Foreign exchange exposure of US tourism-related firms

Foreign exchange exposure of US tourism-related firms

Tourism Management 32 (2011) 934e948 Contents lists available at ScienceDirect Tourism Management journal homepage: www.elsevier.com/locate/tourman ...

333KB Sizes 2 Downloads 48 Views

Tourism Management 32 (2011) 934e948

Contents lists available at ScienceDirect

Tourism Management journal homepage: www.elsevier.com/locate/tourman

Foreign exchange exposure of US tourism-related firms Seul Ki Lee 1, SooCheong (Shawn) Jang* Department of Hospitality and Tourism Management, Purdue University, West Lafayette, IN 47907-0327, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 12 January 2010 Accepted 10 August 2010

To date, literature on foreign exchange risk has paid a particular attention to multinationals in traderelated industries. The tourism sector is also sensitive to the exchange rates between travelers’ home countries and their destinations. Suspecting that the exposure of domestic tourism-related firms to foreign exchange risk results from price elasticity of demand, the current study tested the cash flow exposure of sample firms, accounting for nonlinearity, asymmetry, and lagged effects. As a result, a significant percentage (78%) of domestic tourism-related firms was found to have significant foreign exchange exposure. This study also found that exchange rate exposure for tourism-related firms was nonlinear, asymmetric, and lagged. The evidence implied that several tourism-related firms are passive regarding their exposure and may face financial burdens caused by demand fluctuations. Implications and suggestions are presented along with the findings of the study. Ó 2010 Elsevier Ltd. All rights reserved.

Keywords: Tourism industry Foreign exchange exposure Cash flow Trade-weighted index

1. Introduction Since bilateral exchange rates have been dominated by floating regimes, the extent to which a firm’s value is affected by changes in currency valuation has received considerable attention from the academy and the industry alike. The formidable scholarly interest on the relationship between exchange rates and the firms’ ability to generate cash flows is evidenced by Bartram’s (2008) identification of 34 publications on the topic between 1990 and 2006. While the literature to date has concentrated on multinational firms and traded industries such as automotives (Williamson, 2001) or manufacturing (Miller & Reuer, 1998), finance theory stipulates that firms without foreign accounts may also face exchange risk when the industry is characterized by demand, activities, or competition influenced by changes in currency conversion rates (Adler & Dumas, 1984). However, the difficulties that often arise in documenting empirical evidence of exchange rate exposure for multinational corporations (MNCs) and trading firms (Bartram, 2007) has hindered the extension of such efforts to firms with less obvious sources of exposure. The impact of exchange rates on international tourism has been observed by numerous studies, and exchange rate is generally considered an important determinant of tourism demand (Witt &

* Corresponding author. Tel.: þ1 765 496 3610. E-mail addresses: [email protected] (S.K. (SooCheong(Shawn) Jang). 1 Tel.: þ1 765 337 6249.

Lee),

[email protected]

0261-5177/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.tourman.2010.08.008

Witt, 1995). Tourism products are exports, per se, when consumed by international travelers (Divisekera, 2003), where the producers may engage in multilateral trade regardless of their intention, strategy, or targeted clientele. To this end, we expected to find that tourism-related firms are exposed to the uncertainty of demand fluctuations incurred by exchange rate changes and, in turn, are exposed to exchange rate risks. Consequently, questions regarding the nature of this exposure arise. Commonly exemplified by the utility industry (Dumas, 1978), the cross-elasticity of demand produces indirect exposure for firms without trade accounts when their competitors are subject to exposure risks. As a result, a firm is highly likely to be exposed when the industry or its competition as a whole is exposed. Among tourism industries, airlines are intrinsically dependent on overseas activities (Carter et al., 2006), while evidence shows that major US firms in hotel (Jang & Tang, 2009; Lee, 2008; Lee & Jang, 2010), restaurant (Hua & Upneja, 2007; Park & Jang, 2010) or casino (Economist, 2004) sectors actively pursue internationalization. Exchange rates may affect firms in these sectors directly or indirectly, the latter referring to naturally unexposed firms ‘picking up’ exchange rate risks by competing with directly exposed firms (Lee & Jang, 2010). In contrast, largely domestic sectors without noticeable foreign input or activities constitute a unique question in this consideration. For example, with few exceptions, travel agencies and recreation firms have not signaled significant overseas expansion and may be exposed to exchange rate risks primarily through demand shifts. Our suspicion that these firms may be exposed to exchange rates without foreign accounts raises some important concerns, as the firms’ ability to operationally hedge by using

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

foreign inputs to match foreign sales or to prevent pass-through by selectively pricing for specific markets are often limited. Meanwhile, the economic significance of travel services and recreation sectors in the US is quite sizable. The US Bureau of Economic Analysis (Bureau of Economic Analysis, 2008) reported that the estimated industry output for travel arrangement/reservation services and recreation/entertainment in 2007 were approximately 43,226 and 85,794 million US dollars, which accounted for 5.79% and 11.49% of all travel and tourism-related output respectively. According to the Office of Travel and Tourism Industries (Office of Travel & Tourism Studies, 2008), 48% and 6% of all overseas visitors booked their air trips through travel agents or tour operators. 25%, 16%, 15% and 6% of all overseas visitors attended amusement/theme parks, guided tours, concerts/plays/musicals, and sports events, respectively. Fig. 1 displays the historical series of the broad real effective exchange rate (REER) published by the US Federal Reserve. The Federal Reserve defines the broad index as “a weighted average of the US dollar against the currencies of a large group of major US trading partners.” It also shows the tourism commodity output in the travel services sector, the recreation sector, and the total travel and tourism industry between 1999 and 2006 (Bureau of Economic Analysis, 2008). At a glance, the movements of travel and tourism outputs and the exchange rate index display an inverse relationship, supporting the primary question of this study. However, the literature on firms without foreign accounts is scarce and shows a lack of anticipation or interest in the exposure of such firms in general. Even as firm values are being significantly affected by exchange rate changes, theoretical and empirical gaps may result in a lack of understanding and countermeasures to combat exposure risks. Further, among those examining foreign currency risks, much is dedicated to investigation of the exposure of stock returns, perhaps due to the availability of data and the wide use of the factor models. Use of stock return data to estimate exposure nevertheless has caveats in studying the effect of currency value changes on firm values. As the firms engage in financial and operational hedging, the estimate of the firm’s true risk may not be accurately quantified. Moreover, the stock return incorporates the market’s perception on the firms’ susceptibility to foreign currency risk and therefore may not yield the true change in the firm values. For domestic firms, the source of exposure is rather limited. Without any foreign account assets in their balance sheets or income statements, the only source of exposure would be the operating cash flow, governed by the demand change induced by purchasing power parity movements. By using the more recently

favored cash flow model of exposure estimation, the firms’ hedging and investing activities, which may influence the net exposure at the corporate level, can be isolated out. Therefore, the purpose of this study is to suggest a theoretical framework and develop an appropriate methodology to identify the exchange rate exposure of domestic travel agencies and recreation firms, using the cash flow approach to examine the source of exposure and its effect on other activities of the firms. As a result of the analysis this study finds that the domestic tourismrelated firms are indeed exposed to the exchange rate risks through the demand changes incurred by currency value movements, while the firms’ financing and investing activities are indirectly affected by the volatility in operating cash flows. Firms with foreign income are found to have no exposure at the corporate level, presumably due to operational hedging effect. Implications for the industry and suggestions for further research are presented with the findings of the study.

2. Literature review 2.1. Exchange rate exposure Seminal works by Dumas (1978) and Adler and Dumas (1980, 1984) defined exchange rate exposure as “the measure of what one has at risk,” while risk is the “statistical quantities that summarize the probability” of a currency value deviation from the originally anticipated value. The exposure of an asset value is econometrically estimated by a single-equation regression:

Exposure of P to Si ¼ EðDP=DSi Þ

(1)

where E is the expectation operator, P is the dollar price of a risky asset on a given future date and S is the vector of macroeconomic variables ¼ {S1,.,Sn}. It immediately follows that exposure is the amount of change that is induced by the exogenous change in Si. Simply put, exposure is the partial effect of Si on P, estimated as the regression coefficient of the exchange rate variable Si. Eq. (1) is specified so that the asset value and exchange rates are expressed as levels. In practice, elasticity models are often used, but the difference is negligible. While many studies use variations of such factor models as Eq. (1), which employs stock returns as regressands (Bartram, 2008), another stream of research (Bartov & Bodnar, 1994; Bartram, 2008; Oxelheim & Wihlborg, 2003; Stulz & Williamson, 1997) advocates using cash flow variables as the dependent variable. For the cash flow model, a firm’s cash flows at period t, CFt, is regressed on the innovations of exchange rate, DEt:

CFt ¼ b0 þ b1 DEt þ 3t

Fig. 1. Historical output by US travel & tourism and by sectors with REER, 1999e2006.

935

t ¼ 1; .; T

(2)

The two models are theoretically identical, for a firm’s value is the sum of its current and future cash flows. Nevertheless, there are two important aspects to this approach. First, market perception and its view on a firm’s exposure, which is aggregated in the stock return model, are isolated out (Bartram, 2008). Second, cash flow allows estimating the firm’s exposure on respective activities of operation, financing, and investment, prior to inter-activity hedging (Flood & Lessard, 1986). Since the three types of cash flows are likely to be highly correlated (Bartram, 2008), as the firm’s financing and investing activities are subject to the availability of internal funds from operations, financing and investing cash flows are likely to show significant correlation with the exchange rate movements for firms with ‘operating’ exposures. The total exposure in this case would be the exposure to total cash flows and is equivalent, but not equal, to

936

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

the exposure estimated by stock return, as stock return incorporates the perceptions of the market and investors (Bartram, 2008). Estimation of the stock return model in earlier studies has mainly used the exchange rate variable and the market or national index as regressors, thus usually referred to as the two-factor model (Choi & Prasad, 1995). An implicit assumption in such a practice would be that other macroeconomic effects will be captured by greater movements of the index. Using only the exchange rate variable, or orthogonalizing the market index, may improve the significance of the results, but could lead to an unexpected outcome due to omitted-variable bias (Kiymaz, 2003). For example, Dumas and Solnik (1995) regressed the market return on the currency variables while Priestley and Ødegaard (2006) regressed the market return with the currencies and also a set of macroeconomic factors to use the residuals as the orthogonal market return uncorrelated with the changes in exchange rates, hence the term orthogonal market return. As studies using cash flow models do not have a readily available control variable, Bartram (2008) and Oxelheim and Wihlborg (1995) have advocated the control of macroeconomic effects, such as inflation and interest rates, through control variables or orthogonalizing the exchange rate variable with respect to these factors (Bartram, 2008; Choi & Prasad, 1995; Jorion, 1991). 2.2. Exposure of US industries A number of studies have examined the exchange rate exposure of US firms and industries since the initial work of Jorion (1990), who tested the exposure of 287 multinational firms and 40 US industry portfolios to find that only a small percentage of the sample was exposed. He found that as little as 5e16 out of 287 firms and 6 out of 40 industries were significantly exposed to exchange rates. In his subsequent study, Jorion (1991) also examined 20 US portfolios using CRSP data on NYSE firms and reported that 7 out of 20 industry portfolios had significant exposure, while Bodnar and Gentry (1993) examined 39 industry portfolios and found that 11 of them were significantly exposed. Table 1 summarizes the studies on exchange rate exposure of US industries. It can be seen that the percentage of US industries with exposure is somewhat low. In essence, extensively traded industries such as chemicals, energy, textiles, and other retail sectors are often found to be exposed. However, the exposures estimated through the factor models seem to be insignificant and small. These are residual exposures after hedging at the corporate or industry level, which is again also influenced by the perceptions of the market (Williamson, 2001). The implication is that the use of the cash flow model will disaggregate the exposures cancelling each other out in each of the firms’ activity (i.e. operating, investing, financing), unless the firm is operationally hedged. As an alternative approach, studies have concentrated on identifying the number of firms in a given industry that are exposed to exchange rate risk rather than testing the industry as a whole (Muller & Verschoor, 2006, 2008). 2.3. Considerations in measuring exposure Through the years researchers have attempted a number of theoretical and methodological developments to refine exposure estimation procedures and, in turn, solve the problem of weak empirical evidence, often known as the “exchange rate exposure puzzle” (Bartram, 2007). Considerations made by researchers are discussed in the following subsections. Before discussing the various approaches, it is worthwhile to note that in general it is difficult to assume the validity of estimation through the simplest model form, as it

