the valuation-relevance of the foreign translation adjustment: the effect of barriers to entry

Available online at www.sciencedirect.com

The International Journal of Accounting

The valuation-relevance of the foreign translationadjustment: The effect of barriers to entry☆

Suresh Radhakrishnana,⁎, Albert Tsangb

a School of Management, University of Texas at Dallas, Richardson, TX, United Statesb School of Accountancy, Chinese University of Hong Kong, Shatin, Hong Kong

Received 23 July 2009

46 (2011) 431–458

Abstract

We examine the economic effects of barriers to entry on the association between foreign currency trans-lation adjustments and the stock returns of multinational firms operating in the manufacturing and serviceindustries. Firms that are innovation leaders, that is, firms that are research and development (R&D) inten-sive and firms with high foreign asset intensity (i.e., asset-intensive firms), are our proxies for firms oper-ating in environments with barriers to entry (i.e., environments in which competition is less intense). Wehypothesize and find that foreign currency translation adjustments are positively associated with abnormalstock returns for firms operating in environments with barriers to entry in both manufacturing and serviceindustries. This finding highlights the importance of assessing the valuation-relevance of foreign currencytranslation adjustments by considering the economic contexts of foreign currency movements. Overall, theevidence shows that the accounting rules governing foreign currency translations generally produce resultsconsistent with the economic effects of foreign exchange rate changes.© 2011 University of Illinois. All rights reserved.

Keywords: Foreign currency translation adjustment; Valuation-relevance; Country growth; Wage rigidity

☆ We thank Gerald Lobo, Rashad Abdel-Khalik, Yuan Ding, an anonymous reviewer, and participants at the 2009Illinois International Accounting Symposium, Catania, Italy for their useful comments and suggestions.We also grate-fully acknowledge comments on earlier versions of the paper from Ashiq Ali, Eli Bartov, Qiang Cheng, Bill Cready,Ferdinand Gul, Bin Srinidhi, and participants at the 2007 AAA annual meeting at Chicago, Illinois.⁎ Corresponding author.E-mail address: [email protected] (S. Radhakrishnan).

0020-7063/$ - see front matter © 2011 University of Illinois. All rights reserved.doi:10.1016/j.intacc.2011.09.008

mailto:[email protected]

http://dx.doi.org/10.1016/j.intacc.2011.09.008





432 S. Radhakrishnan, A. Tsang / The International Journal of Accounting 46 (2011) 431–458

1. Introduction

We examine the economic effects of barriers to entry on the association between foreign cur-rency translation adjustments and the stock returns of multinational firms. Louis (2003) con-siders the economic effect of rigidity of wages and shows that foreign currency translationadjustments are negatively associated with multinational manufacturing firms' stock returns;foreign currency translation adjustments produce results opposite to the economic effects of ex-change rate changes. The argument for rigidity of wages is based on the assumption of openmarkets, that is, unhindered competition. The objective of this paper is to examine the economiceffects of barriers to entry (i.e., we relax the assumption of open markets).

Research in economics shows that foreign currency exchange rates are associated with a for-eign country's economic growth relative to U.S. economic growth, and vice versa (see Balassa,1964; Bhagwati, 1984; Kravis & Lipsey, 1983; Samuelson, 1964).1 The economic growth of aforeign country attracts increased competition because imports are cheaper; this exerts down-ward pressure on output prices. However, input factor prices (especially labor prices) are sticky,resulting in a decrease in future profits (see White, Sondhi, & Fried, 1998, p. 861). Two impor-tant assumptions underpin this argument: (1) the foreign country's product market is open and(2) wages (labor prices) are sticky. We relax the open-market assumption by considering firmsoperating in environments with barriers to entry. Barriers to entry help mitigate the downwardpressure on output prices and allow firms to pass through some of their increased costs to cus-tomers (Chen, Rogoff, & Rossi, 2008; Clark, Masaaki, & Rajaratnam, 1999). In addition, firmsoperating in environments with barriers to entry can benefit from increased demand due to eco-nomic growth in the foreign country, as reflected in foreign currency appreciation (seeBaginski,Lorek, Willinger, & Branson, 1999; Mark, 1995). Thus, a foreign country's economic growthwill result in a higher profit potential for firms operating in environments with barriers toentry. This leads to our prediction that foreign translation adjustments (a proxy for a foreigncountry's economic growth) are positively associated with stock returns (a proxy for higher fu-ture profit potential).

We examine the effect of barriers to entry in the manufacturing and service sectors sepa-rately. We do so for two reasons. First, prior research considers the rigidity of wage effectfor the manufacturing sector alone; accordingly, we extend this prior work by examiningwhether both the rigidity of wages and barriers to entry effects exist for the manufacturingindustry.2 Second, while the rigidity of wage effect is likely to exist for the manufacturingsector, the economic growth and pass-through arguments suggest that the barriers to entry ef-fect should also exist for the service industry (discussed in the next section). Thus, we extendprior research by examining the economic effects of barriers to entry for the service industry.3

1 The Balassa–Samuelson models do not assume purchasing power parity (PPP). Empirical evidence supportsthe view that productivity differentials are an important determinant of long-term exchange rates (see Cheung,Chinn, & Pascual, 2005; Chinn, 1999; Lee & Chinn, 2006; Mark, 1995).2 A firm operating in a foreign country that experiences economic growth will be subject to both effects: the ri-

gidity of wages and the barrier to entry. The barriers-to-entry effect will help to increase revenues, and the rigidityof wage effect will decrease profitability. Thus, both effects can coexist, partly because the pass-through rate is not100% (see Clark et al., 1999; Krugman, 1987).3 In our research design, we include the rigidity-of-wage effect for the service industries as an exploratory analysis.

433S. Radhakrishnan, A. Tsang / The International Journal of Accounting 46 (2011) 431–458

We consider firms with high R&D intensity and asset intensity to be firms operating in en-vironments with barriers to entry. In particular, firms in the top (bottom) 50% of the industry-adjusted R&D capital intensity are considered to be R&D leaders (followers) (see Ciftci, Lev,& Radhakrishnan, 2011). Similarly, firms in the top 50% of industry-adjusted foreign assets(number of foreign employees) to foreign sales ratio each year are considered to be highasset- (labor-) intensive firms. Firms with high industry-adjusted foreign asset (labor) intensityare our proxy for firms operating in environments with barriers-to-entry (rigidity-of-wages)effect.4 We hypothesize that foreign currency translation adjustments are positively associatedwith stock returns for R&D leaders and high asset-intensive firms. Consistent with our hypoth-esis, we find that foreign currency translation adjustments are positively associated with stockreturns for R&D leaders and asset-intensive multinational firms operating in both themanufacturing and service industries.5

An increase in the foreign currency exchange rate (positive foreign currency translation ad-justment) by itself may not be indicative of the foreign country's economic growth because ex-change rates are based on a relative concept (in our case, relative to the U.S. dollar). Forinstance, an appreciation in the foreign currency exchange rate could be associated with eco-nomic growth in the foreign country or an economic downturn in the United States. According-ly, we condition our analysis on the years when theUnited States experienced economic growth,so as to identify whether the foreign currency translation adjustment is indicative of growth inthe foreign country. In particular, we consider whether the National Bureau of Economic Re-search identified a particular year as recessionary or expansionary. We find a positive associa-tion between stock returns and foreign currency translation adjustments for R&D leaders andasset-intensive firms in expansionary periods, and no association in recessionary periods. Thisis likely to be because recessionary periods include very few observations and, consequently,lack statistical power.

