finance 3

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=rejf20

The European Journal of Finance

ISSN: 1351-847X (Print) 1466-4364 (Online) Journal homepage: https://www.tandfonline.com/loi/rejf20

Exchange rate risk exposure and the value of European firms

Fabio Parlapiano, Vitali Alexeev & Mardi Dungey

To cite this article: Fabio Parlapiano, Vitali Alexeev & Mardi Dungey (2017) Exchange rate risk exposure and the value of European firms, The European Journal of Finance, 23:2, 111-129, DOI: 10.1080/1351847X.2015.1072570

To link to this article: https://doi.org/10.1080/1351847X.2015.1072570

Published online: 19 Aug 2015.

Submit your article to this journal

Article views: 799

View related articles

View Crossmark data

Citing articles: 4 View citing articles

The European Journal of Finance, 2017 Vol. 23, No. 2, 111–129, http://dx.doi.org/10.1080/1351847X.2015.1072570

Exchange rate risk exposure and the value of European firms

Fabio Parlapianoa,b∗, Vitali Alexeevb and Mardi Dungeyb,c aManagement Department, Sapienza University of Rome, Rome, Italy; bTasmanian School of Business and Economics, University of Tasmania, Hobart, Australia; cCenter for Financial Analysis & Policy, University of Cambridge Judge Business School, Cambridge, England

(Received 25 March 2013; final version received 23 December 2014)

This paper presents a new assessment of the exposure of European firms to exchange rate fluctuations which takes into account the potential common drivers of exchange rates and equity market conditions. Using monthly data for European firms from 1999 to 2011, we assess the impact of unexpected fluctua- tions in the USD, JPY, GBP and CHF against the Euro, and show that the proportion of firms subject to exchange rate risk is considerably larger when estimation accounts for potential common drivers and firm- specific factors than otherwise. Firm exposure to exchange rate risk is affected by the level of international involvement, industry, firm size and country of origin. European firms with largely domestic operations reveal the greatest vulnerability to unexpected exchange rate movements, suggesting an opportunity to improve risk management for these companies.

Keywords: exchange rate risk; firm value; European firms

AMS Subject Classifications: G32; F31; F23; G15

1. Introduction

Currency volatility is a leading concern for companies around the world; risk managers consider currency fluctuations as amongst the top three risk factors, dominating overall economic and mar- ket risk (Economist Intelligence Unit 2011). Currency fluctuations may affect firm value via cash flows, and although firms often use hedging strategies, currency exposure is not completely elim- inated. The existing literature reports weak, and mixed, evidence on the significance of exchange rate risk; see, for example, for the USA, Jorion (1990), Amihud and Levich (1994) and Bodnar and Wong (2003) and, for international evidence, Bodnar and Gentry (1993) and Dominguez and Tesar (2001). The majority of the existing evidence considers US data for non-financial firms and employs an estimation strategy of regressing individual firm or portfolio returns on the con- temporaneous exchange rate change and an indicator of market portfolio return – an augmented capital asset pricing model (CAPM) style model promoted in Jorion (1990) and subsequent work. However, Doukas, Hall, and Lang (2003) convincingly argues that at least part of the lack of evidence for the impact of exchange rate changes on individual firms arises from the commonly implemented modelling assumption that aggregate stock markets are independent from exchange

∗Corresponding author. Email: [email protected]. Current address: Risk Management Department, Bank of Italy, Rome, Italy.

© 2015 Taylor & Francis

112 F. Parlapiano et al.

rates; see also Priestley and Ødegaard (2007). There is significant evidence that aggregate stock markets and exchange rates are related, albeit exchange rates may exert lower influence on stock markets than other factors; see, for example, Roll (1992) and Granger, Huang, and Yang (2000); a theoretical model in Hau and Rey (2006) and recent evidence on bi-directional Granger causality between exchange rate and stock market returns in Inci and Lee (2014).

This paper considers the impact of exchange rate fluctuations on European firms for the period 1999–2011 using the Doukas, Hall, and Lang (2003) orthoganalized model approach. European firms have not previously been examined in a framework accounting for the com- mon forces driving exchange rates and equity market changes. We extend the existing literature on European firms, summarized in Table 1, with the use of firm-level data from all sectors of the economy, including financial and non-financial firms, and firms which are active in international markets and those concerned only with domestic markets. Firms which solely con- centrate on domestic markets are often omitted from analysis, but recently Amiti, Itskhoki, and Konings (2014) provide both a theoretical framework and empirical evidence (from Belgium) that these firms are likely to be considerably more exposed to exchange rate fluctuations than their more internationally active counterparts. Our findings confirm this evidence for Europe as a whole. We find evidence that a larger fraction of firms are affected by exchange rate fluctu- ations in the value of the Euro against other currencies than identified under the independence assumptions used in the Jorion (1990) framework. For example, in the case of fluctuations of the value of the JPY against the Euro the fraction of firms almost doubles to 27.60% using the orthoganalized model compared with the 15.88% identified with the Jorion (1990) method. Similar results apply for fluctuations in the value of the USD, GBP and CHF against the Euro.

The impact of exchange rate fluctuations may also vary with firm characteristics, and we make a thorough exploration of the role of industry sector, firm size and level of international activity as potential determinants. In addition, we consider the relevance of the country of origin of the firm, reflecting the potential importance of sovereignty within a currency union. Although it is well documented that the European common currency area resulted in reduced exchange rate risk for firms in member countries, particularly for those with foreign business activity (Bartram and Karolyi 2006; Koutmos and Knif 2011), significant risk remains (Adjaouté and Danthine 2004; Muller and Verschoor 2006a,b; Hutson and O’Driscoll 2010) even after accounting for improved risk management practices via hedging (Nguyen, Faff, and Marshall 2007). We find that the industry sector and country of origin of a firm are important determinants of the exchange rate exposure for an individual firm. The financial sector is more exposed to unexpected exchange rate fluctuations, and countries that became deeply embroiled in the European sovereign debt crises from 2010 onwards had greater exposure, particularly affecting firms in Portugal, Ire- land, Greece and Spain. We also find that the relationship between the degree of response to unexpected market shocks and unexpected exchange rate shocks is not monotonic. Defensive securities tend also to be those with the lowest response to exchange rate risks. Firms below the 50th percentile of estimated beta display a positive relationship between increasing market risk and increasing exchange rate risk. However, after the 50th percentile (β = 0.8), the sensitivity of firms to exchange rate risk tends to decline with increasing market risk. That is, aggressive securities are not necessarily as exposed to exchange rate fluctuations than their more defensive counterparts.

The paper proceeds as follows. Section 2 outlines the methodology for empirically detecting exchange rate risk for individual firms. Section 3 describes the newly collected data set. The results are presented and discussed in Section 4, and Section 5 concludes.

T h e

E u ro

p ea

n Jo

u rn

a l

o f

F in

a n ce

113

Table 1. Summary of empirical literature on exchange rate exposure for the Eurozone.