requires three rather large foundational assumptions: contemporaneity, linearity, and symmetry. 2.3.1. Contemporaneous vs. lagged exposure Amihud (1994) and Bartov and Bodnar (1994) suggested that the limited evidence of exposure in former studies (i.e. Bodnar & Gentry, 1993) may be attributed to the non-contemporaneous effect of exchange rates. Walsh (1994) further found that a sixmonth lagged term (consistent with Amihud, 1994) is most significant. The complexity of the interaction between exchange rates and performance and mispricing (the effect of exchange rate change not correctly reflected on price) were cited by the authors as potential reasons for the lagged relationship. Bartov, Bodnar, and Kaul (1996) added to this school of thought by suggesting information asymmetry and market inefficiency as possible causes while theoretical justification of lagged exposure was provided by Adler and Dumas (1984) and Chang (2009). In contrast, in their study of Japanese firms He and Ng (1998) tested both contemporaneous and lagged effects and concluded that exposure is contemporaneous. Supporting the study by He and Ng (1998), a number of studies did not find significant results associated with lagged exposures, concluding that the market is efficient (Bartram & Bodnar, 2007). Theoretically, under the efficient market hypothesis the stock returns should not be exposed to the past changes of exchange rate as the market would have already incorporated the necessary changes in the firms’ future cash flows induced by currency conversion rates. However, the cash flows of the firms can be largely affected by past changes of the exchange rate, especially when the firms do not engage in active financial hedging. 2.3.2. Nonlinearity and asymmetry Some studies (Bartram, 2004; Koutmos & Martin, 2003; Priestley & Ødegaard, 2006) examined the nonlinear and asymmetric forms of exposure, citing these as possible reasons for partial success of obtaining evidence on exposure. Limited evidence suggests that the models are marginally more significant when exposures are specified nonlinear (Bartram, 2004) and asymmetric (Koutmos & Martin, 2003). Nonlinearity of exposure is grounded in the concept that a firm’s earnings and therefore its value is a nonlinear function of exchange rates. Reasons such as the firms’ ability to adjust production, imports, or exports pertaining to exchange rates (Ware & Winter, 1988), risk of defaulting on a foreign debt (Stulz, 2003), or foreign currency options with nonlinear payoff (Giddy & Dufey, 1995) were cited as possible reasons. Bartram (2004) further commented that in estimation, the effect of nonlinear exposure may appear more significant as firms mainly use linear tools to hedge. Eq. (3) illustrates the basic form for the nonlinear exposure model suggested by Bartram (2004):

CFt ¼ b0 þ b1 f ðDEt Þ þ 3t

t ¼ 1; .; T

(3)

where CFt is the cash flow of the firm at period t and f(DEt) is a nonlinear function of the rate of change in exchange rates at period t. Asymmetric exposure stands for the influence of the exchange rate changes on the firms’ earnings differing over appreciation and depreciation cycles respectively. Koutmos and Martin (2003) argued that exchange rate exposure can be asymmetric due to the pricing-to-market behavior of firms (Goldberg, 1995; Miller & Reuer, 1998), hysteresis (Christophe, 1997), or conversely, because the firms do not engage in symmetric hedging. Knetter (1989) also described a situation in which the asymmetric reactions were amplified due to greater pricing-to-market by firms intending to

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

937

Table 1 Studies on exchange rate exposure of US industries. Study

Period

Sample (industries/firms)

Results (sig. exposure)

Industries/firms with significant exposures

Jorion (1990)

1971e1987

6/40 Industries

Not specified

Jorion (1991)

1971e1987

40 Industries (287 multinationals) 20 Industries

7/20 Industries

Bodnar and Gentry (1993)

1979e1988

39 Industries

11/39 Industries

Choi and Prasad (1995)

1978e1989

20 Industries (409 multinationals)

5/20 Industries

Prasad and Rajan (1995)

1981e1989

20 Industries (765 NYSE firms)

8/20 Industries

Allayannis (1997)

1978e1986 1987e1990 1977e1989

137 Industries 124 Industries 65 Industries

1976e1994

Gold mining industry (2 gold mining firms) 82 Manufacturing industries

30/137 Industries 39/124 Industries 8e55 Industries (in 48-mo horizon) 2/2 Firms

Chow et al. (1997) Petersen and Thiagarajan (2000) Allayannis and Ihrig (2001)

1979e1995

Koutmos and Martin (2003)

1992e1998

9 Industries

4/9 Industries

Williamson (2001)

1973e1995

3/3 Firms

Muller and Verschoor (2006)

1971e1987

Automotive industry (GM, Ford, and Chrysler) 12 Industries (935 foreign operating firms)

All industries (272/935 firms)

Priestley and Ødegaard (2006)

1979e1998

28 Manufacturing industries

11/28 Industries

Muller and Verschoor (2008)

1970e2001

20 Industries (1075 multinationals)

4/18 Industries

All industries (170/1075 firms)

-

Mining Machinery Chemical Department stores Textile and apparel Other retail Other industries Mining Apparel Transport equip. Heavy construction other than buildings - Petroleum refining - Air transport - Motor freight transportation - Wholesale trade, durable goods - General merchandise stores - Misc. retail - Business services - Mining - Paper products - Chemical - Other retail - Finance and real estate - Textile and apparel - Machinery - Transport equip. - Other transport - Utilities - Dept. stores - Other retail - Financial and real estate Not specified (all samples, however, are manufacturing industries) Not specified American Barrick and Homestake Mining - Furniture and fixture - Chemicals - Stone/clay/concrete - Industrial machines/computers - Finance - Industrial - Cyclical - Noncyclical GM, Ford, and Chrysler At least some %age of firms in all industries were exposed to at least one currency, at a certain point Extensively traded industries: - Recreation - Automobiles - Machinery - Measuring/control equip. - Consumer goods - Electrical equipment - Business supplies Low trade industries: - Pharmaceutical products - Ship and rail equipment - Fabricated products - Business services At least some %age of firms in all industries were exposed to at least one currency, at a certain point

938

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

build market share. The asymmetric model suggested by Koutmos and Martin (2003) is represented by Eq. (4):

CFt ¼ b0 þ b1 DEtþ þ b2 DEt þ 3t ;

t ¼ 1; .; T

(4)

DEþ t

where CF is the cash flow of the firm at period t and and DE t respectively the positive and negative exchange rate changes (appreciations and depreciations of home currencies) in period t. A significant difference in coefficients b2 and b3 provides evidence that exposure is asymmetric between appreciation and depreciation of the home currency. 2.3.3. The exchange rate variable The discussion on exchange rate variables revolves around three aspects. First is the use of nominal and real exchange rates, where the latter is adjusted by domestic and foreign inflation rates. When tested by Choi and Prasad (1995), some differences in the results were observed, although they were described as negligible by the authors. Bartram and Bodnar (2007) also reported that nominal and real exchange rates are highly correlated. Second is the selection between bilateral exchange rates or the trade-weighted index. While a majority of studies have used the trade-weighted basket of currencies, the difference in results is not considerably noticeable (Bartram, 2004). However, the tradeweighted index may not represent the currencies that respective firms are exposed to and may introduce the effect of currency diversification for firms when exposure is examined. Lastly, expectations regarding the value of currency may affect exposure. As Adler and Dumas (1984) stated, “a currency will not be risky if appreciation or devaluation is expected.” Forward premiums can be used as a proxy. Therefore, the firm’s value is exposed only to unexpected changes in the currency value. Jorion (1991) addressed this issue and argued that indicators such as forward premium on exchange rate are a biased proxy for such expectations. He used an ARIMA model to conclude that forward premiums do not even outperform the spot exchange rate in predicting future changes. 3. Theoretical framework Without foreign accounts, domestic firm i’s operating cash flows at time t is simply written as:

  OCFi;t ¼ Mi;t DDOM þ DINT i;t i;t

(5)

where OCFi,t is the operating cash flows of firm i at time t, Mi,t is the unit margin for firm i’s product at time t, and Di,tDOM and Di,tINT are, respectively, US and international demand for firm i’s products at time t. Substitutability among international tourism products was shown by Divisekera (2003) and, accordingly, both demands are convex and anticipated to change due to own and cross-elasticity. The operating cash flow is determined by the difference between the price and cost multiplied by total purchases made by the demand. Ceteris paribus, the demands can be represented as functions of exchange rate as:

h    i DOM INT 1 þ fFX þ DINT 1 þ fFX OCFi;t ¼ Mi;t DDOM i;t i;t

(6)

DOM INT where fFx and fFx are percentage changes, or price elasticities, of domestic and international demand incurred by exchange rate innovations for firm i’s products, expressed as functions of the exchange rate. Such situation is likely to occur for domestic firms whose prices or the costs are uncorrelated with the exchange rates, whereas the demand sees changes in the real prices due to the

dollar conversion rate or because of the substitutability of the products. Although the existence of foreign income introduces some complications, the model may be extended to consider these aspects. Eq. (7) represents the foreign income of firms:

 i h INT 1 þ fFX FIi;t ¼ ðM=FXÞ DINT i;t

(7)

where FI is the foreign income of firm i at time t and FX is the exchange rate of the US Dollar to the currency of interest. Under this scheme, the effect of exchange rate variation will be at least partially hedged, as the demand alteration will be set off by the changes in dollar cash flows induced by the currency conversion rate. A firm’s total operating cash flows would be the sum of Eq. (6) and Eq. (7). Other variations such as foreign sales, operational hedging, and pricing-to-market can be modeled using similar notation. Based on this framework, some important hypotheses can be formed. First, we expect the exposure to be nonlinear for travel agencies and recreation firms. Since the demand function is nonlinear with respect to exchange rate (Divisekera, 2003), nonlinearity is incurred even when the firms do not adjust for exchange rate changes. Second, we expect the exposure to be asymmetric, as the demand is unlikely to respond symmetrically to exchange rate changes due to friction. For example, a certain number of travelers may decide not to visit a destination where the price of foreign currency has changed unfavorably, but it is counterintuitive that a symmetric response will take place (the same number of travelers will visit the destination) solely because the country’s currency became relatively cheaper. The hypothesis is also consistent with the stream of literature on push and pull factors of destination choice (Yoon & Uysal, 2005). Third, we expect the exposure to incorporate lagged effects. On average, the decision to visit the US is made at least two months ahead (Office of Travel & Tourism Studies, 2008). If part of the future demand is shifted through changes in the currency value at present, the current operating cash flows of the firm will depend on both the current exchange rate and its past changes. Including foreign input or income in the equation changes little in terms of nonlinearity and asymmetry, as additive and multiplicative combinations of nonlinear and linear functions are not expected to be either linear or symmetric. By using foreign input, however, the direction of exposure may change when the inflow of cash from changes in dollar-denominators outweighs the outflow of cash from depressed demand. Therefore, it is hypothesized that domestic tourism-related firms will enjoy increased operating cash flows from the depreciation of the dollar. Depreciation of the dollar will stimulate the foreign demand as the actual prices paid by the international travelers will be lower, while also increasing domestic demand because the US travelers will have less budget (when converted to foreign currency) to travel abroad and substitute to domestic products, where the prices in dollars presumably remain the same. Appreciation of the dollar, in the same fashion, will influence both the domestic and foreign demand to turn away from domestic products. On the contrary, through the line of reasoning above, the operationally hedged firms may experience no gain, or even a loss of operating cash flows from appreciation or depreciation of the dollar. Adapting Chow, Lee, and Solt’s (1997) terminology, hereafter the former type of exposure is referred to as “positive exposure,” whereas the latter type of exposure is referred to as “negative exposure.” Since respective cash flows of a firm are highly correlated, it is also of significant interest to examine the exposure of financing,

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

investing, and total cash flows. Operating, financing, and investing cash flows cannot be independent of one another, but rather are dependent on one another or assist the others. Therefore, if the operating cash flows are exposed to the exchange rates, it is intuitive to expect the effect on other cash flows as well. By substituting the left-hand side of the equation with financing, investing, or total cash flows, the exposure of non-operating activities of a firm can be observed. 4. Data and methods

Shapiro, 1984) or differencing them and taking the absolute values (Oxelheim & Wihlborg, 1995). Oxelheim and Wihlborg (1995) explained that using levels or rates of change is practically irrelevant in terms of outcome as the results can simply be recalculated. Accordingly, the appropriate form should be decided based on potential autocorrelation or stationarity in the variables. Econometrically, regressing the first differences renders a slightly different interpretation of the regression coefficients. Regressing the level on percentage changes simply yields one hundredth of the elasticity coefficients, as shown in Eq. (8) (Wooldridge, 2002).