We also condition our analysis for both domestic and foreign growth because exchange ratesdepend on relative growth rates and our hypothesis is based on the link between exchange ratesand real economic growth. For this purpose, we use theMajor Currencies Index, which is basedon the level of trade, provided by the Federal Reserve. Conditioned on U.S. economic growth(decline), a negative change in the index indicates growth (decline) in the foreign economy.We call this sub-sample the domestic growth (decline) and foreign growth (decline) sub-sample.However, a positive change in the annual index when the U.S. economy is growing does notindicate whether the foreign economy is growing or declining. We call this sub-sample the do-mestic growth and foreign growth indeterminate sub-sample. We find that the positive associa-tion between foreign currency translation adjustments and stock price occurs in both thedomestic growth and foreign growth sub-sample and the domestic growth and foreign growthindeterminate sub-sample. Taken together, our evidence suggests that foreign currency transla-tion adjustments provide information on future growth prospects for R&D leaders and asset-in-tensive multinational firms operating in environments with barriers to entry. By implication, the

4 In the interests of brevity, we refer to firms with high industry-adjusted foreign asset (labor) intensity as asset-intensive (labor-intensive) firms. We use the terms “innovation-intensive” and “R&D-intensive” interchangeablyfor firms with high industry-adjusted R&D capital intensity.5 As an additional test, we examine the relationship between foreign currency translation adjustments and future

foreign income and find that foreign currency translation adjustments are positively associated with future foreignincome for R&D leaders and asset-intensive firms.


economic effects of foreign exchange rate changes have differing impacts on firms dependingon their respective strategies and characteristics (i.e., their innovation, asset, and labor intensity).Our evidence shows the importance of considering the economic effects of foreign currencymovements in assessing the valuation-relevance of foreign currency translation accountingrules.

Studies on the valuation-relevance of foreign currency translation adjustments have pro-vided mixed results.6 Griffin and Castanias (1987) show that analyst earnings forecast accu-racy improved after SFAS 52, suggesting that the standard enhanced earnings quality. Collinsand Salatka (1993) find that the perceived noise in earnings increases when foreign currencytranslation adjustments are included in net earnings, also suggesting that the earnings qualityimproved under SFAS 52. Gilbert (1989) finds that since the adoption of SFAS 52, foreigncurrency translation adjustments are not valuation-relevant. Similarly, Soo and Soo (1994)express a concern that the relevance of foreign exchange translations for stock returns maybe hard to document due to their small size relative to net income. Bartov (1997) examinesthe association between stock price changes and alternative foreign currency translation ad-justments (under the temporal and current rate methods) and finds that foreign currency trans-lation adjustments are valuation-relevant. Louis (2003) examines the valuation-relevance offoreign currency translation adjustments for manufacturing firms and finds that they are neg-atively associated with stock returns; this finding is consistent with the economic effects ofrigidity of wages. Sabac, Scott, andWeir (2005) provide an analytical insight that the negativeassociation between foreign currency translation adjustments and stock returns depends onwhether the firm's foreign subsidiary is a net producer or a net seller. They consider a sampleof Canadian firms and find a negative (positive) association between foreign translation ad-justments and stock returns for foreign net producers (sellers).

We extend these studies by developing the economic context in which foreign currencytranslation adjustments are positively associated with stock returns. We find that foreigncurrency translation adjustments convey differential information according to the econom-ic contexts and firm strategy as represented by its innovation, asset, and labor intensity.The implication of our findings is that in assessing the valuation-relevance of foreign cur-rency translation adjustments, the economic context is important. Specifically, we providean economic rationale for and evidence of why a positive foreign currency translation ad-justment can be deemed to be an increase in income.

The rest of the paper is organized as follows: Section 2 develops the hypothesis; Section 3provides the empirical analysis; and Section 3.4 provides concluding remarks.

6 Under SFAS 52, translation gains and losses of foreign subsidiaries whose functional currency is the local cur-rency are reported as a cumulative translation adjustment in stockholders' equity; the change in the cumulativetranslation adjustment is part of comprehensive income. Prior to SFAS 52, SFAS 8 required that total foreign ex-change gains and losses were reported as part of net income. A number of studies have examined the valuation-relevance of comprehensive income. For instance, O'Hanlon and Pope (1999) show that “dirty surplus account-ing,” that is, items that bypass reported earnings, are not valuation-relevant. Their findings provide insights intowhether such dirty surplus accounting leads to undesirable earnings management. Similarly, Dhaliwal, Subramanyam,and Trezevant (1999) show that comprehensive income is no more strongly associated with stock returns than is netincome, thus questioning the validity of the notion of comprehensive income.


2. Motivation and hypothesis development

White et al. (1998) discuss settings in which foreign currency translation adjustments dis-tort economic reality. First, they consider the case of a manufacturing subsidiary, the output ofwhich is sold entirely in the foreign country and the costs of which are incurred in the foreigncurrency. If the value of the foreign currency increases, imports are likely to enter that country,resulting in lower sales, lower earnings, and lower cash flows because output prices will haveto be reduced to maintain market share. However, where the foreign currency is used as thefunctional currency, the foreign currency translation adjustment will be positive, even thoughthe economic value of the foreign subsidiary has declined. Louis (2003) makes this argumentprecise by considering the economic effect of rigidity of wages: to remain competitive, thefirm will reduce its selling prices, although its labor costs will remain rigid for an extended pe-riod. Thus, the economic effect of rigidity of wages when a foreign currency appreciates is todecrease the firm's future profitability. Bringing these arguments together, a foreign currencyappreciation leads to a positive foreign currency translation adjustment (FTA) and an increasein reported accounting/comprehensive income; however, it also causes the firm'smarket valueto fall due to the negative impact on the firm's future profitability. Accordingly, Louis (2003)shows that stock returns are negatively associated with foreign translation adjustments, that is,on average, a positive FTA is associated with a loss of value rather than an increase in value,and vice versa. This negative association is more pronounced for labor-intensive firms.

Second, White et al. (1998) consider a case in which a U.S. company has a foreignmanufacturing subsidiary that exports all its goods to the United States for sale. If the foreigncurrency appreciates and the U.S. dollar prices remain unchanged, the subsidiary's profit mar-gins will be squeezed because its revenues will decline as a result of increased productioncosts. If the foreign subsidiary maintains its revenues, U.S. dollar prices will have to increase,causing the quantity of goods sold to fall. Overall, the sales and/or earnings of the companywill decrease, leading to a fall in the value of the company. Sabac et al. (2005) make this in-tuition precise and show that foreign currency translation adjustments are negatively associ-ated with economic value when the foreign subsidiary is a net producer.7

Central to the arguments and findings of Louis (2003) and Sabac et al. (2005) is the assump-tion that product markets are open. For instance, Louis (2003) argues that the entry of potentiallycheaper imported goods exerts downward pressure on the output prices of the subsidiary. Sim-ilarly, Sabac et al. (2005) assume that foreign currency exchange rate shocks cannot be passedthrough to the product market.8 Thus, we consider the effects of relaxing the open-market as-sumption and refer to this as “barriers to entry.”