Authors Exchange rate indexa

Exchange rate exposure

Financial marketsb

Sampled firms Data period Frequency Market index

% Significantly exposedc

Muller and Ver- schoor (2006a)

Cross USD, GBP, JPY

Contemp. Eurozone (20) 817 non- financial MNCs

1988–2002 1;4;12;54 weeks

VW EMU 13%JPY; 22% GBP; 14% USD;

Bartram and Karolyi (2006)

TW ER index Contemp. Eurozone (18), US, JP

3220 non- financial

1990–2001 Weekly Market indices for each country

7.3% negative, 4.7% positive

Nguyen, Faff, and Mar- shall (2007)

French TW ER index

Contemp. France 99 non- financial MNCs

1996–2000 Monthly France CAC 40 30.30%

Hutson and O’Driscoll (2010)

TW ER index Contemp. Eurozone (7) and non- Eurozone (4)

1154 1990–2008 Monthly DataStream weighted for each country

8.4% pre-Euro; 9.1% post-Euro

Koutmos and Knif (2011)

Cross USD Contemp. Finland 6 industry and 4 size portfolios

1992–2006 Weekly NASAQ OMX Helsinki market index

4 industry and 4 size portfolios pre-Euro; 2 industry portfolios post-Euro

Inci and Lee (2014)

TW ER index Lagged France, Germany, Italy

MNCs 1984–2009 Yearly MSCI country indices

5 out of 8 country portfolios

This study Cross USD, GBP, JPY, CHF

Contemp. Eurozone (15) 650 financial and non- financial, MNCs and domestic

1999–2011 Monthly Euro Stoxx TMI and Euro Stoxx 50

StoxxTMI 27.60% JPY; 13.04% USD; 10.78% GBP. Stoxx 50 42.86% JPY; 26.53% USD; 20.41% GBP

aAbbreviations used: trade weighted (TW); value weighted (VW); exchange rate (ER); cross denotes cross-rates of (the Euro) with other currencies. bFigure in parentheses denote the number of countries in the study. cWe report the main result from each study. Figures represent the percentage of firms significantly exposed to fluctuations of exchange rates considered.

114 F. Parlapiano et al.

2. Modelling framework

A model commonly used in the literature considers the direct regression of contemporaneous exchange rate changes on individual firm stock returns, controlling for market conditions via a common market indicator. That is, an augmented CAPM model is as follows:

rit = β0i + β1irst + β2irmt + μit, (1) where the stock return, rit, for a firm i is regressed on the contemporaneous exchange rate return, rst, and the market portfolio return, rmt. The loading β1i represents the exposure of the individual firm to exchange rate movements having conditioned on the broad equity market. When β1i is positive (negative), a rise in the exchange rate increases (decreases) the value of the stock for the individual firm in excess of market fluctuation.

A problem with the specification given in Equation (1) is that it is predicated on the exchange rate movements and market conditions being uncorrelated. As there is significant evidence that aggregate equity market returns are related to exchange rate movements via common factors (see, for example, Phylaktis and Ravazzolo (2005)) this means that Equation (1) is subject to collinearity, with consequent unknown bias in the estimated coefficients.

An alternative is to apply the three step procedure of Doukas, Hall, and Lang (2003) which orthogonalizes the market and exchange rate risk factors; see also He, Ng, and Wu (1996). The first step removes from the exchange rate returns the effect of common macroeconomic funda- mental influences which might be expected to influence both equity markets and exchange rates. There is abundant evidence linking indicators such as aggregate activity, inflation, employment, interest rates and other indicators to both equity market returns and exchange rates; for example, Chordia and Shivakumar (2002) and Flannery and Protopapadakis (2002) on macro fundamen- tals in forecasting and announcement effects on stock returns; Beckmann, Belke, and Kuhl (2011) for recent evidence on macro variables and the Euro and Taylor (1995) on exchange rate mod- els. Regressing these control variables on the exchange rate returns, we obtain the unexpected exchange rate returns from

rst = β0 + 6∑

j=1 βjCVj,t−1 + β7rst−1 + εst, (2)

where CVj,t−1 are lagged control variables and rst−1 controls for any autocorrelation in exchange rate returns. The unexpected exchange rate returns are the fitted residuals from Equation (2), that is ε̂st. The second stage orthogonalizes equity market returns to the same set of common control variables and the unexpected exchange rate returns from step (2), while also controlling for autocorrelation in the market returns as follows:

rMmt = β0 + 6∑

j=1 βjCVj,t−1 + β7rMmt−1 + β8ε̂st + eMmt, (3)

where rMmt is the stock market return. The orthogonalized indicator of the market returns will be

given by the estimated residuals, êMmt. There remains the possibility that exchange rates and market returns retain a common factor

through some alternate effect which has not been captured in the model, something which can be addressed with latent factor models or principal components but with the accompanying dis- advantage of being unable to observe the contributing feature. For example, Doukas, Hall, and

The European Journal of Finance 115

Lang (2003) include Fama–French factors in their analysis (Fama and French 1992, 1993). While these factors have been successfully associated with returns premia around the world, they dif- fer across markets. Dimson, Marsh, and Staunton (2002) document country-specific differences among the European countries both for size and growth premia, hence a unique measure of these factors cannot be applied at the Eurozone level.

Finally, the third stage regression estimates the sensitivity of individual firm stock returns to the unexpected exchange rate and unexpected market returns as follows:

rit = αi,0 + 6∑

j=1 βijCVij,t−1 + βi7rit−1+βimêMmt + βisε̂st + γit, (4)

where we now consider the effect of the economic factors and the unexpected components of the exchange rate and market returns. The coefficient βis represents the sensitivity of the returns for an individual firm, rit, to unexpected exchange rate movements.

3. Data

Focusing on firms with operations in the Eurozone, we consider all constituents of a broad market index, the Euro Stoxx TMI (600 firms), and constituents of a large capitalization index, the Euro Stoxx 50 (50 firms) for the period 1999–2011. Both indices have diverse coverage of the Eurozone countries and sectors, with an economic performance benchmark of the Eurozone in the former, and a large capitalization focus and value weighted selection criteria in the latter.1

The large number of firms in the sample allows for a breakdown based on geographical location, industry and level of international involvement. Monthly data used in this study are obtained from Bloomberg and Stoxx Ltd.

Two main sources of biases related to the inclusion of large stocks in the market index have been discussed in the literature. The size (or positive) bias affects the estimation of exchange rate exposure due to higher proportion of exports of large firms (Bodnar and Wong 2003). The success bias, instead, inflates the market risk premium estimation when stocks included in value weighted indices experienced a sustained growth path (Dimson, Marsh, and Staunton 2002). We are able to account for the mentioned size and success biases by the use of a broad market index such as the Euro Stoxx TMI.

In order to account for the firm’s level of involvement in international operations or, more precisely, non-Eurozone operations, we construct the ratio of each firm’s non-Eurozone revenues to total revenues, hereafter the Foreign Exchange Exposure (FEE) index. The average degree of international involvement in period from 2005 to 2010 is higher for Euro Stoxx 50 constituents (at 46%) than for Euro Stoxx TMI (at 39.5%). On average, large European firms make approxi- mately half of their revenues outside the Eurozone countries. The FEE index for Euro Stoxx 50 constituents is more homogeneous than for Euro Stoxx TMI constituents, which is not surprising given the wide industry coverage that characterizes the TMI.