4.1. Sample and the dependent variable All domestic US firms in the NAICS four-digit code of 5615 (Travel Arrangement and Reservation Services) and the two-digit code of 71(Arts, Entertainment, and Recreation) were searched between the years 2001 and 2008, yielding 29 firms. After excluding 11 firms with missing values or insufficient time spans (to estimate the 15 parameters specified in Eq. (11)), 18 firms were selected. 8 were classified as travel agencies/tour operators and 10 were grouped as recreation firms. A total of 482 firm-quarter observations were obtained, which was equivalent to 1928 firmquarter-cash flow observations, as COMPUSTAT offers information on the 4 types of cash flows: operating, financing, investing, and total. Two travel agencies and three recreation firms reported foreign income. Unlike the small foreign income of travel agencies, which can be generated directly or indirectly through commissions or fees, the sizable foreign income of recreation firms was counterintuitive. An examination of the Security Exchange Commission filings revealed that these recreation firms operated in foreign countries either through owned subsidiaries or partnerships/ leasehold agreements. Nevertheless, all 18 domestic firms were tested in the current study for comparative purposes. Table 2 lists the samples firms and their descriptive statistics. Since a firm’s cash flows can be either positive or negative, the use of an elasticity model, such as the method used by Stulz and Williamson (1997) based on percentage changes or logarithmic differences, is limited to firms that consistently have positive cash flows throughout the duration of the sample period. For this reason, diverse techniques are employed by researchers on cash flow variables. Bartram (2008) noted the variety of techniques, such as simply differencing raw cash flows (Garner &

939

Regression of :

lnðyÞ on lnðxÞ : Y on lnðxÞ :

%DY ¼ b1 %Dx DY ¼ ðb1 =100Þ%Dx

(8)

The current study uses levels of cash flows, which is theoretically analogous to the original model proposed by Adler and Dumas (1984), as first-order autocorrelation was not identified due to the cyclical nature of the quarterly data used in the current study. Scaling was done by adjusting the cash flows with prices (consumer price indices), consistent with Oxelheim and Wihlborg’s (1995, 2003) approach, rather than by assets. By scaling the cash flows with assets, the interpretation of the dependent variable changes to the firm’s performance in use of its assets to generate cash flows. This deviates a bit from Adler and Dumas’ (1984) original definition of exposure: “how much one has at risk.” A simple correlation test of the sample firms also revealed that the average correlation between assets and operating cash flows was less than 0.1, while one third of the firms had negative correlation.

4.2. Travel and tourism-weighted index (TTWI) In order to develop an index applicable to the travel and tourism industry, the currency weights of countries in travel and tourism trade with the US were identified from the Office of Travel and Tourism Industries data on international travel receipts and spending. Combining international travel and tourism receipts and spending during the period 2001e2008, all currencies accounting for more than one percent (Kite, 2007) of total trade, defined as the sum of total international travel receipts and spending, were identified. As a result, thirteen currencies (see Table 3) were included in the index. The thirteen currencies accounted for 86.72%

Table 2 Descriptive statistics of sample firms. ID Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

NAICS classification

NAICS code

Assetsa

Average OCFa

s(OCF)

Foreign income

% Total income

Obs

Tour operator Travel agency Travel agency Travel agency Travel agency Travel agency Travel agency Travel agency Sports teams/clubs Sports teams/clubs Sports teams/clubs Racetracks Racetracks Racetracks Racetracks Amusement/theme parks Amusement/theme parks Promoters of performing arts

561520 561510 561510 561510 561510 561510 561510 561510 711211 711211 711211 711212 711212 711212 711212 713110 713110 711320

122 0.23 8295 84.29 62.89 93.98 583.46 4.65 14.70 7.63 12.03 624.8 1982 519.9 1578 2945 2419 2752

5.93 0.33 128.13 0.97 0.63 1.15 6.57 0.6 0.46 1.25 0.18 11.78 50.85 4.52 33.21 23.54 41.21 2.74

22.04 0.68 218.95 2.75 1.54 1.45 10.18 1.29 1.31 2.22 1.92 19.46 26.10 31.48 12.44 112.80 82.28 120.88

e e O e e e O e e e e e e O e e O O

e e 1.61 e e e 4.4 e e e e e e 24 e e 13 60.

32 16 25 19 29 32 22 21 32 31 19 32 32 32 32 32 28 16

OCF ¼ operating cash flows, s(OCF) ¼ standard deviation of OCF; income is derived from pretax figures; Obs is the number of quarterly observations. a In millions.

940

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

Table 3 Currency weights for TTWI. Year

CA (%)

UK (%)

BR (%)

MX (%)

VE (%)

AU (%)

CN (%)

HK (%)

IN (%)

JP (%)

KR (%)

TW (%)

EU (%)

2001 2002 2003 2004 2005 2006 2007 2008 Average

11 11 12 11 11 12 12 12 11

15 14 15 14 14 13 13 13 14

3 2 2 2 2 2 2 2 2

10 11 12 11 11 11 10 10 11

2 1 1 1 1 1 1 1 1

3 2 3 3 3 3 3 3 3

2 2 1 2 2 3 3 3 2

2 1 1 2 1 2 2 2 2

1 1 2 2 2 3 3 3 2

11 11 10 10 12 10 10 9 10

3 3 3 3 3 3 3 3 3

2 2 2 2 2 2 2 2 2

38 37 38 38 37 36 36 38 37

Abbreviations denote: (CA) Canada, (UK) United Kingdom, (BR) Brazil, (MX) Mexico, (VE) Venezuela, (AU) Australia, (CN) China, (HK) Hong Kong, (IN) India, (JP) Japan, (KR) Korea, South, (TW) Taiwan, (EU) Euro Area.

of all US travel and tourism trade between 2001 and 2008. Table 3 lists the currencies and their pertaining weights. Consistent with the approach of The US Federal Reserve, bilateral exchange rates were adjusted by prices changes in respective countries so that the outcome index represents a real effective exchange rate (REER) rather than a nominal rate. Consumer price indices (CPIs) were retrieved from OECD statistics portal and government databases for China, Hong Kong, Taiwan, and Venezuela. Adjusted exchange rates were aggregated by taking the geometric mean of the thirteen currencies. The weights are defined in Table 3. The TTWI was calculated as:

100 

Y w i;t ei;t =ei;0

(9)

Q where is the product of the bracketed term over the thirteen currencies, ei,t the amount of foreign currencies i at period t, ei,0 the amount of foreign currencies i at base period (1st quarter of 2001), and w i,t the travel and tourism weight for currency i at time t in Table 3. Table 4 compares the travel and tourism-weighted index and the broad real effective exchange rate published by the US Federal Reserve. The correlation between the two indices was found to be considerably high at 0.9689; yet, it can be seen that the two indices move in opposite directions in three quarters (boldfaced) during the sample period. This is an indication that using the broad REER for tourism-related firms may not yield efficient estimation.

Table 4 Comparison of TTWI and broad REER series between 2006 1Q and 2008 4Q. Period

TTWI Index

D%

Index

D%

2005 2005 2005 2005 2006 2006 2006 2006 2007 2007 2007 2007 2008 2008 2008 2008

83.16 84.01 84.54 85.53 86.14 83.33 82.39 82.83 83.00 79.67 78.20 75.06 74.73 72.30 74.11 86.71

0.01 0.01 0.01 0.01 0.01 0.03 0.01 0.01 0.00 0.04 0.02 0.04 0.00 0.03 0.03 0.17

88.33 89.36 89.61 90.23 88.99 87.62 87.17 86.67 86.52 84.45 82.89 80.04 78.56 77.34 79.02 87.48

0.01 0.01 0.00 0.01 L0.01 0.02 0.01 L0.01 L0.00 0.02 0.02 0.03 0.02 0.02 0.02 0.11

1Q 2Q 3Q 4Q 1Q 2Q 3Q 4Q 1Q 2Q 3Q 4Q 1Q 2Q 3Q 4Q

4.3. The nonlinear function Various forms of nonlinear functions were suggested for modeling exposure (Bartram, 2004, 2002; Priestley & Ødegaard, 2006). In general, nonlinear functions can be categorized into two groups: convex or concave. While concave functions imply that small shifts in the exchange rate lead to a great volatility in the firm’s earnings (Bartram, 2004), they do not seem to be applicable to the sampled firms in the current study, as small changes in the exchange rate may not be as influential to the international demand as greater movements. In addition to the better fit of the convex functions, which has been reported by the previous studies (Bartram, 2004; Priestley & Ødegaard, 2006), the convex function is deemed theoretically more feasible as international demand is likely to be more responsive to greater movements of exchange rates. Therefore, the hyperbolic sine function suggested by Bartram (2002, 2004) was used. Eq. (10) and Fig. 2 display the form and the shape of the hyperbolic sine function.

f ðxÞ ¼ sinhðxÞ ¼



ex  ex

. 2

(10)

4.4. Controlling for macroeconomic impacts Former studies have chosen either to orthogonalize the exchange rate variable with macroeconomic factors and use the residuals to test the exposure (Choi & Prasad, 1995; Priestley & Ødegaard, 2006) or to include them along with the exchange rate variable as control variables (Bartram, 2007; Jorion, 1991). While it

10

Broad REER

Percentage change denoted by D%; 2001 1Q used as base ¼ (100).

y= sinh(x)

-10 5

-5 Fig. 2. Hyperbolic sine function.

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

 distributed lags for f(DEþ t ) and f(DEt ). Control variables used were P Q2e4, quarter dummies (1st quarter as the base), RSTt and RDSt, short term interest rate and term spread divided by long term interest rate (Bartram, 2002), and CCI, the Composite Coincident Indicator published by The Conference Board (2009).

has been reported that the results are practically not different (Bartram, 2007), the current study used macroeconomic indicators as control variables to simplify the interpretation of the results. Short-term and long-term interest rates were obtained from the public data made available by the US Federal Reserve as 3-month and 10-year T-bill rates. Other macroeconomic movements were controlled for by the Composite Coincident Indicator published by The Conference Board. The Composite Coincident Indicator is made up of four economic indices of employees on nonagricultural payrolls, personal income less transfer payments, industrial production, and manufacturing and trade sales. Further, it “historically coincides with the cyclical turning points in the aggregate economic activity of the US” (Conference Board, 2009). In general, controlling for macroeconomic effects is advocated, as it tends to yield more stable exposures and avoid potential multicollinearity problems than omitting them in the regression framework (Bartram, 2008).

4.6. Heteroscedasticity and autocorrelation Due to the nature of quarterly data (Wallis, 1972), heteroscedasticity and fourth-order autocorrelation were suspected. Durbin’s (1970) test and the BreuschePagan test were used to test for potential fourth-order autocorrelation and heteroscedasticity. As a result, half of the samples displayed significant fourth-order autocorrelation in their error terms, while 17% displayed heteroscedasticity. Even though presence of autocorrelation and heteroscedasticity do not introduce bias in the parameter estimates of OLS regression (Wooldridge, 2002), influence on standard errors of the parameters was expected. Therefore, NeweyeWest standard errors, which account for up to four lags of error terms, were used to test the coefficients with robustness to autocorrelation and heteroscedasticity in time-series models. Estimation through OLS and using adjusted robust errors is cited as a common practice in finance studies (Chow et al., 1997; Petersen, 2008).