7 Sabac et al. (2005) also consider the differential stickiness of input and output prices and argue that the stickiness ofwages is not likely to be a reason for the negative association between stock prices and foreign currency translation ad-justments. This is consistent with the empirical evidence in the economic literature, which suggests that when exchangerates change, wages are onlymarginally more rigid than output prices (see Froot &Rogoff, 1995; Giovanni, 1988; Haskel& Wolf, 2001).8 The foreign operation can be either a self-sustained foreign operation (SSFO) or an integrated foreign opera-

tion (IFO). Foreign operations are typically a combination of these two types. Therefore, we make no distinctionbetween SSFOs and IFOs either in developing our hypothesis or in the empirical analyses. Sabac et al.'s (2005)categorization of firms as net producers and net sellers represents an attempt to distinguish between the two onan empirical basis. In our sample of U.S. firms, we find that about 80% of firms' foreign operations are net sellers.


2.1. Barriers to entry, foreign currency exchange rates, and stock returns

White et al. (1998) consider a case in which a foreign subsidiary operates in an environmentwith no export or import competition. In this case, the parent company data represent the perfor-mance of the subsidiary in the foreign currency. In other words, if the foreign currency appreci-ates, the firm's net investment in the subsidiary increases in foreign currency terms and viceversa. This suggests that net investment is a good proxy for the value of the investment to theparent. In comparison with the scenarios considered by Louis (2003) and Sabac et al. (2005)(see the above discussion), the main assumption that determines the positive association of for-eign currency translation adjustments with stock price is the restriction on competition. Moreimportantly, when competition is restricted, the FTA provides aggregate information onchanges in the foreign currency exchange rate and thus conveys information on the foreigncountry's economic growth.

Stigler (1963), Mueller (1977), and Kamerschen (1968) suggest that more barriers to entryrestrict competition by limiting entry to the industry. Firms can create barriers to entry bychoosing product differentiation strategies that ensure that their products do not have directsubstitutes. Caves and Porter (1977) develop a framework for intra-industry profit differen-tials based on pre-commitment to specialized resources such as R&D and capital-intensivetechnologies. Eaton and Lipsey (1981) show a positive association between industry profit-ability and capital intensity, which suggests that capital intensity is a potential source of bar-riers to entry. For foreign operations, especially those in emerging markets where foreigndirect investments are key drivers of economic growth, capital intensity is likely to be an im-portant barrier to entry. Evidence shows that foreign direct investments are an important de-terminant of economic growth (Borensztein, DeGregorio, & Lee, 1998; Froot & Stein, 1989;Griffin & Stulz, 2001). Firms invest in capital-intensive technologies in foreign countries toimprove productivity and to take advantage of local economic growth. Thus, firms thatmake strategic choices to differentiate their products through R&D activities and/or investheavily in foreign country assets are likely to create barriers to entry.9

Firms that follow a strategy of creating barriers to entry are positioned to take advantage of aforeign country's economic growth in twoways. First, the downward pressure on output prices islikely to be muted for such firms because of restricted competition. This allows these firms to in-crease their revenues based on the increased demand resulting from economic growth. Second,input prices and labor costs are either sticky (see Louis, 2003) or increase on average due to eco-nomic growth (see, for instance, Bailliu & King, 2005). Clark et al. (1999) argue that pricing ininternational markets depends on firms' ability to pass through the increased costs, which in turndepends on their product differentiation strategy. Krugman (1987) finds that foreign exporters tothe United States during the early to mid-1980s passed through only 60 to 65% of the real appre-ciation of the U.S. dollar to their U.S. customers. Chen et al. (2008) have recently shown that for-eign currency exchange rates are positively associated with future commodity prices, indicating arelationship between economic growth and pass-through rates. The effects of barriers to entryapply equally to foreign production or marketing subsidiaries: if the foreign subsidiary is a

9 Human capital measured in terms of management quality, innovation ability, etc. is also likely to create barriers to en-try. However, superior human capital resources may not be reflected in the total number of employees. Consequently, la-bor intensity is more likely to be subject to the economic effect of rigidity of wages than is total number of employees.


production subsidiary, then the parent will pass through its increased costs in the United States.Thus, firms with barriers to entry can insulate themselves from the adverse effects of high/increased input prices by passing through their increased costs to their customers.10

The role of the service sector in economic growth is well known (for example, see Jorgenson,Ho,& Stiroh, 2005). Recently, the growth in the service sector is attributable to firms in businessservices, including information technology firms and technical and engineering consultingfirms.11 Similar to manufacturing firms, these service firms are also subject to increasedforeign competitive pressures when the foreign economy grows. Furthermore, some servicefirms employ differentiation-related strategies (i.e., innovation strategies), and some, such asIBM, have large brick and mortar assets as well (see Hall, 2000; Lev & Radhakrishnan,2005). Accordingly, firms in the service industry are likely to enjoy the barriers-to-entry effectto the extent that their services are differentiated (see discussion on differentiation above). Thus,service firms with barriers to entry can insulate themselves from the adverse effects of high inputprices (typically high personnel costs) by passing the increased costs on to their customersand can reap the benefit from economic growth through increased demand for their services(Nakamura, 2003). On the other hand, service firms who are not subject to the barriers-to-entry effect may not be able to pass through the increased personnel related costs or commandrents due to increased competition. However, if the nature of service is such that increased laborcosts can be passed through to consumers, then the rigidity-of-wage effect may not apply for theservice sector. Thus, we do not hypothesize the rigidity-of-wage effect for the service sector.

We perform the analysis separately for firms in the manufacturing and service industries.12

2.2. Hypothesis

H1. Foreign currency translation adjustments are positively associated with stock returnsfor firms with barriers to entry in the manufacturing and service industries.