We group the firms by their degree of international involvement on the basis of the FEE index threshold as in Doukas, Hall, and Lang (2003). We consider high exporters (or MNCs), low exporters, and non-exporters (or domestic firms). The highest concentration of MNCs (high exporters) is in the industrial sector (73%) followed by utilities (50%) and real estate (37.5%). On the other side of the spectrum, the proportion of domestic firms is highly concentrated in non-industrials, that is financials, banking, and insurance sectors.

116 F. Parlapiano et al.

Monthly data for the set of economic variables and nominal bilateral exchange rates are obtained from the ECB Statistical Data Warehouse. We adopt the indirect exchange rate quotation from the point of view of the Eurozone, therefore, positive exchange rate returns imply apprecia- tion of the EURO. To be completely clear, this implies that we express all exchange rates in terms of the foreign currency cost of one Euro; thus when the value of the Euro falls, the exchange rate falls. Following Chen, Roll, and Ross (1986) and Doukas, Hall, and Lang (2003), we rely on six control variables to express macroeconomic conditions: unexpected inflation (UI), industrial production (IP), term premium (TP), money supply (MS), interest rate spread (IRS), and trade balance (XM ). With the exception of the UI, all of the control variables are selected and computed following Chen, Roll, and Ross (1986) and Doukas, Hall, and Lang (2003) specifications.2

4. Results

Table 2 reports the results of the first stage regressions for all four currencies. The currency pairs are only weakly influenced by the set of macroeconomic fundamentals. Only the first lag of the exchange rate return (in the case of JPY and USD), the Euro area money supply (in the case of USD), unexpected inflation (in the case of CHF) and interest rate spread (in the case of CHF) are significant determinants of exchange rate changes. While the joint hypothesis of zero coefficients for all explanatory variables is rejected for the Euro exchange rates against the JPY, USD and CHF, the Euro exchange rate with the GBP is found to be independent of all the macroeconomic variables considered. The poor performance of macroeconomic fundamentals in forecasting short run exchange rates is consistent with existing literature. Meese and Rogoff (1983) first pointed to the random walk behavior of exchange rates, but subsequent work of Engel and West (2005) and Andersen, Tim Bollerslev, and Vega (2003) point to the existence of some, although weak, links between economic fundamental news and exchange rates.

The estimate of the unexpected exchange rate return from stage 1 is incorporated into the second stage regression for the equity market returns proxies, Euro Stoxx TMI or Euro Stoxx 50, also reported in Table 2. The Eurozone stock market is only weakly influenced by the set of macroeconomic fundamentals, but in each case the models reject the joint hypothesis of no influence of all macroeconomic variables. There is some evidence that realized inflation is nega- tively correlated with stock market returns, consistent with the inflation puzzle of Fama (1981); see also Geske and Roll (1983) and Flannery and Protopapadakis (2002) for later evidence.

The influence of the unexpected exchange rate movements in the Euro against the JPY and CHF, generated from the first stage regressions, on the stock market returns is found to be signif- icant and positive for both stock market proxies. This result is consistent with those of Doukas, Hall, and Lang (2003) for Japanese firms, but differs from the earlier studies of Chen, Roll, and Ross (1986) and Hamao (1988) for US and Japanese markets, respectively. In the cases of the other currencies, GBP and USD, there is no significant effect from the generated unexpected exchange rate changes.

The effect of unexpected exchange rate changes on individual stock returns is the primary interest of this paper. The third stage regression considers this via the impact of the generated orthogonal unexpected exchange rate change and the market return residuals on the individual stock returns, as per Equation (4).

Table 3 summarizes the heteroskedasticity robust ordinary least-squares (OLS) estimates of the exchange rate exposure for Euro Stoxx TMI and Euro Stoxx 50 constituent stocks, respec- tively. We report the percentage of firms significantly affected by unexpected exchange rate

T h e

E u ro

p ea

n Jo

u rn

a l

o f

F in

a n ce

117

Table 2. Results of the first and second stage regressions in the orthogonalized market model.

Second stageb

First stagea Euro Stoxx TMI Euro Stoxx 50

Variables JPY USD GBP CHF JPY USD GBP CHF JPY USD GBP CHF

Intercept 0.02 0.02 0.00 0.00 0.07** 0.02 0.02 0.05 0.06* 0.01 0.01 0.04 Unexpected

inflation UIt−1 − 3.72 − 4.38 2.88 − 5.28** − 21.60** − 14.72 − 24.32*** − 22.95*** − 22.64** − 13.60 − 25.07*** − 24.32***

Industrial production

IPt−1 0.14 0.15 − 0.11 0.05 0.13 − 0.09 − 0.01 0.12 0.24 − 0.01 0.09 0.28

Term premium TPt−1 − 0.81 − 0.24 − 0.22 − 0.26 − 0.76 0.66 0.65 0.05 − 0.46 0.70 0.68 0.11 Money supply MSt−1 − 0.28 − 0.57** 0.10 0.14 − 0.17 − 0.32 − 0.46 − 0.19 − 0.17 − 0.36 − 0.52 − 0.22 Interest rate

spread IRSt−1 0.28 − 0.18 − 0.18 − 0.58** 0.78 0.81** 1.46*** 0.39 0.47 0.95** 1.67*** − 0.01

Trade balance XM t−1 0.09 0.05 − 0.02 0.02 0.00 0.06 − 0.12 − 0.02 − 0.03 0.06 − 0.15 − 0.06 Lagged exchange

rate return rst−1 0.25*** 0.25*** 0.11 − 0.07

Lagged market return

rMmt−1 0.06 0.03 0.03 0.02 0.04 − 0.01 − 0.01 − 0.03

Unexpected exchange rate

εst 0.43*** 0.21 0.27 1.18*** 0.44** 0.23 0.28 1.21***

F-statistic 2.73** 3.25*** 1.67 2.22** 2.88*** 2.49** 3.16*** 3.33*** 2.49** 2.45** 3.23*** 3.05*** R2 adjusted 0.07 0.09 0.03 0.10 0.09 0.07 0.10 0.16 0.07 0.07 0.11 0.15 JB-statistic 12.55*** 59.31*** 24.46*** 184.32*** 27.39*** 33.62*** 35.47*** 34.38*** 27.37*** 31.79*** 29.39*** 34.19***

Notes: Estimated using monthly data from 1999 to 2011. The JB-statistic reports the result of the Jacque–Bera normality test on the residuals. aFirst stage results are from Equation (2). bSecond stage results are from Equation (3). *Significance at 10%. **Significance at 5%. ***Significance at 1%.

118 F. Parlapiano et al.

Table 3. Exchange rate exposure of European firms (orthogonalized market model).

