the euro would be more transparent because there is no longer the risk of exchange rate exposure. With the introduction of the euro, trade and intra trade increased significantly (ECB, 2008). This makes it difficult to predict the exact effect of the euro on exchange rate exposure beforehand, since from a theoretical point of view more intra euro-area trade decreases the degree of exchange rate exposure whereas more non euro-area increases the degree of exchange rate exposure. Furthermore, it is argued whether the increase in trade is fully due to the introduction of a single currency.

Despite the increase in intra trade, non euro-area trade, (weak) signals of price

convergence and proved decrease of foreign exchange rate exposure, the continuation of the European monetary union is still under discussion. Due to the recent economic crisis and instability of today, it is argued that the euro-area should again make use of their old currencies. Countries such as Greece and Ireland already needed financial support, while the financial position of Spain, Portugal and Italy is unstable. The IMF already

announced that the emergency fund of the Eurozone is insufficient to cope with the financial problems. Andrew Bosomworth, top manager of worlds’ largest obligation investment company Primco, already announced in the German paper ‘Die Welt’ on the 20th of January 2011 that the only possibility for Greece, Portugal and Ireland to pay off its debts is to leave the Eurozone. Only with an own currency they will be able to stimulate export and boost their domestic economy.

Section 7.2 Limitations and Recommendations

Despite the contributions of this paper, already discussed in previous sections, I also have to acknowledge several key limitations. The first limitation is that, due to the limited availability of stock returns in datastream in the examined period, this investigation only investigates a limited number of firms. This holds especially for Luxembourg, Portugal and Ireland, were country specific results are respectively based on 10, 21 and 24 company returns. Including more firms would improve the explanatory power of the paper. In total, this paper investigates the stock returns of 969 firms. The second

limitation is, as already mentioned, that the findings of low significance and decrease in foreign exchange rate exposure can be due to an increase of firms using derivatives. It could be that more firms are hedging over time and the euro dummy variable may pick up the hedging effect instead of a euro effect. Up to now there is little evidence for the latter, however more research is needed to completely reject the possibility of a reduction in foreign exchange rate exposure after the euro due to the increased use of hedging instruments. Another possible extension for further research is the focus on the

determinants of exchange rate exposure with respect to firm characteristics. This paper already showed that firm characteristics are important for explaining the degree of exposure. However, due to data limitations, this paper did not focus on how firm specific characteristics are related to the degree of exposure found. Finally, one could also

examine why and how results change when using different time horizons

References

Adler, Michael and Bernard Dumas, 1984, Exposure to Currency Risk: Definition and Measurement, Financial Management, 13, PP 41-50.

Allayannis, George S. and Eli Ofek, 1997, Exchange Rate Exposure, Hedging, and the Use of Foreign Currency Derivatives, Journal of International Money and Finance, 20, PP 273-296.

Allayannis, George S., 1997, The Time-Variation of the Exchange-Rate Exposure: An Industry Analysis, Unpublished University of Virginia working paper

Allington, Nigel F.B., Kattuman, Paul A. and Florian A. Waldman, 2005, One Market, One Money, One Price?, International Journal of Central Banking, 12, PP 73-115.

Bartov, Eli and Grodon M. Bodnar, 1994, Firm valuation, earnings expectation, and the Exchange-Rate Exposure Effect, The journal of Finance, 44, PP 1755-1785.

Bartram, Söhnke M. and Andrew G. Karolyi, 2006, The Impact of the Introduction of the Euro on Foreign Exchange Rate Risk Exposures, Journal of Empirical Finance, 13, PP 519-549.

Beck, Nathaniel L., 2001, Time-Series Cross-Section Data: What Have We Learned in the Past Few Years?, Annual Review of Political Science 4, PP 271-93.

Bilson, John F.O., 1981, The “Speculative Efficiency” Hypothesis, The Journal of

Business, 54, PP 435-451.

Blanchard, Oliver J. and Lawrence H. Summers, 1994, Perspectives on High World Real Interest Rates, Brookings Papers on Economic Activity, 2, PP 273-334.

Bodnar, Gordon M., Dumas, Bernard and Richard C. Marston, 2002, Pass-through and Exposure, The Journal of Finance, 57, PP 199-231.

Bodnar, Gordon M., Hayt, Gregory S., and Richard C. Marston, 1998, Wharton Survey of Financial Risk Management by US Non Financial Firms, Financial Management, 27, PP 70-91.

Bodnar, Gordon M. and William M. Gentry, 1993, Exchange rate exposure and industry characteristics: Evidence from Canada, Japan and the USA, Journal of International

Money and Finance, 12, PP 25-49.

Bodnar, Gordon M. and Franco M.H. Wong, 2003, Estimating Exchange Rate Exposures: Issues in Model structure, Financial Management, 32, PP 35-67.

Bris, Arturo, Koskinen, Yrjö and Mattias Nilsson, 2003, The Euro and Corporate Valuations, Yale School of Management, Working Paper

Brown, Gregory W., 2001, Managing foreign exchange risk with derivatives, Journal of

Financial Economics, 60, PP 401-448.

De Jong, Abe, Ligterink, Jeroen and Victor Macrae, 2002, A Firm-specific analysis of the Exchange-rate Exposure of Dutch Firms, Financial Management, 24, PP 77-88.

De Nardis, Sergio, De Santis, Roberta and Claudio Vicarelli, 2007, The Single Currency’s Effects on Eurozone Sectoral Trade: Winners and Losers?, Economics

E-journal, Discussion Paper No. 2008-1

Doidge, Craig, Griffin, John and Rohan Williamson, 2002, Does exchange rate exposure matter?, Working Paper

Doidge, Craig, Griffin, John and Rohan Williamson, 2006, Measuring the economic importance of exchange rate exposure, Journal of empirical finance, 13, PP 550-576.

Dominguez, Kathryn M.E., 2006, The European Central Bank, The Euro, and Global Financial Markets, Journal of Economic Perspectives, 20, PP 67-88.

Dominguez, Kathryn M.E. and Linda L. Tesar, 2001, A Reexamination of Exchange Rate Exposure, American Economic Review, 91, PP 396-399.

Dominguez, Kathryn M.E. and Linda L. Tesar, 2005, Exchange Rate Exposure, Journal

of International Economics, 68, PP 188-218.

Dominguez, Kathryn M.E., 2006, The European Central Bank, the Euro, and Global Financial Markets, Journal of Economic Perspectives, 20, PP 67-88.

Dumas, Bernard, 1978, The Theory of the Trading Firm Revisited, Journal of Finance,

33, PP 1019-1029.

European Union, 1985, Completing the internal market: White Paper from the Commission to the European Council, Milan, 28–29 June 1985