Foreign currency exchange rate is a relative concept. An appreciation in the foreign currencyexchange rate could be associated with economic growth in the foreign country or an economic

10 Consistent with this notion, Caves andGhemawat (1992) show that differentiation-related strategies play an importantrole in sustaining profit differentials across firms. Differentiation-related strategies include innovation activities such as theintroduction of new products/services and brands. In essence, barriers to entry restrict competition and therefore allow foroutput prices to be raised to absorb increased costs without any adverse effect on earnings or profits. Baginski et al. (1999)show that firms operating in environments with barriers to entry have more persistent earnings because they can passthrough increases in input prices while protecting their market share.11 Manufacturing (service) firms are firms operating in SIC codes 2000–3999 (7200–8799). Firms in SIC 73, 80, and 87are typically the ones with large R&D expenditures. In SIC 73, the large R&D outlays are for firms (1) firms engaged insoftware design and production, such as Adobe Systems Inc. and American Software Inc.; (2) firms providing IT supportand data services, such as IBM and American Management Systems; (3) firms providing business services, such as ItexCorp., Pacificnet Inc., and Telecredit Inc.; and (4)firms providing advertising services, such asAdvanced Promotion TechInc. and Mediaplex Inc. In SIC 80, firms engaged in health-related services have large R&D outlays. Examples include:Clinical Homecare Ltd.,Medgroup Inc., Neuromedical System Inc., andOncormed Inc. In SIC 87, the large R&Doutlaysare for (1) firms engaged in physical or biological research, such as Arena Pharmaceuticals Inc., Commonwealth Biotech-nologies, Integrated Genetics, and Superconductor Technologies; (2) firms providing management consulting services,such as GIGA Information Group, Accenture Ltd., and Venture Catalyst Inc.; and (3) testing laboratories, such as ASETest Ltd., Cambridge Association.12 If we combine the manufacturing and service sectors and find that the rigidity-of-wage effect does not exist, we willnot be able to isolate the reason for such a result. That is, we will not be able to attribute the finding to the non-robustnessof the Louis (2003) result to a different time-period or to the difference in the sample of industries.


downturn in the domestic market. If a foreign currency appreciation is concurrent with an eco-nomic downturn in the domestic market, it could be bad news for firms in that market. In otherwords, in the arguments leading to our hypothesis, we implicitly assume that the U.S. economyis not declining when the foreign currency is appreciating. We test this assumption by consid-ering economic growth in the United States relative to economic growth in the foreign economyin our empirical tests.

The hypothesis can be articulated on an intuitive basis. In a process that started in the late1980s, many countries transformed their controlled economies into market-based economies.These transitional economies had/have the potential for considerable future economic growth.Lured by this growth potential, many U.S. companies invested in such countries. When a for-eign country's exchange rate rises as a result of economic growth, the future cash flows expectedfrom such investments are more likely to be realized. For instance, consider a pharmaceuticalcompany that follows an innovation strategy. Innovation provides the company with an edgenot only in the U.S. market but also in foreign markets due to the economic growth they expe-rience; that is, the company is likely to be able to realize higher profits in its foreign markets dueto increased aggregate demand. Innovation and patent protection enable such companies to re-alize such higher profits unencumbered, at least for a period of time. The key aspect here is thatmultinational firms are able to reap the benefits of increased aggregate demand through barriersto entry, such as innovation.

3. Empirical analysis

3.1. Proxy and variable definitions

We use two proxies for barriers to entry: innovation strategy and foreign asset intensity.

3.1.1. Proxy for innovation strategyWe use industry-adjusted capitalized R&D intensity to classify firms as R&D leaders or fol-

lowers (see Ciftci et al., 2011). Capitalized R&D is computed by capitalizing and amortizingfirm R&D expenditures (Compustat data item 46) over 5 years (see Lev & Sougiannis,1996). Firm R&D capital intensity is defined as capitalized R&D divided by sales (Compustatdata item 12).13 Industry-adjusted R&D capital intensity (R&D_ADJ) is the difference betweena firm's R&D capital intensity and the weighted average R&D capital intensity of the industry.A firm is classified as an R&D leader in year t if its industry-adjusted R&D capital intensity inyear t−1 is above the median. Correspondingly, all other firms are classified as R&D followers.

We use total R&D capital intensity as opposed to foreign R&D capital intensity for two rea-sons. First, the knowledge/innovations gained from R&D activities are transferable. Thus, mostR&D efforts, regardless of where they are made, can be transferred to other geographic marketsto reap the benefits of innovation. Second, as a practical matter, less than 30 firms in our samplereport foreign R&D expenditures in the Compustat segment database. Thus, most firms do notvoluntarily provide such information in their annual reports, even if they conduct R&D activitiesin foreign countries. We thus use total R&D capital intensity as the measure of a firm's innova-tion strategy.

13 Accounting rules (SFAS 2) mandate that firms provide information on R&D expenditures. We consider R&Dexpenditures to be 0 for firms with missing R&D expenditure data, because these are mandated disclosures.


3.1.2. Proxy for foreign capital intensityWe use a firm's industry-adjusted foreign asset intensity (ForAsset_ADJ) as the proxy for its

foreign capital intensity. ForAsset_ADJ is calculated as the difference between a firm's foreignasset intensity and the mean foreign asset intensity of the two-digit SIC code. Data for foreignsales and assets are obtained from the Compustat segment database. Firms with industry-adjust-ed foreign assets to foreign sales ratio greater than (lower than or equal to) the median ratio inyear t−1 are classified as asset-intensive (labor-intensive) firms in year t.14 Asset-intensivefirms are considered to be operating in an environment with barriers to entry.

3.2. Research design

To empirically examine the hypothesis, we augment the Louis (2003) model and esti-mate the following equation:

SARETit ¼ β0 þ β1RDLeader þ β2HighAsset þ β3HighLaborþβ4FTAit þ β5FTAit∗RDLeader þ β6FTAit∗HighAsset þ β7FTAit∗HighLaborþβ8NIADJit þ β9ΔNIADJit þ β10FTAXit þ β11TADJit þ error

ð1Þ

where SARET is the annual size-adjusted abnormal return for the period ending 3 months afterthe close of the fiscal year.15 The annual return is adjusted for delisting bias according to themethod used by Shumway (1997). Size-adjusted returns are computed using the companionsize portfolio approach. Market capitalization is the proxy for size, and the size-decile break-points used to identify the companion portfolio are based on NYSE/AMEX stocks. Thus, thesize-adjusted abnormal return is the annual raw return minus the annual companion size-decile return. Data used for computing the size-adjusted returns are obtained from the Centerfor Research in Security Prices (CRSP).

RDLeader and HighAsset are our proxies for barriers to entry. In particular, RDLeader is adummy variable that equals 1 if the industry-adjusted R&D intensity of a firm is greater thanindustry R&D intensity, and 0 otherwise. HighAsset (HighLabor) is a dummy variable thattakes the value of 1 if a firm is classified as an asset-intensive (labor-intensive) firm. Asset-in-tensive (labor-intensive) firms are classified using industry-adjusted foreign asset (foreign em-ployee) intensity.16 FTA is the change in cumulative foreign translation adjustment (changein Compustat data item 230).