4.5. Functional form The model shown in Eq. (11) is based on a number of preceding studies (Oxelheim & Wihlborg, 1995; Petersen & Thiagarajan, 2000; Stulz & Williamson, 1997) using cash flow models, but explicitly addresses nonlinearity, asymmetry, and lagged effects. As it was difficult to determine the length of lagged effects a priori, the lag structure was tested using a step-by-step procedure. Based on the two criteria of model fit (R2) and significance (F-test) it was concluded that the three-lag models performed the best for sampled firms:

CFi;t ¼ b0 þ

X

941

5. Results and discussion 5.1. Operating exposure of travel agencies and recreation firms For reporting purposes, summarized results of the analysis are used in this section. Complete results for all 72 regressions (18  4) are appendicized. Table 5 details the outcome of the regressions on the operating cash flows of the sample firms. The variables of primary interest are presented in Table 5. 14 out of 18 firms were significantly exposed, which exceeded our expectations based on past literature. More specifically, 10 of the 13 firms (77%) without foreign income were significantly exposed to exchange rate risks, while 4 out of 5 (80%) firms with foreign income were significantly exposed.

b1e3 Q2e4 þ b4 RSTt þ b5 RDSt þ b6 CCIt

 X      þ b8e10 f DEtn þ b11 f DEt þ b7 f DEtþ þ   X  b12e14 f DEtn þ þ 3i;t n ¼ 1; 2; 3 t ¼ 1; .; T ð11Þ

 where CFi,t is firm i’s cash flows at quarter t, f(DEþ t ) and f(DEt ) respectively, the hyperbolic sine functions of positive and negative P þ  ) þ f(DEtn ), the exchange rate returns at quarter t, and f(DEtn

Table 5 Operating exposure of tourism-related firms.

b7

ID

b8

b9

b10

b11

b12

b13

(0.8) 0.02 (0.7) 0.16 (0.5) 0.13 (0.1) 0.01 (0.2) 0.08 (0.1) 0.49 (0.1) 0.36 (0.1) 0.05 (0.0) (0.5) 3.75 (0.9) 5.92c (0.9) (30.4) 34.15 (22.8) 18.35 (20.9) 0.94 (2.4) 1.72 (3.6) 1.60 (4.3) c 0.06 (0.1) 0.19c (0.1) (0.3) 0.43 (0.4) 0.10 (0.3) 0.21 (0.1) (0.1) 0.19 (0.2) 0.01 (0.2) 0.01 (0.0) 0.03 (0.0) 0.02 (0.0) (0.1) 0.26b (0.1) 0.10 (0.1) 0.03 (0.0) 0.03 (0.0) 0.03b (0.0) (0.9) 2.32a (0.5) 1.00 (0.8) 0.02 (0.1) 0.04 (0.2) 0.50a (0.1) (0.5) 0.72 (0.6) 0.69 (0.4) 0.07c (0.0) 0.03 (0.0) 0.00 (0.1) (0.2) 0.24 (0.2) 0.06 (0.2) 0.02 (0.0) 0.02 (0.0) 0.01 (0.0) (0.3) 0.55 (0.4) 0.17 (0.3) 0.01 (0.0) 0.07 (0.1) 0.07 (0.1) (0.3) 2.55 (1.4) 4.17 (2.6) 0.32 (0.2) 0.23 (0.1) 0.05 (0.0) (1.0) 0.06 (1.4) 0.07 (1.5) 0.14 (0.3) 0.24 (0.2) 0.01 (0.2) 0.03 (0.3) 0.18 (0.4) (1.7) 0.27 (1.5) 0.81 (2.3) 0.50c (0.3) (4.4) 5.28 (5.8) 11b (4.1) 0.94 (0.6) 0.34 (0.6) 1.06 (1.0) 0.29 (0.3) 0.30 (0.2) 0.26 (0.2) (1.1) 1.38 (1.3) 2.17c (1.2) (5.4) 7.31 (5.7) 3.84 (2.9) 0.26 (0.6) 1.29 (0.8) 1.34b (0.6) (2.3) 3.58 (3.6) 2.47 (4.0) 0.47 (0.6) 0.02 (0.4) 0.11 (0.3) (18.6) 109.8c (14.3) 36.07 (8.2) 0.45 (0.8) 1.70 (0.9) 0.72 (0.6) P þ þ þ ) þ b f( D E ) þ b f( D E ) þ b f( Coefficients from regression: CFi,t ¼ b0 þ b1e3Q2e4 þ b4RSTt þ b5RDSt þ b6CCIt þ b7f(DEþ t 8 t1 9 t2 10 DEt3) þ    b11f(DE t ) þ b12f(DEt1) þ b13f(DEt2) þ b14f(DEt3) þ 3i,t. a Significant at 1% level. b Significant at 5% level. c Significant at 10% level. d Firms with foreign income. Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm

1 2 3d 4 5 6 7d 8 9 10 11 12 13 14d 15 16 17d 18d

0.00 1.13 0.00 0.59 0.33 0.00 0.91 0.84c 0.00b 0.19 1.51c 1.91 0.00 0.00 0.00 0.00 0.00 0.00

(0.0) (0.4) (0.0) (0.4) (0.2) (0.0) (1.4) (0.4) (0.0) (0.3) (0.5) (1.4) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0)

0.82 1.72 14.99 0.52 0.23 0.22c 1.19 0.74 0.19 0.45 0.20 1.30 0.75 3.02 0.28 8.81 0.88 35.79

b14 0.17 0.27 7.21a 0.04 0.02 0.01 0.22a 0.03 0.00 0.02 0.21 0.25c 0.78 1.02 0.47b 0.08 0.05 4.21c

R2

F (0.2) (0.1) (2.2) (0.1) (0.0) (0.0) (0.1) (0.0) (0.0) (0.0) (0.1) (0.1) (0.7) (0.8) (0.2) (0.7) (0.7) (0.6)

a

138.86 230.56c 124.92a 20.35a 1.53 112.63a 53.09a 1.57 168.38a 0.92 15.91b 209.1a 139.70a 7.25a 12.18a 91.65a 326.94a 5585.4b

0.9455 0.9876 0.7810 0.7937 0.5112 0.6692 0.9404 0.6955 0.3319 0.3422 0.9500 0.8128 0.6050 0.5471 0.5135 0.9374 0.9571 0.9991

942

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

Although the number of firms in the sample is limited, exchange rate exposure seems to be prevalent for travel-related firms. The overall model fit and significance appeared to be good for the selected nonlinear specification, reporting an average R2 of 0.74 and only 3 insignificant models based on the F-test. It should be noted that these firms do not necessarily coincide with those without significant exchange rate exposure. Firm 8 was significantly exposed to exchange rates but the overall model was insignificant. The opposite was found for Firm 1 and Firm 17. These results implied that for Firm 8, whereas its operating cash flows were exposed to exchange rate the effects were not very big. Firm 1 and Firm 17’s operating cash flows were affected by cyclical effects and other macroeconomic conditions, as quarter dummies or control variables appeared to be highly significant in the model. 5.2. Effect and asymmetry of operating exposure To determine whether domestic tourism-related firms were positively exposed, the net exposures of the firms were determined by summing the significant coefficients and subtracting the summation of coefficients for depreciation (b11 þ b12 þ b13 þ b14) from that for appreciation (b7 þ b8 þ b9 þ b10). A difference greater than zero signaled negative exposure, whereas a difference smaller than zero signaled positive exposure. Concurrent with the theoretical framework, positive exposure indicates that the tourismrelated firms will enjoy increased cash flows from stimulated foreign and domestic demand from depreciation of the dollar. Negative exposure implies that the firms will gain when the dollar gains value, which is unlikely for the subject of the study. The outcome is summarized in Table 6. Among the firms with foreign income, two firms had positive exposure to exchange rates while the other two had negative exposure. Since the firms were found to be operationally hedged prior to analysis, negative exposure was not surprising. All but one of the firms without foreign income showed positive exchange rate exposure as expected. The negative exposure of Firm 15 was somewhat counterintuitive. A closer examination of the SEC filings revealed that the firm, although now discontinued, had operated in foreign oil and gas import business. Oil and gas imports are generally found to react favorably to appreciation of the dollar (Bodnar & Gentry, 1993). Thus, we believe negative exposure was acquired through the implicit diversification of the firm’s business.

Table 6 Sign and asymmetry of operating exposure. ID

b7 þ b8 þ b9 þ b10

Firms without foreign income Firm 2 5.92 Firm 4 e Firm 6 0.48 Firm 8 0.84 Firm 9 0.00 Firm 11 1.51 Firm 12 e Firm 13 e Firm 15 2.17 Firm 16 e Firms with foreign income Firm 3 e Firm 7 2.32 Firm 14 11.00 Firm 18 109.80

b11 þ b12 þ b13 þ b14

Difference

Exposure

Table 7 Contemporaneous vs. lagged exposure coefficients. ID Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm Firm

0.40 0.03 0.07 e e 0.25 0.50 0.47 1.34

5.92 0.40 0.45 0.77 0.00 1.51 0.25 0.50 2.17 1.34

Positive Positive Positive Positive Positive Positive Positive Positive Negative Positive

7.21 0.72 e 4.21

7.21 1.60 11.00 105.59

Positive Negative Positive Negative

P

Coefficients from the regression: CFi,t ¼ b0 þ b1e3Q2e4 þ b4RSTt þ b5RDSt þ b6CCIt þ þ þ   þ b7f(DEþ t ) þ b8f(DEt1) þ b9f(DEt2) þ b10f(DEt3) þ b11f(DEt ) þ b12f(DEt1) þ b13f   ) þ b14f(DEt3 ) þ 3i,t. (DEt2

Contemporaneous

Lagged

e e 0.21 e e 0.77 0.00 1.51 e 0.50 e e e e

5.92 7.21 0.19 0.45 3.04 e e e 0.25 e 11 1.7 1.34 105.59

Contemporaneous exposure is calculated as: (b7  b11), lagged exposure is calculated as: (b8 þ b9 þ b10)  (b12 þ b13 þ b14). a Firms with foreign income.

Table 6 also clearly illustrates the asymmetry of the exposure coefficients. While many of firms are significantly exposed to only one direction, exposures to two directions of innovation seem to differ considerably in size. The results support the theoretical reasoning that response to international demand is asymmetric to the changes in exchange rates. 5.3. Non-contemporaneous effect of exchange rate As suspected, tourism-related firms showed significant lagged exposures to exchange rate changes. Table 7 compares the sum of contemporaneous and lagged exposure coefficients. Since part of the demand response is likely to occur after the exchange rate change, many firms displayed cash flow sensitivity to past changes of the exchange rate. These results are consistent with the survey results of the Office of Travel & Tourism Studies (2008). 5.4. Financing, investing, and total cash flow exposure Excluding Firms 10 and 17, all firms showed significant exposure on at least one of their cash flows. Table 8 illustrates the presence and sign of exposure for individual cash flows for these firms. For Firms 10 and 17, given that total cash flow is the sum of three

Table 8 Investing, financing, and total cash flow exposures. ID

e

2 3a 4 6 7a 8 9 11 12 13 14a 15 16 18a

ICF

FCF

TCF

Firms without foreign income Firm 1 ns Firm 2 Positive Firm 4 Positive Firm 5 ns Firm 6 Positive Firm 8 Positive Firm 9 Positive Firm 11 Positive Firm 12 Positive Firm 13 Positive Firm 15 Negative Firm 16 Positive

OCF

Positive ns Positive Negative ns Positive Positive ns Negative ns ns Negative

Negative ns ns ns Negative ns Positive ns Negative Positive ns Positive

Negative Positive ns Negative Negative Positive Positive ns ns ns ns Negative

Firms with foreign income Firm 3 Positive Firm 7 Negative Firm 14 Positive Firm 18 Negative