Euro Stoxx TMI Euro Stoxx 50

JPY

Currency trend − 12.05% Average β

(a) s 0.82 0.67

Total significance 146 (27.60%) 21 (42.86%)

(N +)β(b)s 142 20 (N −)β(c)s 4 1

USD

Currency trend 30.80% Average βs 0.68 0.59 Total significance 69 (13.04%) 13 (26.53%) (N +)βs 62 12 (N −)βs 7 1

GBP

Currency trend 30.85% Average βs 0.70 0.95 Total significance 57 (10.78%) 10 (20.41%) (N +)βs 48 10 (N −)βs 9 0

CHF

Currency trend − 24.38% Average βs 1.88 1.64 Total significance 279 (52.74%) 33 (67.35%) (N +)βs 279 33 (N −)βs

Notes: Equation (4) estimated using monthly data from 1999 to 2011. Heteroskedasticity robust OLS estimates of the exchange rate exposure, βis , are reported along with the percentage of firms significantly affected by exchange rate fluctuations, the average magnitude of significant exposure coefficients and the number of firms positively, N +, and negatively, N −, affected by currency variations. (a)Average of significant βis from Equation (4) on page 8. (b)Number of significant and positive βis from Equation (4). (c)Number of significant and negative βis from Equation (4).

fluctuations, the average magnitude of the coefficients, βis and the number of firms positively and negatively affected by currency variations.

The first panel of Table 3 reports the results for unexpected changes in the exchange rate of the JPY against the Euro. Just over 27% of the Euro Stoxx TMI constituents have a significant response to unexpected changes in the value of the Yen against the Euro, with an on-average appreciation of 8.2% in the stock returns for each 10% appreciation of the Euro. The impact of an appreciation in the Euro against the USD and GBP is also positive at an average of around 7% (the second and third panels of Table 3), but the proportion of firms with significant impact is fewer than half that in the Japanese case.

An astonishing 52% of firms are significantly affected by CHF fluctuations with as many as 72% of affected firms among the financial firms with an overall average appreciation of 18.8% in the stock returns for each 10% appreciation of the Euro against the CHF. This large proportion of

The European Journal of Finance 119

European firms exposed to the Swiss Franc may reflect a speculative demand of the CHF, in the form of foreign debt (see Keloharju and Niskanen 2001), which might be driven by the interest rate differentials between the Eurozone and Switzerland making it convenient to borrow funds denominated in Swiss Francs (see Brunnermeier, Nagel, and Pedersen 2008; Galati, Healt, and Guire 2007 on the role of the CHF and JPY as global funding currencies).

The increase in stock returns in response to an appreciation of the Euro is consistent with results reported in Muller and Verschoor (2006b): the authors find that appreciating Euro against foreign currencies had a positive effect on the returns of European stocks. They argue that Euro- pean firms are mainly net-importers and thus benefit from a strong Euro which makes both domestic consumption and exports of products, including materials supplied from overseas, more competitive. The results in our paper differ from Campbell, Medeiros, and Viceira (2010) which report a negative correlation between the European stock market and the Euro exchange rate for the period from 1975 to 2005 from the perspective of an international investor. These results are not inconsistent with ours, as Campbell, Medeiros, and Viceira (2010) use the US dollar as the reference currency, whereas we use the Euro, so that in comparing directly the results for the exchange rate between the Euro and the US dollar the results should be inverted. Indeed where Campbell, Medeiros, and Viceira (2010) find a negative relationship between the exchange rate expressed as the number of Euros per US dollar and their proxy for European stock markets (comprising a portfolio that includes German, French, Italian and Dutch stocks), we find a pos- itive relationship between the number of US dollars per Euro and the Euro Stoxx TMI. These results are directly consistent, as our exchange rate is a direct inversion of the quotation con- vention adopted in Campbell, Medeiros, and Viceira (2010). There are important differences between the research question and approach in Campbell, Medeiros, and Viceira (2010) and this paper. Campbell, Medeiros, and Viceira (2010) consider excess returns on equity portfolios and currencies, whereas we adopt a procedure to procure unexpected (as opposed to excess) returns; to check our results we conducted data transformations following Campbell, Medeiros, and Viceira (2010) to obtain excess returns in the Euro exchange rate against the US dollar and the Euro Stoxx and MSCI World index and obtained the same (negative) sign on the correlations between these series as in Campbell, Medeiros, and Viceira (2010) for the total sample period. Thus, our results should not be interpreted as inconsistent with those of Campbell, Medeiros, and Viceira (2010), but have a common underlying result which is then used to consider somewhat different aspects of the empirical relationship (in their case hedging in international investor port- folios and in ours the impact of exchange rate fluctuations on the equity value of individual firms involved in different underlying lines of business).

The percentage of firms with a significant response to unexpected exchange rate fluctuations in the Euro Stoxx 50 sample is considerably higher than for the Euro Stoxx TMI (almost 43% for the JPY, 26.5% for the USD, 20.41% for the GBP and 67.35% for the CHF). Given that the Euro Stoxx 50 comprises the largest MNCs in Eurozone this result is unsurprising, although one cannot directly compare the estimated βis across the two tables as the market indices for which they are constructed differ. Table 3 reports that for the Euro Stoxx 50 constituents, the highest average impact is from unexpected moves in the Euro against the CHF at almost 16.4% for every 10% exchange rate variation, followed by the GBP (9.5%), JPY (6.7%) and USD (5.9%).

4.1 The role of firm characteristics

A firm’s level of international involvement is a recognized determinant of its exchange rate risk exposure. However, the sign, significance and magnitude of the exposure coefficients

120 F. Parlapiano et al.

are ambiguous. Choi and Jiang (2009) found evidence that the multinationality affects a firm’s exchange exposure, however, this effect is not consistent with theoretical predictions. Empirically, exchange rate risk exposures are smaller and less significant for MNCs than for non- multinationals and this evidence may be the outcome of operational hedging which decreases a firm’s exchange risk exposure and increases its stock returns. According to Davies, Eckberg, and Marshall (2006), Bartram (2008) and Bartram, Brown, and Fehle (2009), there is strong evidence indicating that firms with a higher proportion of international sales are more likely to hedge FEE. Nevertheless, it is plausible that the effectiveness of these financial and operational hedging strategies may be incomplete when future cash flows are considered. While hedging strategies effectively mitigate current exposure in the short run (i.e. the transaction exposure), the effectiveness of these hedging practices is weak in the long run (i.e. the economic exposure). An alternative explanation offered by Amiti, Itskhoki, and Konings (2014) provides a theoretical framework where the typical large importer is also a large exporter and hence manages both sides of any exchange rate fluctuation, so that less internationally active firms are likely to experience more exchange rate pass-through to prices than their more internationally focused counterparts.