14 We also use total assets to proxy for firms' asset intensity, that is, gross total assets (Compustat data item 6 plus Com-pustat data item 196) scaled by net sales (Compustat data item 12), and we obtain qualitatively similar results. In addition,we conduct a robustness test using a benchmark other than the median to classify high and low asset-intensive firms,whereby a firm is classified as an asset-intensive firm when its industry-adjusted foreign asset intensity is above 0; theresults are qualitatively similar. The Pearson (Spearman) correlation between the foreign asset intensity measure andthe total asset intensity measure is 0.79 (0.89). These high correlations between the foreign and total asset intensity mea-sures suggest that firms use similar business models for their foreign and domestic operations.15 We use raw returns instead of size-adjusted returns and obtain qualitatively similar results. We also estimateEq. (1) separately for each group without the interaction variables and obtain qualitatively similar results.16 Unlike foreign sales and foreign assets, which are mandatory disclosures under FASB Statement number 131“Disclosures about Segments of an Enterprise and Related Information,” the number of foreign employees is avoluntary disclosure. If foreign employee data are missing, we use the firm-level number of employees multipliedby the foreign assets over total assets ratio. The Pearson (Spearman) correlation between foreign labor intensityand total labor intensity is 0.78 (0.90), based on observations for which foreign employee data are available.


TADJ is foreign transaction gains or losses (Compustat data item 150), which we use asa separate control variable. TADJ controls for the portion of foreign earnings that are real-ized and are included in net income. Correspondingly, we use adjusted net income NIADJ,which is computed as net income (Compustat data item 18) minus foreign transaction gainsor losses (Compustat data item 150).17 ΔNIADJ is the change of adjusted net income fromyear t−1 to year t. We include this variable as a control variable because prior researchsuggests that both levels and changes in earnings are important to consider. In essence,we disaggregate earnings into the foreign component of earnings and other earnings.FTAX is foreign income tax (Compustat data item 64). All the explanatory variables, in-cluding the intercept, are scaled by the initial market value.

We estimate Eq. (1) for manufacturing (SIC 2000–3999) and service firms (SIC 7200–8799) separately. The research specification in Eq. (1) allows for the economic effects ofboth rigidity of wages and barriers to entry to co-exist. For instance, firms with RDLeaderand HighAsset of 0 are likely to be subject to the rigidity-of-wage effect. Similarly, the totaleffect of both barriers to entry and rigidity of wages for firms with RDLeader, HighAsset,and HighLabor all equaling 1 is the sum of β5, β6, and β7. In short, the hypothesis does notimply that the barriers-to-entry effect subsumes the rigidity-of-wage effect (see Krugman,1987). We thus expect the coefficient on the interaction term FTA∗HighLabor (β7) to benegative for manufacturing firms. Based on the barriers-to-entry hypothesis, we expectthe coefficient on the interaction terms FTA∗RDLeader (β5) and FTA∗HighAsset (β6) tobe positive for both manufacturing and service firms. In addition, we expect the effect ofbarriers to entry on the firm's abnormal stock returns to be positive as well, that is, thesum of the coefficients β5, β6, and β4 is expected to be positive.

3.3. Sample

The sample selection procedure is similar to that adopted by Louis (2003). The sample in-cludes all multinational firms that report at least two consecutive non-zero cumulative foreignexchange adjustments (Compustat data item 230). These multinational firms are required tohave the same fiscal year-end for all years and to have the relevant data available from boththe CRSP and Compustat databases to estimate Eq. (1). Furthermore, extreme observationsin the top and bottom 1% of the annual empirical distribution for all explanatory variablesin Eq. (1) are deleted. The final sample contains 8089 (2506) firm-year observations, repre-senting 1607 (703) multinational firms in the manufacturing (service) industry for the period1985 to 2006.

Panel A of Table 1 provides the sample distribution by two-digit SIC codes. Formanufacturing firms, roughly 70% of the sample operates in four industries: SIC 35, Indus-trial Machinery and Equipment (18.54%); SIC 38, Instruments and Related Products(18.35%); SIC 36, Electronic and Other Electric Equipment (18.10%); and SIC 28, Chem-ical and Allied Products (14.49%). For service firms, more than 78% of the sample oper-ates in one industry: SIC 73, Business Services (78.41%). The industries represented in

17 In unreported analysis, we include only the level of adjusted net income as in Louis (2003) and find qualita-tively similar results.


the sample are generally characterized by high levels of innovation. Panel B of Table 1 pro-vides the sample distribution by year. Both the manufacturing and service firms exhibita similar temporal pattern: the number of multinational manufacturing (service) firmsis 271 (49) in 1985, 623 (207) in 1997, and 294 (131) in 2006, suggesting that moreU.S. firms have established a global presence over time, while the decrease in the num-ber of multinational firms from 1998 may be due to the trend of outsourcing. Overall, thetrend emphasizes the increasing importance of foreign operations for U.S. multinationalfirms.

Table 2, Panel A (B) provides the descriptive statistics for the variables used in estimatingEq. (1) for manufacturing (service) firms, the unscaled explanatory variables used in Eq. (1),the variables used to measure firm R&D capital, firm foreign assets, and firm labor intensity(both on an industry-adjusted basis and without adjusting for the industry average), and cer-tain variables used to describe firm characteristics. The mean (median) size-adjusted return is0.025 (−0.002) and 0.018 (−0.007) for manufacturing and service firms, respectively, all ofwhich are statistically insignificant values showing that the size-adjusted return is 0 on aver-age. The mean unscaled FTA for manufacturing (service) firms is $0.08 ($−0.01) million witha standard deviation of 16.73 (9.52). This shows that while the FTA is 0 on average, there isconsiderable variation between firms. The median market value of equity for manufacturing(service) firms is $189.73 ($204.80) million with Q1 of $50.62 ($53.54) million and Q3 of$740.05 ($735.25) million, showing that firms of varying sizes have foreign operations.The median market-to-book ratio of manufacturing (service) firms is 2.06 (2.84), with Q1of 1.31 (1.59) and Q3 of 3.41 (4.94), indicating that the sample contains firms with consider-able variation in intangible intensity. The median sales growth for manufacturing (service)firms in years t, t+1, and t+2 are 9%, 21%, and 34% (13%, 31%, and 53%), respectively,suggesting that the typical multinational firm in our sample has positive future sales growth.The market-to-book ratio and future sales growth profiles provide support for the importanceof examining the economic effects of barriers to entry. In computing future sales growth pro-files, we include only observations of firms that survive in future years. Thus, the sales growthmeasures are likely to be biased upward.

Our measure of firm innovation strategy is industry-adjusted R&D capital intensity. Themean industry-adjusted R&D capital intensity for manufacturing (service) firms is −0.28(−0.68), showing that only a few firms use innovation as a strategy when benchmarkedagainst the industry. The firm asset intensity measure, the industry-adjusted foreign assetsto foreign sales ratio, ranges from −1.15 (−4.61) in Q1 to 0.16 (1.14) in Q3 formanufacturing (service) firms; the firm labor intensity measure, the number of foreign em-ployees to the foreign sales ratio adjusted by industry, ranges from −6.97 (−25.95) in Q1to 0.86 (0.63) in Q3 for manufacturing (service) firms. Overall, the sample characteristicspresented in Panels A and B demonstrate that the economic context of barriers to entry iswell represented in the sample.

Table 2, Panel C shows both Pearson and Spearman correlations. The Spearman corre-lation of FTA and size-adjusted return is 0.02 and is statistically significant at the 5% level(the Pearson correlation is 0.01 and is statistically significant at the 12% level). This pro-vides initial evidence that FTA is value-relevant, although the effect is weak. FTA isalso positively associated with barriers-to-entry and the rigidity-of-wages proxies. Thisshows the importance of controlling for the economic effects, because the two effects

Table 1Sample distribution.