Negative ns ns ns

Negative Positive ns Negative

ns ns ns ns

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

different types of cash flows, it was intuitively appealing that their total cash flow was not significantly exposed since none of the cash flows from the three types of firm activities were exposed. It was observed that the exposure of operating cash flows necessarily leads to exposure of investing and financing cash flows. Higher average R2 (74%) for the OCF models also supported this notion (FCF and ICF: 60% and 61%). The investing cash flows of 7 domestic firms without foreign income were exposed to the TTWI. As the firms decide to invest according to the availability of their current and future funds, investing cash flows may be exposed simultaneously with operating cash flows while others may choose to invest in some internationally-priced asset. Firms 1 and 5 can be characterized as the latter. Their operating cash flows were not exposed, but their investing cash flows showed significant exposure which, in turn, affected their total cash flows. The financing cash flows capture a firm’s financing activities to assist the firm’s operations and investment. Therefore, the financial exposure of the domestic firms may be due to the fact that the firm’s operating or investing cash flows were significantly exposed to exchange rate changes. As a firm moves to stabilize its cash flows, the firm is likely face financial burden in order to complement the shortage (or excess) of funds from operation or investment. In this sense, the lack of exposure in the total exposure estimates does not imply that the firms are unaffected by the exchange rate movements, but rather are experiencing volatility in operating or investing cash flows but stabilizing or perhaps hedging them through financing activities. No firm had significant exposure on its financing cash flow alone, supporting the idea above. For Firms 1, 3, 6, 7, 12, and 16 the financing cash flows were exposed in the opposite direction of the operating and/or investing cash flows, which is consistent with Bartram’s (2008) results. As a result, the exposure of total cash flows was insignificant or had a different sign. Firms 2, 5, 8, and 9, did not display financial exposure that complemented (has different signs from) the operating and/or investing activities, and instead carried the effects of operating exposure on their total cash flows. Alternatively, the cash flow volatility arising from exchange rate changes may not be considerable for the firms’ total cash flows, and the firms may not choose to assume the financial position to stabilize. Firms 4, 11, 13, 14, 15, and 18 seemed to fall into this category. Although their total cash flow exposure was insignificant, the exposure of their operation was nevertheless significant. This suggested that the stability of the total cash flows could be improved by financing activities that complement operating activities in terms of exchange rates. It should be noted that while none of the total cash flows of the firms with foreign income were exposed, 7 out of 12 firms without foreign income were exposed to exchange rate risks. Among these firms, Firms 2, 5, 8, and 9 carried the operating or investing exposure on their total cash flow exposure. This may be an indication that the domestic firms may not be aware of their sources of exposure, or as purely domestic firms they may have a limited ability or intention to hedge. Another implication for firms with foreign income would be that the exposure, although significant, may be operationally hedged with even a small percentage of foreign income or awareness of foreign currency risk. Finding a hedged position for domestic firms may not be an easy task, but the results support that operational hedging is viable in mitigating the risk at the corporate cash flow level. 6. Conclusions Based on the theory of exposure by Adler and Dumas (1984), which stated that a firm’s value may be sensitive to currency value changes even without any foreign currency accounts, the current

943

study examined the exchange rate exposure of US tourism-related firms. As expected, a significant percentage (78%) of the tourismrelated firms was exposed, in terms of their operation, to changes in exchange rates. Even when considering only the firms without any foreign income, practically the proportion of exposed firms did not change. Regarding its form, the results supported the idea that exposure can be nonlinear, asymmetric, and lagged, as the source of exposure is primarily from demand changes resulting from shifted purchasing power parities. A considerable percentage of domestic firms (58%) without foreign income were also exposed to exchange rate risk in terms of their financing and investing activities, suggesting that as a firm’s activities are interconnected, the impact of exchange rate on a firm’s operations can influence all activities of the firm. It was also noted that among the exposed firms that partially internationalized firms (domestic firms with foreign income) did not have any significant exposure on their total cash flows, but approximately half of the purely domestic firms (domestic firms without any foreign income) were significantly exposed in terms of their total cash flows. Several important implications can be drawn from these findings. This study found that even domestic firms are not free from risk of exchange rate movements. Accordingly, firms in tourismrelated industries need to make an effort to identify and address the risks associated with exchange rate changes. If a firm’s products are purchased by consumers whose buying power is subject to a foreign exchange rate, or can be substituted with any good priced by foreign currency, the lurking effect of exchange rates may significantly influence corporate cash flows. In this light, for tourism-related firms business diversification seems to be of considerable importance in order to operationally hedge against exchange rate risks. Internationalizing and diversifying into other sectors may introduce even more complicated exposure (as in the case of Firm 13), however, since pass-through is inevitable and the ability to price-to-market is limited, without diversification the firms seem to have a few options. Active targeting and concentration of marketing efforts on the segments insensitive to exchange rates (i.e. Japan; Marvel & Johnson, 1997) may also be a desirable approach, but may not be feasible for all firms and products. Along these lines, constructing a hedge against currency value deviation among tourist markets by using historical covariances (Jang, 2004; Jang & Chen, 2008) may be an attainable countermeasure as well. Foreign inputs should also be considered where commercially feasible; some potential examples for the subject of this study include overseas call centers or outsourcing of corporate website maintenance. The presence of nonlinear, asymmetric, and lagged exposure for these firms suggests that identification and hedging efforts of the firms through linear, symmetric and contemporaneous analysis and tools may achieve only limited success. Specifically, estimation of exposure or hedging through financial derivatives solely on a concurrent basis may not necessarily yield desirable results for the firms, because the earnings may reflect past exchange rates. Yet, one advantage of these tourism-related firms may be that some lead time exists for the effect of exchange rates to be translated into their income statements, thereby allowing them to prepare for inadequate cash flows in advance. Some of the domestic firms without foreign income carried the operating exposure onto their total cash flows. This may be an indication that some of the domestic firms in the travel and tourism industry may not be aware of their businesses’ exposure to exchange rates, have limited ability to operationally or financially hedge, accept such exposure as given due to the nature of exchange rates as a macroeconomic phenomenon, or assume a combination of these positions.Nevertheless, the issue must be carefully addressed, as cross-elasticity of demand implies that local demand is subject to the

0.78

0.53 37.85a

0.99 830.61b

124.92a

0.90

0.97 69.6c

6.86

0.99

0.73 a

230.56c

0.60 33.33a

27.35

0.71 3.76c

0.95 138.86a

93.55 (71.2) 201.4 (138.7) 7.21a (2.2) 2.85 (5.9) 1.60 (4.3) 6.56 (12.0) 1.72 (3.6) 17.20c (8.0) 0.94 (2.4) 18.02 (15.1) 18.35 (20.9) 35.53 (27.6) 34.15 (22.8) 35.39 (40.9) 14.99 (30.4) 3.98 (60.4) 0.00 (0.0) 0.00 (0.0) 19,042b (7395) 21,686 (27,842) 158.87 (352.9) 823.69 (1285.3) 90.85 (215.5) 149 (571.6) 283b (120.7) 128.38 (277.7) 180.3c (96.4) 742.55b (272.1) 21.03 (92.3) 160.81 (263.0) FCF

TCF

ICF

Firm 3 OCF

32.10 (6.3) 20.05 (5.2) 1.76 (4.9) 13.82 (3.8) 0.27 (0.1) 0.21 (0.0) 0.02 (0.0) 0.08 (0.0) 0.05 (0.0) 0.02 (0.0) 0.02 (0.0) 0.05 (0.0) 0.36 (0.1) 0.25 (0.1) 0.03 (0.1) 0.14 (0.0) 0.49 (0.1) 0.40 (0.1) 0.02 (0.1) 0.11 (0.1) 5.92c (0.9) 2.83 (0.8) 0.58 (0.7) 3.67c (0.6) 3.75 (0.9) 2.88 (0.7) 0.31 (0.7) 1.18 (0.5) 1.72 (0.5) 1.99 (0.4) 0.04 (0.4) 0.23 (0.3) 1.13 (0.4) 2.27 (0.4) 0.01 (0.3) 1.13 (0.3) 3175 (665) 2316 (549) 145 (520) 1005 (404) 3.73 (1.6) 4.48 (1.3) 1.77 (1.3) 1.01 (1.0) 51.20 (9.7) 30.14 (8.0) 2.74 (7.6) 23.79 (5.9) 22.41 (4.1) 13.70 (3.4) 1.09 (3.2) 9.80 (2.5) 37.40 (7.5) 24.46 (6.2) 1.93 (5.9) 14.87 (4.5) 12.66 (2.4) 9.19 (2.0) 0.28 (1.9) 3.75 (1.5) FCF

TCF

ICF

Firm 2 OCF

15.32a (5.0) 1.23 (5.0) 5.47 (4.2) 0.31b (0.1) 0.00 (0.1) 0.48a (0.1) 0.28c (0.1) 0.24 (0.3) 0.03 (0.3) 0.16 (0.1) 0.18 (0.2) 0.03 (0.2) 0.26b (0.1) 0.00 (0.2) 0.13 (0.2) 1.62 (1.0) 0.56 (0.7) 2.02c (1.0) 2.55a (0.7) 1.36 (0.8) 1.17 (0.8) 2.23a (0.7) 2.46c (1.3) 0.59 (0.6) 0.00b (0.0) 0.00 (0.0) 0.00 (0.0) 379 (332) 688 (505) 255 (562) 59.41b (24.9) 15.59 (18.9) 23.30 (32.5) 36.10b (12.9) 8.54 (7.7) 15.64 (13.9) 2.82 (4.8) 6.49c (3.7) 3.56 (5.2) 1.27 (3.8) 15.01a (5.2) 31.81a (7.5) (3.5) 6.70 (5.2) 3.86 (4.7) 9.23 (7.7) FCF

45.54a (5.7)

7.24b (3.3)

11.9c (6.7)

20.52c (10.5)

564c (280)

0.00 (0.0)

0.82 (0.8)

0.02 (0.7)

0.16 (0.5)

0.13 (0.1)

0.01 (0.2)

0.08 (0.1)

0.17 (0.2)

22.02a (3.8)

F CONS f  (DEt3 ) f  (DEt2 ) f  (DEt1 ) f (DE t ) f þ (DEt3 ) f þ (DEt2 ) f þ (DEt1 ) f (DEþ t ) CCI RDS RST Q4 Q3

19.79a

Complete regression results

Firm 1 OCF

Appendix.

Q2

risk as well. For example, a firm may decide not to hedge the operating risks incurred by international demand and concentrate on local demand; however, appreciation of the home currency will not only depress international demand but also stimulate local demand to substitute away to foreign products from domestic products. This study makes three unique contributions to the literature. Most conspicuously, this study revealed that travel services and recreation industries constitute a unique case for exchange rate exposure. This study also found that a relatively high percentage of firms significantly exposed to exchange rate risks, even compared to preceding studies on multinationals (Bartram, 2007). Consistent with the theoretical assumption that exposure originates from the responsiveness of demand, nonlinear, asymmetric, and lagged forms of exposure were found among the sample firms. Furthermore, since identifying exposure to exchange rate requires choosing which exchange rates firms are influenced by, the current study developed and used the travel and tourism-weighted index (TTWI). The TTWI was found to significantly differ from the Broad REER, published by US Federal Reserve that weighs the currencies by general US trades, in 3 out of 32 quarters. Identification of exposure is often difficult and puzzling (Bartram, 2007), and firms in tourism and related industries may significantly benefit by utilizing the index for future analytical purposes. The study is somewhat limited in its generalizability and applicability due to the lack of publicly traded firms and the availability of data in relevant industries, as the number of firms is relatively small. Even though the selected nonlinear specification was theoretically the most appealing, as Bartram (2004) and Priestley and Ødegaard (2006) have noted, nonlinear specification of the exchange rate is a complex function that incorporates firmspecific characteristics and various macroeconomic conditions. Also, as the purpose of this study was mainly to identify the presence and sign of exposure, the standardized research model was applied uniformly for all firms. We suggest that future studies attempting to further investigate exposure should improve the model by adjusting it to their specific needs. The existence and characteristics of foreign exchange exposure is dependent on the respective firm’s operating, investing, and financing profiles. Therefore, constructing a new weighted exchange rate index by using specific currencies of interest, modifying the lag structure for the exchange rate variables, or controlling for other macroeconomic, industry, or firm-specific effects are all viable options to improve model performance and obtain more effective results. Nevertheless, the importance of the study’s findings regarding the nature and form of exchange rate exposure in tourism-related firms remains unchanged. The understanding of exchange rate exposure may be broadened by conducting further research on the characteristics of firms that magnify or alleviate exposure. As internationalization, business diversification, and franchising are not irrelevant to this phenomenon, the risk-averse behavior of firms may be jointly studied. Approximating the nonlinearity of international tourism demand, with respect to exchange rate changes, will further extend the efficiency of the exposure models. In order to better understand a firm’s macroeconomic risks, interest rate exposure of the firm and its relationship with exchange rate exposure can be taken into account as well. As long as currencies are translated among one another, research efforts in tourism academia on this topic must continue.