Table 4 reports the exposure coefficients grouped by the level of international involvement as approximated by the FEE index. A clear pattern emerges: low exporters and domestic firms are more sensitive to unexpected exchange rate fluctuations than MNCs. In fact, the magnitude of the average exposure coefficient of low exporters and domestic firms is greater, confirming the findings in Davies, Eckberg, and Marshall (2006) and Choi and Jiang (2009). This result does not support the theoretical premise that the greater the firm’s level of international involvement, the greater the impact of currency fluctuations on firm’s market value. In contrast, our results highlight the relevance of the exchange rate exposure to firms which do not report operations in foreign currencies and the need for a better control of this risk. These results are consistent with Amiti, Itskhoki, and Konings (2014), who provide Euro relevant evidence for reduced exchange rate pass through for internationally active firms from Belgian data. Coupled with risk manage- ment policies, these effects may explain why MNCs have less stock price response to exchange rate exposure than low exporter firms.

The breakdown by the level of international involvement, given in Table 4, displays two additional features: the combined proportion of low exporters and domestic firms significantly affected by fluctuations of the Euro is larger than MNCs in the case of USD and GBP, while this picture is reversed in the case of JPY and CHF. One possible explanation is that MNCs concen- trate on hedging against risks from only their main trading partners’ currencies (e.g. USD and GBP), while the CHF and JPY are simply overlooked. Another possible explanation could stem from the use of foreign debt by European firms and the ability of these firms to raise capital by borrowing from Swiss and Japanese institutions with unhedged loans denominated in JPY and CHF; traditionally low yielding currencies (Brunnermeier, Nagel, and Pedersen 2008; Galati, Healt, and Guire 2007; Hattori and Shin 2009).

When controlling for the level of international involvement, large capitalization firms are more likely to be exposed to exchange rate risk. Consider, for example, that 43% of Euro Stoxx 50 MNCs are significantly affected by exchange rate fluctuations compared with 27% of Euro Stoxx TMI MNCs (see Table 4).

Most of the existing literature has focused on industrial firms, assuming that the exposure of financial firms may be driven by different aims and factors; particularly, the possibility of taking advantage of better forecasts of future exchange rates by financial institutions. Here we also include financial firms in our analysis. Estimates in Table 5 confirm that firms within the financial industry experienced a much larger positive impact of exchange rate fluctuations than

T h e

E u ro

p ea

n Jo

u rn

a l

o f

F in

a n ce

121

Table 4. Exchange rate exposure of European firms with a breakdown by international involvement.

JPY USD GBP CHF

MNCs Low export. Dom. MNCs Low export. Dom. MNCs Low export. Dom. MNCs Low export. Dom.

Euro Stoxx TMI

Sample 340 82 71 340 82 71 340 82 71 340 82 71

Total significance

27% 18% 24% 10% 16% 13% 10% 9% 13% 54% 40% 52%

Average β (a) s 0.77 0.87 0.85 0.47 1.00 0.76 0.47 1.10 0.91 1.90 1.90 1.72

(N +)β(b)s 87 15 17 27 13 9 24 7 9 182 33 27 (N −)β(c)s 4 7 9

Euro Stoxx 50

Sample 37 5 37 5 37 5 37 5

Total significance

43% 40% 24% 40% 19% 40% 65% 80%

Average βs 0.69 0.92 0.65 0.89 0.89 1.21 1.64 1.60 (N +)βs 16 2 9 2 7 2 33 24 (N −)βs

Notes: Equation (4) estimated using monthly data from 1999 to 2011. Heteroskedasticity robust OLS estimates of the exchange rate exposure, β is

, for the Euro Stoxx TMI and the Euro Stoxx 50 constituents are reported along with the percentage of firms significantly affected by exchange rate fluctuations, the average magnitude of significant exposure coefficients and the number of firms positively, N +, and negatively, N −, affected by currency variations. Each firm is grouped as MNCs, low exporters or domestic according to the ratio of non-Eurozone revenues on total revenues. Equation (4) estimated using monthly data from 1999 to 2011. Heteroskedasticity robust OLS estimates of the exchange rate exposure, βis , for the Euro Stoxx TMI and the Euro Stoxx 50 constituents are reported along with the percentage of firms significantly affected by exchange rate fluctuations, the average magnitude of significant exposure coefficients and the number of firms positively, N +, and negatively, N −, affected by currency variations. Each firm is grouped as MNCs, low exporters or domestic according to the ratio of non-Eurozone revenues on total revenues. (a)Average of significant βis from Equation (4). (b)Number of significant and positive βis from Equation (4). (c)Number of significant and negative βis from Equation (4).

122 F.

P a rla

p ia

n o

et al.

Table 5. Exchange rate exposure of European firms with a breakdown by industry.

JPY USD GBP CHF

Financials(a) Non Financials(b) Financials Non Financials Financials Non Financials Financials Non Financials

Euro Stoxx TMI

Sample(c) 83 446 83 446 83 446 83 446

Total significance

49% 24% 35% 9% 20% 9% 72% 49%

Average β (d) s 0.90 0.79 0.82 0.58 0.99 0.58 1.90 1.88

(N +)β(e)s 41 101 29 33 17 31 60 219 (N −)β(f)s 4 7 9

Euro Stoxx 50

Sample 12 37 12 37 12 37 12 37

Total significance

50% 41% 33% 24% 25% 19% 83% 62%

Average βs 0.67 0.67 0.44 0.66 1.02 0.93 1.75 1.60 (N +)βs 5 15 3 9 3 7 10 23 (N −)βs 1 1

Notes: Equation (4) estimated using monthly data from 1999 to 2011. Heteroskedasticity robust OLS estimates of the exchange rate exposure, β is

, for the Euro Stoxx TMI and the Euro Stoxx 50 constituents are reported along with the percentage of firms significantly affected by exchange rate fluctuations, the average magnitude of significant exposure coefficients and the number of firms positively, N +, and negatively, N −, affected by currency variations. Each firm is grouped as financial or non-financial according to the Bloomberg Industry Classification System. (a)Financials include firms in financial, banking, and insurance sectors. (b)Non-financials are composed of firms in Industrial, Utility and REIT sectors. (c)The sample size is 529 firms from Euro Stoxx TMI. Securities with less than 70 observations have been excluded from the analysis. (d)Average of significant βis from Equation (4). (e)Number of significant and positive βis from Equation (4). (f)Number of significant and negative βis from Equation (4).

The European Journal of Finance 123

Table 6. Exchange rate exposure of European firms with a breakdown by country.