Panel A: distribution of sample by two-digit SIC codes

SIC code Industry grouping No. of obs. % of obs. No. of firms % of firms

Manufacturing firms20 Food and kindred products 254 3.14 52 3.2421 Tobacco products 6 0.07 4 0.2522 Textile mill products 92 1.14 15 0.9323 Apparel and other textile products 135 1.67 29 1.8024 Lumber and wood products 33 0.41 6 0.3725 Furniture and fixtures 111 1.37 23 1.4326 Paper and allied products 149 1.84 27 1.6827 Printing and publishing 188 2.32 35 2.1828 Chemical and allied products 1172 14.49 231 14.3729 Petroleum and coal products 79 0.98 13 0.8130 Rubber and miscellaneous plastics products 252 3.12 46 2.8631 Leather and leather products 39 0.48 8 0.5032 Stone, clay, and glass products 54 0.67 12 0.7533 Primary metal industries 183 2.26 40 2.4934 Fabricated metal products 294 3.63 47 2.9235 Industrial machinery and equipment 1500 18.54 304 18.9236 Electronic and other electric equipment 1464 18.10 295 18.3637 Transportation equipment 424 5.24 80 4.9838 Instruments and related products 1484 18.35 300 18.6739 Misc. manufacturing industries 176 2.18 40 2.49

Total 8089 100.00 1607 100.00

Service Firms72 Personal services 27 1.08 8 1.1473 Business services 1965 78.41 554 78.8175 Auto repair, services, and parking 42 1.68 9 1.2876 Miscellaneous repair services 2 0.08 1 0.1478 Motion pictures 58 2.31 18 2.5679 Amusement and recreation services 56 2.23 15 2.1380 Health services 94 3.75 24 3.4182 Educational services 30 1.20 10 1.4283 Social services 12 0.48 3 0.4387 Engineering and management services 220 8.78 61 8.68

Total 2506 100.00 703 100.00


with which FTA is correlated may cancel each other out. Thus, it is important to examinethe economic effects in the multivariate specification, as in Eq. (1).

3.4. Results

Table 3 provides the estimate of Eq. (1) for manufacturing and service firms. Eq. (1) isestimated using OLS, and because we have panel data, it is well known that the standarderrors are biased downward. We correct for the bias in standard errors using the Huberman,White, and Sandwich procedure with firm as the clusters (see Petersen, 2009 on the

Table 1 (continued)

Panel B: distribution of sample by year

Year Manufacturing firms Service firms

No. of obs. % of obs. No. of obs. % of obs.

1985 271 3.35 49 1.961986 297 3.67 57 2.271987 333 4.12 65 2.591988 318 3.93 72 2.871989 332 4.10 69 2.751990 351 4.34 80 3.191991 357 4.41 83 3.311992 397 4.91 90 3.591993 432 5.34 102 4.071994 496 6.13 126 5.031995 516 6.38 134 5.351996 574 7.10 163 6.501997 623 7.70 207 8.261998 404 4.99 128 5.111999 305 3.77 104 4.152000 288 3.56 119 4.752001 297 3.67 149 5.952002 300 3.71 141 5.632003 304 3.76 138 5.512004 306 3.78 156 6.232005 294 3.63 143 5.712006 294 3.63 131 5.23Total 8089 100.00 2506 100.00

Note: manufacturing (service) firms are firms bearing SIC codes 2000–3999 (7200–8799). The sample is from1985 to 2006.


appropriateness of this procedure).18 For all the multivariate analysis reported here, the t-statistics are corrected using robust standard errors.

Consistent with the rigidity-of-wages argument, the coefficient on the interaction termFTA∗HighLabor (β7) is negative for manufacturing firms (β7=−2.41, t=−2.13), showingthat foreign translation adjustments are negatively associated with abnormal stock returnsfor labor-intensive firms. Consistent with our hypothesis, the coefficients on the interactionterms are 2.26 (t=3.68) for FTA∗RDLeader (β5) and 1.78 (t=2.29) for FTA∗HighAsset(β6) among manufacturing firms, indicating support for the barriers-to-entry effect.19

The results for service firms are qualitatively similar to those for manufacturing firms. Theestimated coefficient on FTA∗HighLabor (β7) is negative (β7=−1.17, t=−1.50), while the es-timated coefficients on FTA∗RDLeader (β5) and FTA∗HighAsset (β6) are significantly positive(β5=1.05, t=2.68; β6=1.27, t=3.16). In addition, the sum of the coefficients β5 (β6) and β4 is

18 Using years or industry-years as the cluster yields qualitatively similar results.19 We estimate the Louis (2003) model without the interaction terms in Eq. (1) for the same sample period con-sidered in his study, that is, 1985 to 2001. The estimated coefficient for FTA is −0.51 (t=−2.84) for manufactur-ing firms, which is consistent with Louis's (2003) result. For our sample period, the coefficient on FTA is 0.60(t=4.08). This result could be attributable to the countervailing economic effects we examine in Eq. (1).

Table 2Descriptive statistics and correlation.

Panel A: manufacturing firms Panel B: service firms

Mean Std. dev. Q1 Median Q3 Mean Std. dev. Q1 Median Q3

Eq. (1) variablesSARETURN 0.025 0.583 −0.100 −0.002 0.070 0.018 0.668 −0.038 −0.007 0.026NIADJ 0.004 0.181 −0.007 0.050 0.081 −0.036 0.199 −0.061 0.025 0.059ΔNIADJ 0.013 0.200 −0.023 0.008 0.033 0.026 0.239 −0.025 0.008 0.037FTA 0.001 0.031 −0.004 0.000 0.004 −0.001 0.020 −0.002 0.000 0.003FTAX 0.008 0.012 0.000 0.002 0.010 0.006 0.010 0.000 0.001 0.007TADJ −0.001 0.004 −0.002 0.000 0.001 0.000 0.003 −0.001 0.000 0.001

Eq. (1) explanatory variables unscaledNIADJ 56.96 208.88 0.15 7.41 63.20 25.68 116.40 −4.00 4.10 26.00FTA 0.08 16.73 −0.60 −0.01 0.54 −0.01 9.52 −0.30 −0.01 0.31FTAX 6.33 30.10 0.01 0.55 8.00 5.81 20.26 0.00 0.36 8.40TADJ −0.55 4.48 −0.50 −0.02 0.23 −0.38 2.34 −0.50 −0.03 0.19

R&D capital, foreign asset and labor intensityR&D 0.31 1.14 0.01 0.07 0.19 0.25 0.47 0.00 0.13 0.34ForAsset 1.28 1.76 0.69 0.89 1.21 1.39 1.23 0.78 1.04 1.55ForLabor 7.87 5.77 4.42 6.64 9.51 9.79 10.56 4.85 6.85 10.25R&D_ADJ −0.28 0.48 −0.52 −0.13 0.01 −0.68 0.94 −0.89 −0.46 0.11ForAsset_ADJ −0.75 1.11 −1.15 −0.59 0.16 −3.45 3.85 −4.61 −0.35 1.14ForLabor_ADJ −3.63 5.44 −6.97 −3.84 0.86 −9.21 15.90 −25.95 −8.05 0.63