R2

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

SORT

944

ICF TCF Firm 4 OCF FCF ICF TCF Firm 5 OCF

ICF TCF Firm 6 OCF FCF ICF TCF Firm 7 OCF FCF ICF TCF Firm 8 OCF FCF ICF TCF Firm 9 OCF FCF

620.3b (210.8) 58.09 (142.1)

35.89 (209.2) 190b (72.4)

273.72 (327.4) 33.91 (245.9)

793.73 (974.0) 188.83 (297.1)

33,870 (23,953) 6858 (11,816)

0.00 (0.0) 0.00 (0.0)

21.70 (51.0) 10.69 (21.9)

11.48 (37.8) 10.24 (20.6)

43.09 (26.9) 10.80 (16.6)

21.72 (13.5) 2.76 (3.4)

13.36 (8.1) 2.12 (2.9)

9.62 (9.3) 1.47 (2.5)

13.8c (6.7) 3.78 (4.2)

183.05 (135.4) 75.22 (90.2)

5.59a

0.61

11.63a

0.45

0.28 (1.6) 0.58 (2.2) 0.46c (0.2) 1.32 (2.8)

2.97 (2.1) 0.15 (1.9) 0.14 (0.3) 2.98 (3.4)

1.64 (1.3) 0.13 (1.8) 0.32 (0.3) 2.09 (2.0)

11.4c (4.5) 2.31 (3.7) 0.35 (0.3) 8.70b (2.6)

8.24 (5.9) 0.55 (7.4) 0.60 (1.3) 7.10 (6.1)

186 (173) 187 (193) 11 (16) 10 (306)

0.59 (0.4) 0.16 (0.6) 0.21c (0.1) 0.21 (0.6)

0.52 (0.3) 0.37 (0.6) 0.11 (0.1) 0.04 (0.5)

0.43 (0.4) 0.29 (0.3) 0.03 (0.0) 0.11 (0.4)

0.10 (0.3) 0.11 (0.4) 0.05 (0.1) 0.04 (0.4)

0.21c (0.1) 0.10 (0.1) 0.03a (0.0) 0.08 (0.0)

0.06 (0.1) 0.05 (0.1) 0.02 (0.0) 0.00 (0.1)

0.19c (0.1) 0.08 (0.1) 0.02c (0.0) 0.09 (0.1)

0.04 (0.1) 0.02 (0.1) 0.00 (0.0) 0.02 (0.1)

6.51b (2.0) 3.06 (2.0) 0.61b (0.2) 2.85 (1.4)

20.35a

0.79

4.04c

0.47

a

0.89

a

43.76

0.72

0.37 (0.7) 0.10 (0.6) 0.13 (0.4) 0.34 (0.4)

1.02 (0.9) 0.76 (0.5) 1.10 (0.9) 0.83c (0.5)

0.21 (0.6) 0.21 (0.5) 0.90 (0.9) 0.91c (0.5)

1.61 (1.8) 1.89 (1.1) 0.83 (1.5) 1.12 (1.7)

7.05 (4.0) 2.56 (2.2) 5.29c (2.6) 0.80 (2.5)

135 (81) 9 (39) 116c (56) 10 (51)

0.33 (0.2) 0.18b (0.1) 0.11 (0.2) 0.04 (0.1)

0.23 (0.1) 0.11 (0.1) 0.09 (0.1) 0.03 (0.1)

0.19 (0.2) 0.00 (0.1) 0.21 (0.1) 0.02 (0.1)

0.01 (0.2) 0.08 (0.1) 0.02 (0.1) 0.08 (0.1)

0.01 (0.0) 0.00 (0.0) 0.00 (0.0) 0.02 (0.0)

0.03 (0.0) 0.05b (0.0) 0.00 (0.0) 0.03 (0.0)

0.02 (0.0) 0.01 (0.0) 0.03 (0.0) 0.00 (0.0)

0.02 (0.0) 0.02 (0.0) 0.03 (0.0) 0.03c (0.0)

0.78 (1.0) 0.02 (0.7) 0.07 (0.7) 0.83 (0.8)

1.53

0.51

0.63 (0.7) 1.43 (1.2) 0.20 (1.7) 2.26 (2.7)

0.64 (1.0) 2.76 (1.9) 7.85 (6.1) 5.73 (6.5)

2.22b (0.8) 1.15 (1.9) 0.01 (2.3) 3.38 (2.9)

0.20 (1.8) 0.51 (4.3) 0.45 (5.5) 1.16 (8.5)

1.11 (2.3) 6.19 (7.0) 4.63 (8.3) 2.68 (9.6)

1 (74) 1 (166) 240 (257) 239 (325)

0.00 (0.0) 0.00 (0.0) 0.00 (0.0) 0.00 (0.0)

0.22c (0.1) 0.34 (0.3) 1.11 (0.8) 0.99 (0.9)

0.26b (0.1) 0.27 (0.4) 0.65 (0.7) 0.64 (0.5)

0.10 (0.1) 0.31 (0.3) 0.01 (0.4) 0.23 (0.3)

0.03 (0.0) 0.11a (0.0) 0.05 (0.1) 0.03 (0.1)

0.03 (0.0) 0.08 (0.1) 0.02 (0.1) 0.07 (0.1)

0.03b (0.0) 0.05 (0.1) 0.12 (0.1) 0.19b (0.1)

0.01 (0.0) 0.09 (0.1) 0.19 (0.1) 0.11 (0.2)

2.35 (3.9) 0.12 (9.6) 0.54 (10.6) 2.78 (2.9)

11.03b (4.6) 19.10 (13.2) 5.98 (12.4) 2.09 (2.8)

3.20 (4.3) 10.02 (7.7) 5.46 (6.0) 1.36 (3.2)

5.06 (6.5) 4.74 (15.8) 7.71 (19.7) 2.09 (4.1)

5.03 (20.5) 25.59 (40.9) 29.86 (29.6) 0.75 (10.7)

1054a (442) 1484 (1121) 509 (920) 79 (156)

0.91 (1.4) 3.10 (3.1) 3.56 (2.4) 0.46 (0.4)

1.19 (0.9) 1.15 (1.7) 0.00 (1.4) 0.05 (0.4)

2.32a (0.5) 3.70b (1.3) 1.26 (1.2) 0.12 (0.3)

1.00 (0.8) 2.32 (2.5) 2.76 (1.8) 0.56 (0.4)

0.02 (0.1) 0.17 (0.2) 0.15 (0.1) 0.01 (0.0)

0.04 (0.2) 0.79 (0.6) 0.63 (0.5) 0.12 (0.1)

0.50a (0.1) 0.11 (0.3) 0.30 (0.2) 0.10 (0.1)

0.30 (0.9) 0.30 (0.9) 0.03 (0.1) 0.03 (1.1)

1.67 (0.9) 0.14 (0.6) 0.02 (0.0) 1.79c (0.9)

2.09 (1.2) 0.21 (1.3) 0.02 (0.1) 2.32 (1.3)

4.74 (3.0) 0.97 (1.6) 0.01 (0.1) 5.70 (2.2)b

2.99 (6.0) 0.79 (3.1) 0.10 (0.2) 3.69 (3.7)

43 (52) 18 (46) 2 (4) 60 (67)

0.84c (0.4) 0.08 (0.1) 0.03 (0.0) 0.74c (0.4)

0.74 (0.5) 0.10 (0.2) 0.02 (0.0) 0.82 (0.4)

0.72 (0.6) 0.06 (0.2) 0.00 (0.0) 0.78 (0.7)

0.69 (0.4) 0.16 (0.3) 0.00 (0.0) 0.53 (0.4)

0.07c (0.0) 0.01 (0.0) 0.00c (0.0) 0.08 (0.0)

0.03 (0.0) 0.02 (0.0) 0.00 (0.0) 0.05 (0.0)

0.72 (1.1) 0.92 (0.6)

1.46 (1.5) 1.69b (0.6)

1.25 (1.7) 1.35b (0.5)

0.99 (2.0) 2.92a (0.8)

0.36 (3.5) 0.49 (1.9)

7 (56) 91 (63)

0.00b (0.0) 0.00 (0.0)

0.19 (0.2) 0.10 (0.1)

0.24 (0.2) 0.07 (0.1)

0.06 (0.2) 0.06 (0.1)

0.02 (0.0) 0.05b (0.0)

0.02 (0.0) 0.01 (0.0)

55.04

a

0.64

5.66a

0.55

2.74b

0.42

1.29 (0.8) 0.75 (1.4) 0.67 (2.0) 0.13 (2.5)

112.63a

0.67

23.14a

0.34

1.22

0.41

c

3.76

0.28

0.22a (0.1) 0.05 (0.3) 0.16 (0.2) 0.01 (0.1)

5.82 (4.3) 1.61 (6.4) 6.58 (6.6) 0.86 (1.9)

53.09a

0.94

a

61.15

0.72

1.04

0.56

12.83a

0.62

0.00 (0.1) 0.00 (0.0) 0.00 (0.0) 0.00 (0.0)

0.03 (0.0) 0.01 (0.0) 0.00 (0.0) 0.04 (0.0)

1.44 (0.8) 0.53 (0.5) 0.07 (0.1) 1.89b (0.7)

1.57

0.70

13.99a

0.49

a

0.71

c

0.70

0.01 (0.0) 0.05c (0.0)

0.00 (0.0) 0.01 (0.0)

0.36 (0.9) 1.25a (0.3)

168.38a

0.33

a

0.69

14.55

30.09

3.35

137.11

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

FCF

125.41 (239.9) 14.37 (107.9)

(continued on next page) 945

ICF TCF Firm 10 OCF FCF ICF TCF

FCF ICF TCF Firm 12 OCF FCF ICF TCF Firm 13 OCF FCF ICF TCF Firm 14 OCF FCF ICF TCF Firm 15 OCF

R2

Q2

Q3

Q4

RST

RDS

CCI

f (DEþ t )

f þ (DEt1 )

f þ (DEt2 )

f þ (DEt3 )

f (DE t )

f  (DEt1 )

f  (DEt2 )

f  (DEt3 )

CONS

F

0.38 (0.5) 0.57 (1.1)

0.04 (0.8) 0.28 (1.1)

0.32 (0.3) 0.23 (1.5)

2.38 (1.5) 0.45 (1.4)

1.04 (3.2) 0.91 (1.9)

43 (68) 56c (31)

0.00a (0.0) 0.00b (0.0)

0.19 (0.1) 0.10 (0.1)

0.13 (0.1) 0.03 (0.1)

0.02 (0.1) 0.10 (0.1)

0.02c (0.0) 0.06b (0.0)

0.01 (0.0) 0.03 (0.0)

0.01 (0.0) 0.05c (0.0)

0.01 (0.0) 0.01 (0.0)

0.66 (0.5) 0.23 (0.7)

0.19 (0.8) 0.60 (1.4) 0.01 (0.0) 0.40 (1.4)

2.15 (1.3) 0.67 (1.5) 0.03 (0.0) 1.45 (1.7)

1.49 (1.5) 3.06 (3.8) 0.04 (0.1) 1.54 (2.8)

2.52 (2.4) 6.73 (5.7) 0.16 (0.1) 9.09 (6.0)