AT BE DE ES FI FR GR IE IT NL PT Others(a)

Number of firms 23 40 85 39 43 100 47 15 79 37 14 7

JPY Proportion significant (%)

4 23 26 15 16 27 26 27 37 22 21 0

Average β (b) s 1.07 1.06 0.89 0.66 0.50 0.78 1.02 0.93 0.80 0.54 0.60 1.75

USD Proportion significant (%)

9 18 14 15 9 10 17 0 10 11 50 14

Average βs 0.80 0.83 0.76 0.70 0.33 0.39 1.18 0.67 0.09 0.68 1.08 GBP Proportion

significant (%)

0 15 8 15 9 10 19 7 10 5 29 0

Average βs 0.49 0.84 0.91 − 0.93 1.03 1.44 − 3.61 0.81 − 0.19 0.88 CHF Proportion

significant (%)

48 48 67 41 30 53 47 27 66 49 79 43

Average βs 2.09 1.72 2.17 1.40 1.75 1.74 2.40 2.35 1.64 1.93 1.72 2.33

Financials (%) 17 28 8 28 5 9 15 7 30 8 29 0 Non-financials (%) 83 73 92 72 95 91 85 93 70 92 71 100 MNCs (%) 75 47 76 51 81 82 40 100 51 86 73 100 Low exporters (%) 15 16 18 29 19 9 33 0 18 14 9 0 Domestic (%) 10 37 6 20 0 8 28 0 31 0 18 0

Notes: Equation (4) estimated using monthly data from 1999 to 2011. Heteroskedasticity robust OLS estimates of the exchange rate exposure, βis , for the Euro Stoxx TMI are reported along with the percentage of firms significantly affected by exchange rate fluctuations. Each firm is grouped according to the country of origin. (a) Countries with less than five firms. (b)Average of significant βis from Equation (4).

firms outside the financial industry. The proportion of financial firms significantly affected is larger and, in most cases, more than double the proportion of non-financial firms and, in addition, the magnitude of exposure coefficients is, on average, greater.

We next consider the sensitivity of firms to exchange rate risk by the firm’s country of origin within the Eurozone; Table 6 presents the results. The list of countries relatively more exposed to exchange rate fluctuations includes Greece, Ireland, Italy, Portugal, Spain (so-called GIIPS countries) and Belgium. This result can be partially explained by examining the concentration of financial firms in each of these countries. In fact, the ‘GIIPS’ countries, which experienced sovereign default risk late in the sample, show the largest concentration of financial firms when compared with other countries; as discussed before, firms in these categories have higher rates of exposure to exchange rate risk (see Table 5). In a recent paper, Acharya and Steffen (2013) provided evidence of a large ‘carry trade’ by European banks towards domestic sovereigns. By accessing short-term unsecured funding in wholesale markets, banks appear to have undertaken long peripheral sovereign bond positions. These carry trades were undertaken for the most part by ‘GIIPS’ banks showing a form of home bias. We suggest that our result may provide some evi- dence for the existence of spillovers between sovereign and corporate risk perception, although this requires further investigation.

124 F. Parlapiano et al.

(a) (b)

(c) (d)

Figure 1. Relationship between exchange rate exposure and the Eurozone stock market exposure. Using quantile regression, we investigate the interdependence between estimated currency betas, β̂is, and market betas, β̂im, where the exposure to market risk, as captured by the market beta, is the response variable as in Equation (5). We plot OLS quantile regression estimated for τ ranging from 0.05 to 0.95 (the solid dotted curve); in particular, each point measures the impact of a one-unit change of the currency beta on the market beta. For each of the plots, the x axis has the quantile scale and the y axis as the response variable scale. The two solid curves represent 95% confidence intervals of the estimated coefficients using quantile regression. The dashed line in each plot shows OLS estimate of the conditional mean effect and the two dashed dotted lines represent conventional 95% confidence intervals for the least squares estimates.

4.2 Relating market and exchange rate sensitivities

The empirical work carried out in this section has generated a pool of estimates of individual firm sensitivities to market risk, βim, and exchange rate risk, βis. Using quantile regressions, we now consider the extent to which these sensitivities may be related. If firms are simply more sensitive to both types of shocks, then we would expect a monotonic relationship between these loadings. We characterize quantiles, τ , of the conditional distribution of market betas as a function of currency betas generated from the quantile regression framework between estimated currency betas, β̂s, and market betas, β̂Mm , as follows:

β̂Mim = θ0 + θ1 ( β̂is

) + μi, (5)

where β̂Mim is the market risk exposure for a firm i and β̂is is the exchange rate risk exposure from Equation (4). Figure 1 plots the estimated θ 1 for τ ranging from 0.05 to 0.95 (the solid dotted curve), where the horizontal axis is the quantile scale and the vertical axis is the response vari- able scale. In particular, each point measures the impact of a one-unit change in the currency

The European Journal of Finance 125

beta on the market beta, holding other covariate fixed. The two solid curves represent 95% confidence intervals of the estimated coefficients using quantile regression. The dashed line in each plot shows OLS estimate of the conditional mean effect and the two dotted lines represent conventional 95% confidence intervals for the least-squares estimates.

It is clear that the relationship is not linear for any of the currencies considered. The lowest quantile of β̂Mim (on the horizontal axis) is associated with the lowest exchange rate exposure sensitivity – that is the figures all begin in the bottom left hand corner – representing that these firms are simply insensitive to both types of risk. As sensitivity to market risks increases to around the 50% quantile, the exchange rate sensitivity of firms also increases. Firms in this range are more sensitive to both types of risk. More interestingly, post the 50% quantile, we typically see a decline in the exchange rate exposure sensitivity of the firms. These firms are more sensitive to unexpected market risk but less so to exchange rate risk. Potentially, this represents the finding in the literature that some firms may not have detectable exchange rate exposure risk due to their actions to hedge exchange rate risk.

Table 7. Exchange rate exposure of European firms (augmented market model of Jorion (1990)).

Market model type

Augmented Non-orthogonalized

Currency trend JPY ( − 12.05%) Average β 0.30 0.20 Total significance 84 (15.88%) 33 (6.00%) (N +)β 59 21 (N −)β 25 12 Currency trend USD (30.80%)

Average β 0.26 0.15 Total significance 72 (13.61%) 47 (8.88%) (N +)βs 49 29 (N −)βs 23 18 Currency trend GBP (30.85%)

Average βs − 0.06 − 0.12 Total significance 61 (11.53%) 32 (6.05%) (N +)βs 32 16 (N −)βs 29 16 Currency trend CHF ( − 24.38%) Average βs 1.05 1.32 Total significance 60 (11.34%) 27 (5.10%) (N +)βs 49 24 (N −)βs 11 3

Notes: Equations (1) and (4) estimated using monthly data from 1999 to 2011. Average β represents the average of significant β1 from Equation (1) in the case of augmented model and the average of significant βis from Equation (4) in the case of non-orthogonalized model. Non-orthogonalized market return and exchange rate return have been employed in Equation (4). OLS estimates of the exchange rate exposure for the Euro Stoxx TMI constituents are reported along with the percentage of firms significantly affected by exchange rate fluctuations, the average magnitude of significant exposure coefficients and the number of firms positively, N +, and negatively, N −, affected by currency variations.

126 F. Parlapiano et al.

4.3 Sensitivity analysis

The results reported in this paper support that a higher proportion of European firms are affected by exchange rate risk than evidenced in the previous literature. As these studies largely follow a Jorion (1990) style model, and do not account for potential collinearity, this may be a contributing factor to the differing results. To explore this, we implement the Jorion model for our data sample with the results reported in Table 7. We report the results for the Euro Stoxx TMI firms and these can be compared with those reported in the first column of Table 3. For the JPY and CHF, the estimated proportion of firms significantly affected by exchange rate risk using the Jorion (1990) model is remarkably lower than that using the orthogonal market model implemented in this paper; in the case of the USD and the GBP the proportion is almost the same. In all cases, the average estimate of the relationship between individual firm equity return and exchange rate risk, the βis, is larger with the orthogonal market estimates than the Jorion (1990) approach. Notably, the average of the significant coefficients in the orthogonal market models are always positive, in contrast to the varying results obtained using Jorion (1990) approach. To conserve space, we omit the Euro Stoxx 50 results as the outcomes are consistent with those reported (a full set of results are available from the authors on request).