Firm characteristicsMarket value of equity ($ million) 1221.55 3336.30 50.62 189.73 740.05 1168.43 3637.86 53.54 204.80 735.25Market/book 3.00 3.18 1.31 2.06 3.41 4.36 4.94 1.59 2.84 4.94

444S.R

adhakrishnan,A.Tsang

/The

InternationalJournal

ofAccounting

46(2011)

431–458

Raw return 0.19 0.66 −0.20 0.08 0.39 0.17 0.78 −0.31 0.03 0.43Sales ($ million) 914.21 2210.52 59.75 177.10 602.90 571.11 1390.16 45.60 123.00 455.00Sales growth t 0.16 0.39 0.00 0.09 0.22 0.23 0.47 0.01 0.13 0.32Sales growth t+1 0.43 0.94 0.03 0.21 0.49 0.74 1.58 0.05 0.31 0.81Sales growth t+2 0.81 1.89 0.08 0.34 0.78 1.77 4.91 0.10 0.53 1.43

Panel C: Pearson (Spearman) correlations are reported above (below) the diagonal. Significant levels are shown in italics

Variables SARET NIADJ ΔNIADJ FTA FTAX TADJ RDLeader HighAsset HighLabor

SARET 1.00 0.21 0.26 0.02 0.12 −0.01 0.00 −0.02 −0.020.00 0.00 0.02 0.00 0.25 0.74 0.05 0.05

NIADJ 0.46 1.00 0.29 0.01 0.09 0.02 0.04 0.05 0.050.00 0.00 0.56 0.00 0.06 0.00 0.00 0.00

ΔNIADJ 0.39 0.47 0.06 0.06 0.00 −0.01 −0.02 −0.010.00 0.00 0.00 0.00 0.62 0.36 0.01 0.53

FTA 0.01 0.01 0.05 1.00 −0.01 −0.02 0.01 0.00 −0.010.12 0.13 0.00 0.28 0.01 0.29 0.66 0.34

FTAX 0.16 0.27 0.06 0.00 1.00 −0.10 0.03 0.03 0.050.00 0.00 0.00 0.85 0.00 0.00 0.00 0.00

TADJ 0.00 −0.03 0.02 0.02 −0.05 1.00 0.01 0.02 0.000.79 0.00 0.00 0.02 0.00 0.13 0.01 0.68

RDLeader 0.03 0.09 0.00 0.03 −0.02 0.03 1.00 0.48 0.420.00 0.00 0.95 0.00 0.05 0.00 0.00 0.00

HighAsset 0.01 0.11 −0.01 0.07 0.01 0.03 0.48 1.00 0.570.17 0.00 0.40 0.00 0.21 0.00 0.00 0.00

HighLabor 0.01 0.11 0.01 0.02 0.01 0.02 0.42 0.57 1.000.16 0.00 0.43 0.06 0.26 0.07 0.00 0.00

Notes: Manufacturing (service) firms are firms operating in SIC codes 2000–3999 (7200–8799). Panel A (B) shows the descriptive statistics for manufacturing (service)firms. All the descriptive statistics are computed after winsorizing the top and bottom one percentile of all the variables. In Panel C, Pearson (Spearman) correlations arereported above (below) the diagonal. The p-value is shown in italics below the correlation coefficient estimate.Variable definitions are contained in Appendix 1.

445S.R

adhakrishnan,A.Tsang

/The

InternationalJournal

ofAccounting

46(2011)

431–458

Table 3R&D leader, foreign asset intensity, foreign labor intensity and foreign translation adjustment.

Expectedsign

Manufacturing firms Service firms

Coef. -stat Coef. t-stat

FTA (β4) − 0.15 0.98 −0.01 −0.01FTA∗RDLeader (β5) + 2.26 ⁎⁎⁎ 3.68 1.05 ⁎⁎⁎ 2.68FTA∗HighAsset (β6) + 1.78 ⁎⁎ 2.29 1.27 ⁎⁎⁎ 3.16FTA∗HighLabor (β7) − −2.41 ⁎⁎ −2.13 −1.17 −1.50RDLeader (β1) ? −0.03 −0.03 0.01 0.97HighAsset (β2) ? −0.03 ⁎ −1.84 0.05 ⁎⁎⁎ 4.85HighLabor (β3) ? −0.03 ⁎⁎⁎ −2.78 −0.02 −1.82NIADJ (β8) + 0.67 ⁎⁎⁎ 52.15 0.74 ⁎⁎⁎ 34.88ΔNIADJ (β9) + 0.38 ⁎⁎⁎ 32.11 0.28 ⁎⁎⁎ 14.87FTAX (β10) + 6.08 ⁎⁎⁎ 25.98 10.16 ⁎⁎⁎ 14.79TADJ (β11) ? −0.75 −0.97 −0.76 −0.34Fixed year dummies Yes YesFixed industry dummies Yes YesNo. of obs. 8089 2506No. (%) of obs. with positive FTA 3133 (38.73%) 983 (39.23%)Adjusted R-square (%) 27.42 25.69Test of total effect F-stat p-value F-stat p-valueFTA+FTA∗RDLeader =0 F=6.36 Prob.=0.01 F=7.83 Prob.=0.00FTA+FTA∗HighAsset =0 F=1.62 Prob.=0.20 F=13.94 Prob.=0.00

Notes: Eq. (1):

SARETit ¼ β0 þ β1RDLeader þ β2HighAsset þ β3HighLabor

þβ4FTAit þ β5FTAit∗RDLeader þ β6FTAit∗HighAsset þ β7FTAit∗HighLabor

þβ8NIADJit þ β9ΔNIADJit þ β10FTAXit þ β11TADJit þ error:

All independent variables in the above model, including the intercept, are deflated by the firm's beginning marketvalue of equity. The t-statistics are based on standard errors using the Huberman–White–Sandwich procedure (seePetersen, 2009) with firm-level clustering. Manufacturing (service) firms are firms operating in SIC codes 2000–3999 (7200–8799). The manufacturing (service) firms sample contains 8089 (2506) firm-year observations span-ning from 1985 to 2006.Variable definitions are contained in Appendix 1.

⁎ Indicates significance at the 10% level.⁎⁎ Indicates significance at the 5% level.

⁎⁎⁎ Indicates significance at the 1% level.


also positive for both manufacturing and service firms: for manufacturing firms, F-value=6.36(1.62) and for service firms, F-value=7.83 (13.94). These results provide support for the argu-ment that the barriers-to-entry effect exists for firms in both the manufacturing and serviceindustries.