5.08 (4.7) 8.06 (7.9) 0.18 (0.2) 2.79 (6.6)

178 (139) 142 (170) 2 (3) 38 (93)

0.19 (0.3) 0.52 (0.5) 0.01 (0.0) 0.32 (0.4)

0.45 (0.3) 0.73 (0.5) 0.01 (0.0) 0.27 (0.3)

0.55 (0.4) 0.89 (0.7) 0.01 (0.0) 0.33 (0.4)

0.17 (0.3) 0.41 (0.5) 0.00 (0.0) 0.23 (0.4)

0.01 (0.0) 0.11 (0.1) 0.00 (0.0) 0.09 (0.1)

0.07 (0.1) 0.06 (0.1) 0.00 (0.0) 0.01 (0.1)

0.07 (0.1) 0.07 (0.1) 0.00 (0.0) 0.01 (0.1)

0.02 (0.0) 0.06 (0.0) 0.00 (0.0) 0.04 (0.0)

1.62 (1.7) 3.71 (2.8) 0.06 (0.1) 2.03 (1.7)

12.01 (5.5) 0.95 (13.2) 3.58 (11.6) 7.48 (5.9)

20.82 (12.3) 5.42 (29.6) 19.42 (26.8) 6.83 (13.9)

15.86 (8.6) 6.45 (20.7) 12.59 (18.5) 9.72 (9.4)

34.66 (18.7) 19.08 (44.2) 38.22 (40.3) 15.52 (20.8)

9.06c (3.6) 3.62 (15.3) 3.64 (11.1) 9.09 (4.8)

2004 (1011) 20 (2352) 1335 (2101) 689 (1156)

1.51c (0.5) 0.38 (2.2) 0.29 (1.3) 0.83 (0.9)

0.20 (0.3) 1.47 (1.7) 0.00 (0.7) 1.67 (1.5)

2.55 (1.4) 0.25 (2.8) 0.53 (3.0) 1.77 (1.8)

4.17 (2.6) 0.53 (6.5) 3.20 (5.8) 0.44 (3.0)

0.32 (0.2) 0.08 (0.4) 0.18 (0.4) 0.22 (0.2)

0.23 (0.1) 0.08 (0.4) 0.20 (0.3) 0.05 (0.2)

0.05 (0.0) 0.05 (0.1) 0.00 (0.1) 0.00 (0.1)

0.21 (0.1) 0.13 (0.2) 0.04 (0.2) 0.03 (0.1)

22.87 (11.8) 0.01 (26.5) 12.84 (24.8) 10.03 (13.9)

24.99a (4.8) 34.55 (23.8) 28.44 (30.3) 18.87a (6.4)

17.36a (5.8) 12.90 (27.7) 27.34 (33.5) 2.92 (7.3)

8.78 (11.5) 5.06 (21.1) 12.26 (28.0) 8.53 (6.4)

4.65 (11.5) 27.79 (33.8) 19.06 (39.4) 4.07 (10.3)

41.63a (13.1) 74.42 (49.3) 26.22 (60.1) 6.57 (13.6)

19 (368) 903 (1842) 956 (2144) 72 (440)

1.91 (1.4) 11.16c (6.3) 8.68 (7.0) 0.57 (1.4)

1.30 (1.0) 2.43 (5.1) 3.60 (6.1) 0.13 (1.0)

0.06 (1.4) 0.85 (3.4) 0.87 (4.3) 0.04 (0.6)

0.07 (1.5) 6.25 (5.5) 6.30 (5.9) 0.13 (1.5)

0.14 (0.3) 2.00b (0.8) 1.90c (1.0) 0.05 (0.2)

0.24 (0.2) 1.19 (1.2) 1.36 (1.4) 0.06 (0.2)

0.01 (0.2) 0.08 (0.9) 0.22 (1.1) 0.15 (0.2)

0.25c (0.1) 2.30 (1.6) 2.39 (1.9) 0.33 (0.3)

9.14 (8.3) 10.87 (14.2) 6.13 (18.9) 7.86 (5.3)

209.1a

2.48 (11.4) 29.15 (24.2) 22.83 (32.7) 3.84 (26.8)

23.90 (17.2) 42.05b (17.0) 25.01 (42.8) 40.94 (29.8)

19.15 (11.3) 33.06 (22.7) 52.56 (56.6) 0.36 (37.9)

0.69 (26.0) 88.81a (23.2) 66.86 (57.4) 22.64 (30.8)

9.39 (26.4) 12.71 (60.9) 40.89 (52.3) 62.99 (74.5)

88 (1240) 141 (1375) 689 (2334) 742 (2215)

0.00 (0.0) 0.00a (0.0) 0.00 (0.0) 0.00 (0.0)

0.75 (1.7) 2.74 (2.3) 0.54 (3.9) 2.95 (5.3)

0.27 (1.5) 3.95b (1.8) 2.59 (6.1) 1.63 (4.3)

0.81 (2.3) 5.33b (2.0) 1.10 (4.8) 5.04 (5.7)

0.50c (0.3) 0.69 (0.4) 0.30 (1.0) 0.49 (0.9)

0.03 (0.3) 0.13 (0.6) 0.20 (0.8) 0.11 (1.1)

0.18 (0.4) 0.85c (0.4) 1.05 (0.8) 0.01 (0.9)

0.78 (0.7) 0.12 (0.2) 0.31 (0.7) 0.35 (0.4)

57.59a (12.7) 23.41c (13.2) 54.37 (31.7) 26.63 (19.3)

30.64b (13.7) 26.20 (22.8) 10.52 (26.7) 46.32b (21.0)

8.25 (17.6) 19.07 (18.3) 19.57 (26.9) 30.39 (28.5)

43.46c (21.4) 12.92 (18.8) 11.71 (24.3) 44.67 (36.3)

53.49 (49.0) 50.36 (79.1) 60.33 (55.0) 57.20 (82.3)

156.9c (81.9) 114.39 (180.5) 117.10 (84.5) 74.61 (172.1)

2274 (1523) 1284 (1679) 2306 (2649) 1316 (1851)

0.00 (0.0) 0.00 (0.0) 0.00 (0.0) 0.00 (0.0)

3.02 (4.4) 2.84 (2.6) 19.93 (12.2) 14.07 (9.7)

5.28 (5.8) 5.51 (3.9) 8.49 (8.1) 2.31 (5.5)

10.99b (4.1) 4.09 (4.0) 4.89 (6.7) 11.79 (8.7)

0.94 (0.6) 0.29 (0.4) 1.53 (1.0) 0.30 (0.6)

0.34 (0.6) 0.26 (0.8) 1.58 (1.3) 1.66 (1.1)

1.06 (1.0) 0.37 (0.8) 2.18 (1.6) 0.75 (0.9)

1.02 (0.8) 0.61 (0.7) 1.07 (0.9) 0.67 (0.9)

24.05 (27.7) 12.27 (17.9) 73.82c (40.0) 62.04b (26.6)

3.98 (5.9)

9.17 (10.7)

4.37 (6.5)

8.49 (18.5)

11.37 (19.0)

234 (627)

0.00 (0.0)

0.28 (1.1)

1.38 (1.3)

2.17c (1.2)

0.29 (0.3)

0.30 (0.2)

0.26 (0.2)

0.47b (0.2)

26.29a (8.4)

2931a

0.73

21.00a

0.58

0.92

0.34

0.33

0.33

0.43

0.30

1.30

0.43

15.91b

0.95

a

0.91

41.88a

0.95

7.81c

0.82

30.12

0.81

11.81a

0.53

b

0.48

14.40a

0.54

139.70a

0.61

3.41

1346.3a

0.75

9.15a

0.37

139.05a

0.52

7.25a

0.55

3.25b

0.32

a

0.59

4.16a

0.60

12.18a

0.51

5.90

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

Firm 11 OCF

946

Appendix (continued) SORT

FCF ICF TCF Firm 16 OCF FCF ICF TCF

FCF ICF TCF Firm 18 OCF

15.85 (32.0) 5.16 (36.8) 19.86 (15.6)

19.74 (17.1) 26.75 (29.8) 42.12 (30.0)

8.09 (51.0) 29.03 (52.5) 12.45 (30.3)

129.81 (87.6) 116.17 (96.2) 2.27 (42.0)

2311 (2701) 3590 (2565) 1045 (1991)

0.00 (0.0) 0.00 (0.0) 0.00 (0.0)

3.71 (7.0) 5.48 (6.4) 2.06 (2.9)

5.50 (4.5) 0.45 (6.7) 7.33 (4.9)

0.48 (3.0) 3.82 (5.6) 6.47 (5.6)

0.46 (0.5) 1.44 (0.8) 0.70 (0.8)

0.85 (1.4) 0.84 (1.4) 0.30 (0.5)

0.30 (0.8) 0.35 (0.8) 0.39 (0.9)

2.40 (1.6) 2.20 (1.8) 0.67 (0.7)

14.76 (22.7) 8.14 (28.8) 3.38 (22.6)

162.34a (16.5) 416.6a (124.5) 303.2a (99.6) 48.97 (35.6)

218.9a (33.4) 273.1a (61.0) 68.50 (54.6) 14.39 (26.6)

28.81c (15.8) 284b (128.7) 207.3b (96.9) 48.29 (48.7)

11.06 (35.0) 232.61 (211.5) 243.7 (165.3) 22.12 (54.8)

85.74 (80.5) 462.35 (426.1) 460.88 (355.2) 84.27 (71.2)

3814 (2066) 3701 (6353) 5734 (6203) 1782 (2246)

0.00 (0.0) 0.00c (0.0) 0.00 (0.0) 0.00a (0.0)

8.81 (5.4) 44.54a (9.9) 36.01a (11.0) 0.27 (4.9)

7.31 (5.7) 10.77 (12.3) 0.71 (15.9) 4.17 (3.3)

3.84 (2.9) 26.03 (15.8) 31.65b (12.8) 1.79 (4.2)

0.26 (0.6) 3.90a (3.2) 3.69a (3.2) 0.05 (0.9)

1.29 (0.8) 9.09 (3.0) 8.43 (2.8) 0.63 (0.9)

1.34b (0.6) 2.81 (2.2) 4.20c (2.3) 0.05 (0.7)

0.08 (0.7) 3.16 (2.6) 2.46 (2.9) 0.78 (0.5)

100a (16.5) 306.4a (91.1) 202.6b (86.0) 3.66 (26.4)

87.49a (11.4) 19.49 (72.0) 55.77 (68.9) 12.23 (7.5)

181.4a (19.4) 97.37 (327.7) 277.20 (315.5) 1.53 (22.7)

19.23 (16.5) 3.25 (70.5) 39.51 (68.0) 17.04 (13.9)

18.15 (28.8) 262.7 (392.8) 254.77 (368.7) 10.27 (30.3)

39.14 (56.3) 720.30 (1031.9) 681.31 (984.0) 0.16 (63.3)

69 (1590) 455 (15,333) 585 (14,713) 970 (1379)

0.00 (0.0) 0.00 (0.0) 0.00 (0.0) 0.00 (0.0)

0.88 (2.3) 46.85 (47.1) 46.23 (46.2) 1.50 (2.2)

3.58 (3.6) 59.51 (81.3) 62.72 (78.6) 0.37 (5.6)

2.47 (4.0) 7.14 (16.6) 5.31 (17.0) 0.63 (3.0)

0.47 (0.6) 0.98 (3.0) 0.56 (2.8) 0.05 (0.4)

0.02 (0.4) 1.82 (4.1) 1.96 (4.0) 0.13 (0.4)

0.11 (0.3) 5.27 (5.9) 5.25 (5.7) 0.13 (0.4)

0.05 (0.7) 7.80 (10.5) 7.79 (10.3) 0.06 (0.8)

36.79 (21.1) 176.35 (122.9) 133.4 (120.3) 6.16 (15.8)

0.45 (0.8) 8.49b (0.1) 5.83 (3.8) 3.11 (4.8)

1.70 (0.9) 1.87b (0.2) 11.72 (4.4) 11.89 (5.4)