The additional control variables in the orthogonal market model may also be playing a role. To explore this aspect, we estimate Equation (4), but instead of using the orthogonalized market and exchange rate shocks, we substitute these with the market returns and the exchange rate changes directly – that is we omit the first two steps of the three step procedure. The results for the full sample are presented in Table 7, and show that the proportion of firms affected by exchange rate changes is uniformly smaller than in the orthogonal model implemented in Section 2 of this paper, and the impact coefficients are smaller and not uniformly positive. The body of evidence on the common factors which affect exchange rate and equity market conditions, coupled with the difference that omitting to orthogonalize exchange rate and market shocks makes to the results, supports the importance of controlling for potential collinearity using an orthogonalized model.

5. Concluding remarks

Previous works on firms’ exchange rate exposure have focused on the US markets and found that the US dollar exchange rate fluctuations have weak effects on US stock returns. In light of the vast empirical evidence, Bartram and Bodnar (2007) argue that the proportion of firms significantly exposed is not as high as literature leads us to believe. This could be partially attributed to failure to recognize reducing exchange rate exposure. If firms react rationally to the exposure by undertaking operational and financial hedging actions, it is plausible that most firms are not exposed. Alternatively, stock returns reflect only the residual exposure of firms, that is, net of corporate hedging policies. The international evidence on currency exposure provided so far has found significant results for just 10–25% of the cases (Bartram and Bodnar 2007).

This study investigates the exchange rate exposure of European firms accounting for the joint influence of macroeconomic fundamentals on stock market and exchange rate returns. The orthogonal market model approach proposed in Doukas, Hall, and Lang (2003) and later in Priestley and Ødegaard (2007) is employed and the results are compared with the mainstream augmented CAPM model of Jorion (1990).

We examine the three major trading partners of the Eurozone by analyzing the impact of fluc- tuations in the Euro exchange rate against the US Dollar, British Pound and Japanese Yen on the value of European firms. In addition, we include the Swiss Franc in our analysis due to close

The European Journal of Finance 127

geographical and economic relationships between the Eurozone and Switzerland and the CHF’s role as a safe-haven currency. We find that 11–52% of the Euro Stoxx TMI constituents are sig- nificantly affected by exchange rate fluctuations. When compared to the results of other studies, our estimates exceed previous findings reported in Muller and Verschoor (2006b), Bartram and Karolyi (2006), Nguyen, Faff, and Marshall (2007), Hutson and O’Driscoll (2010) and Inci and Lee (2014). We also observe that among the currency pairs analyzed the CHF had the highest impact on the returns of the European firms, with the greatest impact on the firms in the financial sector and firms with large capitalization.

We find that stock returns react positively to the Euro appreciation during 1999–2011; our results are consistent with the investor portfolio results based on excess returns in equities and exchange rates in Campbell, Medeiros, and Viceira (2010) once we control for the quotation convention. Segmenting the results by level of internationalization reveals that domestic firms appear more exposed than MNCs, highlighting the need for a better corporate management of exchange rate risk by these firms. As expected, firms operating in the financial sectors system- atically experience greater exposure to exchange fluctuations both in terms of magnitude and percentage of firms significantly exposed. The country breakdown of the firms reveals differ- ent impact of exchange rate fluctuations on these Eurozone national stock markets. In countries that experienced sovereign debt crisis and countries where the concentration of financial firms is larger we observe greater exchange risk exposure.

Our results suggest that domestic firms are more vulnerable to unexpected exchange rate fluc- tuations than MNCs. This presents an opportunity for further research with appropriate controls for differences in risk management policies of firms.

Acknowledgments

We thank Giovanni Palomba, participants at the fifth International Risk Management Conference and two anonymous referees for helpful comments and discussion in improving the paper. This work was made possible by the facilities of the University of Cambridge Judge Business School – Center for Financial Analysis & Policy.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. The number of constituents of the Euro Stoxx TMI varies to coverage of approximately 95% of the free-float mar- ket capitalization of the European Monetary Union, while the Euro Stoxx 50 covers 60% of the free-float market capitalization of the Euro Stoxx TMI super sector index.

2. Data sources and construction: we initially followed Fama and Gibbons (1984) to obtain measures of expected and unexpected inflation. However, the low volatility of inflation within the Eurozone meant that their approach was not appropriate for this sample. We considered three alternative proxies for expected inflation: (i) world oil price inflation (as approximated by the West Texas Intermediate); (ii) the European Central Bank survey of professional forecasts; and (iii) the previous month’s realized inflation, respectively. The results led us to the prefer lagged realized inflation, consistent with the approach of Gao (2000), who employs realized inflation rate as a control variable in identifying unexpected exchange rates variation. Industrial production and money supply are measured as percentage growth rates, the term premium is the 10 year government bond less 3 month note yield, the international interest rate differential is the 3 month spread and trade balance is the log difference between exports and imports. Most data are sourced from the European Central Bank, with the exception of the term premium and interest rate spread data which are calculated from Datastream series.

128 F. Parlapiano et al.

References

Acharya, V. V., and S. Steffen. 2013. “The ‘greatest’ Carry Trade Ever? Understanding Eurozone Bank Risks.” NBER Working Paper 19039, 1–72.

Adjaouté, K., and J. Danthine. 2004. “Equity Returns and Integration: Is Europe Changing?.” Oxford Review of Economic Policy 20 (4): 555–570.

Amihud, Y., and R. M. Levich. 1994. Exchange Rates and the Valuation of Equity Shares in Exchange Rates and Corporate Performance. New York: Beard Books.

Amiti, M., O. Itskhoki, and J. Konings. 2014. “Importers, Exporters and Exchange Rate Disconnect.” American Economic Review 104 (7): 1942–1978.

Andersen, T. G., F. X. D. Tim Bollerslev, and C. Vega. 2003. “Micro Effects of Macro Announcements: Real-Time Price Discovery in Foreign Exchange.” American Economic Review 93 (1): 38–62.

Bartram, S. M. 2008. “What Lies Beneath: Foreign Exchange Rate Exposure, Hedging and Cash Flows.” Journal of Banking & Finance 32 (8): 1508–1521.