0.72 (0.6) 5.72b (0.1) 1.03 (2.8) 3.98 (3.4)

4.21c (0.6) 3.50b (0.1) 3.68 (3.0) 2.98 (3.7)

132.1b (8.7) 102.6b (1.7) 32.03 (43.2) 202.6 (53.6)

158.78 12,130c 0.00 35.79 109.8c 36.07 25.16 356.2c (31.4) (56.0) (1472) (0.0) (18.6) (14.3) (8.2) (33.9) a b b b b b 442.6 311.1 161.8 639 0.00 101.4 144.8 31.60b FCF (6.5) (6.0) (10.7) (281) (0.0) (3.6) (2.7) (1.6) ICF 177.8 19.57 736.71 20,912 0.00 116.89 89.72 117.2 (167.5) (155.4) (276.8) (7273) (0.0) (92.0) (70.8) (40.5) TCF 289.99 648 1057 8142 0.00 182.45 54.72 49.6 (207.9) (192.8) (343.6) (9025) (0.0) (114.2) (87.8) (50.3) P þ þ þ Coefficients from the regression: CFi,t ¼ b0 þ b1e3Q2e4 þ b4RSTt þ b5RDSt þ b6CCIt þ b7f(DEþ t ) þ b8f(DEt1) þ b9f(DEt2) þ b10f(DEt3) þ    b11f(DE t ) þ b12f(DEt1) þ b13f(DEt2) þ b14f(DEt3) þ 3i,t. Average coefficient of determinant, OCF: 0.7400; FCF: 0.5950, ICF: 0.6088, TCF: 0.6148, total: 0.6397. a Significant at 1% level. b Significant at 5% level. c Significant at 10% level. 41.00 (28.3) 272.4b (5.4) 94.50 (139.8) 218.86 (173.5)

5.39 (11.6) 246.34b (2.2) 178.38 (57.3) 73.35 (71.1)

5.64a

0.42

81.98a

0.48

a

0.41

91.65a

0.94

68.03

a

0.74

5.23

a

0.65

331.11a

0.79

326.94a

0.96

2.47c

0.16

0.47

0.24

46.74

278.41

a

5585.4b

0.51

0.99

a

0.99

48.24

0.96

108.22c

0.96

66,058

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

Firm 17 OCF

22.41 (16.6) 1.81 (14.3) 16.62 (13.4)

947

948

S.K. Lee, SooCheong(Shawn) Jang / Tourism Management 32 (2011) 934e948

References Adler, M., & Dumas, B. (1980). The exposure of long-term foreign currency bonds. The Journal of Financial and Quantitative Analysis, 15, 973e994. Adler, M., & Dumas, B. (1984). Exposure to currency risk: definition and measurement. Financial Management, 13(2), 41e50. Allayannis, G. (1997). The time-variation of the exchange rate exposure: An industry analysis. Working paper. New York, NY: New York University. Allayannis, G., & Ihrig, J. (2001). Exposure and markups. Review of Financial Studies, 14(3), 805e835. Amihud, Y. (1994). Exchange rates and the valuation of equity shares. In Y. Amihud, & R. M. Levich (Eds.), Exchange rates and corporate performance (pp. 49e59). New York, NY: Irwin. Bartov, E., & Bodnar, G. M. (1994). Firm valuation, earnings expectations, and the exchange rate exposure effect. Journal of Finance, 44(5), 1755e1785. Bartov, E., Bodnar, G. M., & Kaul, A. (1996). Exchange rate variability and the riskiness of US multinational firms: evidence from the breakdown of the Bretton Woods system. Journal of Financial Economics, 42(1), 105e132. Bartram, S. M. (2002). The interest rate exposure of nonfinancial corporations. European Finance Review, 6(1), 101e125. Bartram, S. M. (2004). Linear and nonlinear foreign exchange rate exposures of German nonfinancial corporations. Journal of International Money and Finance, 23, 673e699. Bartram, S. M. (2007). Corporate cash flow and stock price exposures to foreign exchange risk. Journal of Corporate Finance, 13, 981e994. Bartram, S. M. (2008). What lies beneath: foreign exchange rate exposure, hedging, and cash flows. Journal of Banking and Finance, 32(8), 1508e1521. Bartram, S. M., & Bodnar, G. M. (2007). The foreign exchange exposure puzzle. Managerial Finance, 33(9), 642e666. Bodnar, G. M., & Gentry, W. M. (1993). Exchange rate exposure and industry characteristics: evidence from Canada, Japan, and the USA. Journal of International Money and Finance, 12(1), 29e45. Bureau of Economic Analysis. (2008). U.S. travel and tourism satellite accounts for 2004e2007. Retrieved from. http://www.bea.gov/scb/pdf/2008/06%20June/ 0608_travel.pdf. Carter, D. A., Rogers, D. A., & Simkins, B. J. (2006). Does hedging affect firm value? Evidence from the US airline industry. Financial Management, 35, 53e88. Chang, C. (2009). To hedge or not to hedge. Cornell Hospitality Quarterly, 50, 301e313. Choi, J. J., & Prasad, A. M. (1995). Exchange risk sensitivity and its determinants: a firm and industry analysis of U.S. multinationals. Financial Management, 24(3), 77e88. Chow, E. H., Lee, W. Y., & Solt, M. E. (1997). The exchange-rate risk exposure of asset returns. Journal of Business, 70(1), 105e123. Christophe, S. E. (1997). Hysteresis and the value of the U.S. multinational corporations. Journal of Business, 70, 435e462. Conference Board, The. (2009, April 20). The Conference Board leading economic index (LEI) for the United States and related composite economic indexes for March 2009. Retrieved from. http://www.conference-board.org/pdf_free/ economics/bci/noConn.pdf. Divisekera, S. (2003). A model of demand for international tourism. Annals of Tourism Research, 30, 31e49. Dumas, B. (1978). The theory of the trading firm revisited. Journal of Finance, 33, 1019e1029. Dumas, B., & Solnik, B. (1995). The world price of foreign exchange risk. Journal of Finance, 50(2), 445e480. Durbin, J. (1970). Testing for serial correlation in least squares regression when some of the regressors are lagged dependent variables. Econometrica, 38, 410e421. Economist, The. (2004, September 30). Gambling goes global. Retrieved from. http://www.economist.com/business/displaystory.cfm?story_id¼3242391. Flood, E., Jr., & Lessard, D. R. (1986). On the measurement of operating exposure to exchange rates: a conceptual approach. Financial Management, 15, 25e37. Garner, C. K., & Shapiro, A. C. (1984). A practical method of assessing foreign exchange rate risk. Midland Corporate Finance Journal, Fall, 6e17. Giddy, I. H., & Dufey, G. (1995). Uses and abuses of currency options. Journal of Applied Corporate Finance, 8(3), 49e57. Goldberg, P. K. (1995). Product differentiation and oligopoly in international markets: the case of the U.S. automobile industry. Econometrica, 63, 891e951. He, J., & Ng, L. K. (1998). The foreign exchange exposure of Japanese multinational corporations. Journal of Finance, 53(2), 733e753. Hua, N., & Upneja, A. (2007). Going international? Important factors executives should consider! International Journal of Contemporary Hospitality Management, 19(7), 537e545.

Jang, S. (2004). Mitigating tourism seasonality: a quantitative approach. Annals of Tourism Research, 31(4), 819e836. Jang, S., & Chen, M.-H. (2008). Financial portfolio approach to optimal tourist market mixes. Tourism Management, 29(4), 761e770. Jang, S., & Tang, C. (2009). Simultaneous impacts of internationalization and financial leverage on profitability. Journal of Hospitality and Tourism Research, 33 (3), 347e368. Jorion, P. (1990). The exchange-rate exposure of U.S. multinationals. Journal of Business, 63(3), 331e345. Jorion, P. (1991). The pricing of exchange rate risk in the stock market. Journal of Financial and Quantitative Analysis, 26(3), 363e376. Kite, H. (2007). A review of the trade weighted exchange rate index. Retrieved from. http://www.rbnz.govt.nz/research/bulletin/2007_2011/2007jun70_2kite.pdf. Kiymaz, H. (2003). Estimation of foreign exchange rate exposure: an emerging market application. Journal of Multinational Financial Management, 13(1), 71e84. Knetter, M. M. (1989). Price discrimination by U.S. and German exporters. American Economic Review, 79, 198e210. Koutmos, G., & Martin, A. D. (2003). Asymmetric exchange rate exposure: theory and evidence. Journal of International Money and Finance, 22(3), 365e384. Lee, S. (2008). Internationalization of US multinational hotel companies: expansion to Asia versus Europe. International Journal of Hospitality Management, 27, 657e664. Lee, S., & Jang, S. (2010). Internationalization and exposure to foreign currency risk: an examination of lodging firms. International Journal of Hospitality Management, 29(4), 701e710. Marvel, W. M., & Johnson, C. B. (1997). A crisis of currency or creativity? Problems and prospects for the Swiss hotel industry. International Journal of Hospitality Management, 16(3), 279e288. Miller, K. D., & Reuer, J. J. (1998). Asymmetric corporate exposures to foreign exchange rate changes. Strategic Management Journal, 19(12), 1183e1191. Muller, A., & Verschoor, W. F. C. (2006). Foreign exchange risk exposure: survey and suggestions. Journal of Multinational Financial Management, 16(4), 385e410. Muller, A., & Verschoor, W. F. C. (2008). The Latin American exchange exposure of US multinationals. Journal of Multinational Financial Management, 18(2), 112e130. Office of Travel & Tourism Studies. (2008). Market profile of overseas visitors. Retrieved from. http://tinet.ita.doc.gov/outreachpages/download_data_table/ 2007_Overseas_Visitor_Profile.pdf. Oxelheim, L., & Wihlborg, C. G. (1995). Measuring macroeconomic exposure: the case of Volvo cars. European Financial Management, 1(3), 241e263. Oxelheim, L., & Wihlborg, C. G. (2003). Recognizing macroeconomic fluctuations in value-based management. Journal of Applied Corporate Finance, 15, 104e110. Park, K., & Jang, S. (2010). Firm growth patterns: examining the associations with firm size and internationalization. International Journal of Hospitality Management, 29(3), 368e377. Petersen, M. A. (2008). Estimating standard errors in finance panel data sets: comparing approaches. The Review of Financial Studies, 22(1), 435e480. Petersen, M. A., & Thiagarajan, S. R. (2000). Risk measurement and hedging: with and without derivatives. Financial Management, 29(4), 5e30. Prasad, A. M., & Rajan, M. (1995). The role of exchange and interest risk in equity valuation: a comparative study of international stock markets. Journal of Economics and Business, 47(5), 457e472. Priestley, R., & Ødegaard, B. A. (2006). Linear and nonlinear exchange rate exposure. Journal of International Money and Finance, 26(6), 1016e1037. Stulz, R. M. (2003). Derivatives, risk management, and financial engineering. Cincinnati: Southwestern Publishing Co. Stulz, R. M., & Williamson, R. (1997). Identifying and quantifying exposures. In R. Jameson (Ed.), Financial risk and the corporate treasury (pp. 33e51). Risk Publications. Wallis, K. F. (1972). Testing for fourth order autocorrelation in quarterly regression equations. Econometrica, 40(4), 617e636. Walsh, E. J. (1994). Operating income, exchange rate changes, and the value of the firm: an empirical analysis. Journal of Accounting, Auditing and Finance, 9(4), 703e724. Ware, R., & Winter, R. (1988). Forward markets, currency options and the hedging of foreign exchange risk. Journal of International Economics, 25(3e4), 291e302. Williamson, R. (2001). Exchange rate exposure and competition: evidence from the automotive industry. Journal of Financial Economics, 59(3), 441e475. Witt, S. F., & Witt, C. A. (1995). Forecasting tourism demand: a review of empirical research. International Journal of Forecasting, 11, 447e475. Wooldridge, J. M. (2002). Econometric analysis of cross section and panel data. MIT Press. Yoon, Y., & Uysal, M. (2005). An examination of the effects of motivation and satisfaction on destination loyalty: a structural model. Tourism Management, 26 (1), 45e56.