Bartram, S. M., and G. M. Bodnar. 2007. “The Exchange Rate Exposure Puzzle.” Managerial Finance 33, 642–666. Bartram, S. M., G. W. Brown, and F. R. Fehle. 2009. “International Evidence on Financial Derivatives Usage.” Financial

Management 38 (1): 185–206. Bartram, S. M., and G. A. Karolyi. 2006. “The Impact of the Introduction of the Euro on Foreign Exchange Rate Risk

Exposures.” Journal of Empirical Finance 13 (4–5): 519–549. Beckmann, J., A. Belke, and M. Kuhl. 2011. “The Dollar-Euro Exchange Rate and Macroeconomic Fundamentals: A

Time-Varying Coefficient Approach.” Review of World Economics 147, 11–40. Bodnar, G. M., and M. H. F. Wong. 2003. “Estimating Exchange Rate Exposures: Issues in Model Structure.” Financial

Management 32, 35–67. Bodnar, G. M., and W. M. Gentry. 1993. “Exchange Rate Exposure and Industry Characteristics: Evidence from Canada,

Japan, and the USA.” Journal of International Money and Finance 12 (1): 29–45. Brunnermeier, M. K., S. Nagel, and L. H. Pedersen. 2008. “Carry Trades and Currency Crashes.” NBER Working Paper

14473 (November), 1–35. Campbell, J. Y., K. S.-D. Medeiros, and L. M. Viceira. 2010. “Global Currency Hedging.” Journal of Finance 65 (1):

87–121. Chen, N.-F., R. Roll, and S. A. Ross. 1986. “Economic Forces and the Stock Market.” The Journal of Business 59 (3):

383–403. Choi, J. J., and C. Jiang. 2009. “Does Multinationality Matter? Implications of Operational Hedging for the Exchange

Risk Exposure.” Journal of Banking and Finance 33 (11): 1973–1982. Chordia, T., and L. Shivakumar. 2002. “Momentum, Business Cycle, and Time-Varying Expected Returns.” The Journal

of Finance 57 (2): 985–1019. Davies, D., C. Eckberg, and A. Marshall. 2006. “The Determinants of Norwegian Exporters’ Foreign Exchange Risk

Management.” The European Journal of Finance 12 (3): 217–240. Dimson, E., P. Marsh, and M. Staunton. 2002. Triumph of the Optimists: 101 Years of Global Investment Returns.

Princeton: Princeton University Press. Dominguez, K. M., and L. L. Tesar. February 2001. “Trade and Exposure.” Working Paper 8129, National Bureau of

Economic Research. Doukas, J. A., P. H. Hall, and L. H. P. Lang. 2003. “Exchange Rate Exposure at the Firm and Industry Level.” Financial

Markets, Institutions & Instruments 12 (5): 291–346. Economist Intelligence Unit, 2011. Risk Radar 2011. How Firms Are Navigating Risk. Technical Report, The Economist. Engel, C., and K. D. West. 2005. “Exchange Rates and Fundamentals.” Journal of Political Economy 113, 485–517. Fama, E. F. 1981. “Stock Returns, Real Activity, Inflation, and Money.” The American Economic Review 71 (4):

545–565. Fama, E. F., and K. R. French. 1992. “The Cross-Section of Expected Stock Returns.” The Journal of Finance 47 (2):

427–465. Fama, E. F., and K. R. French. 1993. “Common Risk Factors in the Returns on Stocks and Bonds.” Journal of Financial

Economics 33 (1): 3–56. Fama, E. F., and M. R. Gibbons. 1984. “A Comparison of Inflation Forecasts.” Journal of Monetary Economics 13 (3):

327–348. Flannery, M. J., and A. A. Protopapadakis. 2002. “Macroeconomic Factors do Influence Aggregate Stock Returns.”

Review of Financial Studies 15 (3): 751–782. Galati, G., A. Heath, and P. McGuire. 2007. “Evidence of Carry Trade Activity.” BIS Quarterly Review 3: 27–41.

The European Journal of Finance 129

Gao, T. 2000. “Exchange Rate Movements and the Profitability of U.S. Multinationals.” Journal of International Money and Finance 19 (1): 117–134.

Geske, R., and R. Roll. 1983. “The Fiscal and Monetary Linkage Between Stock Returns and Inflation.” The Journal of Finance 38 (1): 1–33.

Granger, C. W. J., B.-N. Huang, and C.-W. Yang. 2000. “A Bivariate Causality Between Stock Prices and Exchange Rates: Evidence from Recent Asian Flu Data.” The Quarterly Review of Economics and Finance 40 (3): 337–354.

Hamao, Y. 1988. “An Empirical Examination of the Arbitrage Pricing Theory Using Japanese Data.” Japan and the World Economy 1 (1): 45–61.

Hattori, M., and H. S. Shin. 2009. “Yen Carry Trade and the Subprime Crisis.” IMF Staff Paper 56 (2): 384–409. Hau, H., and H. Rey. 2006. “Exchange Rates, Equity Prices, and Capital Flows.” Review of Financial Studies 19 (1):

273–317. He, J., L. Ng, and X. Wu. 1996. Foreign Exchange Rate Exposure, Risk and the Japanese Stock Market. Technical Report,

University of Wisconsin. Hutson, E., and A. O’Driscoll. 2010. “Firm-Level Exchange Rate Exposure in the Eurozone.” International Business

Review 19 (5): 468–478. Inci, A. C., and B. S. Lee. 2014. “Dynamic Relations Between Stock Returns and Exchange Rate Changes.” European

Financial Management 20 (1): 71–106. Jorion, P. 1990. “The Exchange-Rate Exposure of U.S. Multinationals.” The Journal of Business 63 (3): 331–345. Keloharju, M., and M. Niskanen. 2001. “Why Do Firms Raise Foreign Currency Denominated Debt? Evidence from

Finland.” European Financial Management 7 (4): 481–496. Koutmos, G., and J. Knif. 2011. “Exchange Rate Exposure in the Pre- and Post-Euro Periods: Evidence from Finland.”

The European Journal of Finance 17 (8): 661–674. Meese, R. A., and K. Rogoff. 1983. “Empirical Exchange Rate Models of the Seventies: Do They Fit Out of Sample?.”

Journal of International Economics 14 (1–2): 3–24. Muller, A., and W. F. Verschoor. 2006a. “Foreign Exchange Risk Exposure: Survey and Suggestions.” Journal of

Multinational Financial Management 16 (4): 385–410. Muller, A., and W. F. C. Verschoor. 2006b. “European Foreign Exchange Risk Exposure.” European Financial

Management 12 (2): 195–220. Nguyen, H., R. Faff, and A. Marshall. 2007. “Exchange Rate Exposure, Foreign Currency Derivatives and the

Introduction of the Euro: French Evidence.” International Review of Economics and Finance 16 (4): 563–577. Phylaktis, K., and F. Ravazzolo. 2005. “Stock Prices and Exchange Rate Dynamics.” Journal of International Money

and Finance 24 (7): 1031–1053. Priestley, R., and B. A. Ødegaard. 2007. “Linear and Nonlinear Exchange Rate Exposure.” Journal of International

Money and Finance 26 (6): 1016–1037. Roll, R. 1992. “Industrial Structure and the Comparative Behavior of International Stock Market Indices.” The Journal

of Finance 47 (1): 3–41. Taylor, M. P. 1995. “The Economics of Exchange Rates.” Journal of Economic Literature 33 (1): 13–47.