3.4.1. Conditioning on domestic economic growthAs discussed earlier, foreign currency exchange rate is a relative concept. An appreciation

in the foreign currency exchange rate could be associated with economic growth in the for-eign country or an economic downturn in the domestic market. The arguments on whichwe rely in developing our hypothesis assume that the U.S. economy is not declining. To


ensure that the effect of the interaction of barriers to entry and FTA is conditional on foreigneconomic growth rather than on a domestic economic downturn, we estimate Eq. (1) for sub-samples that consist solely of observations for years in which the U.S. economy was growingor declining. We use the recessionary years indicated by the Business Cycle Dating Commit-tee of the National Bureau of Economic Research (NBER). During our sample period, 1990,1991, and 2001 are identified as recessionary years.20 To condition our analysis on U.S. eco-nomic growth, we estimate Eq. (1) separately for the expansionary and recessionary years.

The results are reported in Table 4, Panel A. For manufacturing (service) firms during theexpansionary years, the coefficient estimate on FTA∗RDLeader is 1.72 with t=2.69 (1.44with t=3.41) and the coefficient estimate on FTA∗HighAsset is 0.71 with t=1.91 (1.08 witht=2.42), both of which are statistically significant at the 10% level. This is consistent with thebarriers-to-entry hypothesis for firms in themanufacturing and service industries. For the serviceindustry, the rigidity-of-wage effect is not statistically significant at the 10% level (see thediscussion prior to the hypothesis presentation).

Table 4, Panel B provides the results of estimating Eq. (1) for the recessionary years.Generally, results are not consistent with the barriers-to-entry hypothesis for firms in themanufacturing and service industries. This can be explained by the differential effects offiscal and monetary policy on individual firms in recessionary periods: foreign currencytranslation may only have a second-order effect on firm value.

3.4.2. Conditioning on domestic and relative growthWe now condition the analysis on both domestic growth and relative growth in the for-

eign economy. As discussed later, while an environment in which there is U.S. economicgrowth and a positive FTA is indicative of foreign economic growth, an economic contextof U.S. economic growth and a negative FTA does not necessarily indicate that the foreigneconomy is not growing; it merely indicates that the foreign economy is growing at aslower rate than the U.S. economy. This underlines the importance of conditioning ouranalysis on both domestic and foreign growth. To do so, in addition to identifying reces-sionary years, we obtain a summary measure of the foreign exchange value of the U.S. dol-lar based on the major currencies index from the H.10 Federal Reserve StatisticalRelease.21 In particular, we use the monthly major currencies index obtained from the

20 The Committee of the National Bureau of Economic Research (NBER) uses five indicators including realgross domestic product, real income, employment, industrial production, and wholesale-retail sales as factors toidentify periods of economic recession. A description of the procedure used is available at http://www.nber.org/cycles/recessions.html. The recessionary years identified are available at http://www.nber.org/cycles.html. Weclassify a year as recessionary if at least two quarters were identified as recessionary. There is no year in whichonly one quarter was identified as recessionary. The recessionary quarters identified by the NBER differ fromthose identified by the Bureau of Economic Analysis (BEA). The BEA defines a recessionary period as two ad-jacent quarters of economic decline. Using the BEA definition, only 1991 and 2001 are identified as recessionaryyears. We obtain similar results using the BEA definition.21 The major currency index is an index of the value of the U.S. dollar against a basket of major foreign curren-cies, which is computed on a trade-weighted basis (see http://www.federalreserve.gov/releases/H10/summary.html). See Appendix 2 for details of the major currencies index obtained from the Federal Reserve. The FederalReserve also provides a broad index that takes all foreign currencies into consideration in constructing the index.The results are similar when we use the broad index instead of the major currencies index, because the signs of thechanges in the index values are the same.

http://www.nber.org/cycles/recessions.html

http://www.nber.org/cycles/recessions.html

http://www.nber.org/cycles.html

http://www.federalreserve.gov/releases/H10/summary.html

http://www.federalreserve.gov/releases/H10/summary.html


Federal Reserve and compute the average index value for each calendar year. In usingthis index, we assume that all the multinational firms in the sample have the same dis-tribution of business across countries as indicated by U.S. trade flows. While this as-sumption is not likely to hold for small firms that operate in only a few foreigncountries, it may be a good approximation for large multinational companies that oper-ate in many foreign countries. We also use the foreign currency index to infer or proxyforeign economic growth. This is consistent with the premise underlying the hypothesisthat foreign currency exchange rates are associated with economic growth in foreigncountries.

Table 4Domestic growth and foreign translation adjustment.


Coef. t-stat Coef. t-stat

Panel A: U.S. expansionary yearsFTA (β4) 0.68 ⁎⁎⁎ 4.75 0.33 0.83FTA∗RDLeader (β5) 1.72 ⁎⁎⁎ 2.69 1.44 ⁎⁎⁎ 3.41FTA∗HighAsset (β6) 0.71 ⁎ 1.91 1.08 ⁎⁎ 2.42FTA∗HighLabor (β7) −1.58 ⁎⁎⁎ −3.98 −1.11 −1.34RDLeader (β1) 0.04 ⁎⁎ 2.30 0.02 ⁎ 1.52HighAsset (β2) 0.01 0.27 0.05 ⁎⁎⁎ 2.65HighLabor (β3) −0.02 ⁎⁎⁎ −2.31 −0.02 ⁎⁎⁎ −3.15NIADJ (β8) 0.50 ⁎⁎⁎ 24.60 0.35 ⁎⁎⁎ 9.91ΔNIADJ (β9) 0.59 ⁎⁎⁎ 32.21 0.51 ⁎⁎⁎ 17.02FTAX (β10) 5.76 ⁎⁎⁎ 14.48 7.78 ⁎⁎⁎ 6.93TADJ (β11) −0.75 −0.59 1.66 0.42Fixed year dummies Yes YesFixed industry dummies Yes YesNo. of obs. 7084 2279No. (%) of obs. with positive FTA 2660 (37.55%) 904 (39.67%)Adjusted R-square (%) 21.26 15.94

Panel B: U.S. recessionary yearsFTA (β4) −3.30 ⁎⁎⁎ −3.52 −3.62 ⁎⁎⁎ −2.53FTA∗RDLeader (β5) 4.50 1.12 1.61 1.17FTA∗HighAsset (β6) 0.51 1.18 1.39 1.13FTA∗HighLabor (β7) −0.80 −0.31 −1.46 −0.90RDLeader (β1) 0.24 ⁎⁎⁎ 2.97 0.07 1.36HighAsset (β2) −0.06 −1.11 −0.11 ⁎ −1.87HighLabor (β3) 0.02 0.32 −0.13 ⁎⁎ −2.39NIADJ (β8) 0.32 ⁎⁎⁎ 4.83 0.48 ⁎⁎⁎ 6.84ΔNIADJ (β9) 0.40 ⁎⁎⁎ 6.85 0.64 ⁎⁎⁎ 10.98FTAX (β10) 5.18 ⁎⁎⁎ 3.53 9.56 ⁎⁎⁎ 3.40TADJ (β11) −5.88 −1.57 −3.96 ⁎ −1.66Fixed year dummies Yes YesFixed industry dummies Yes YesNo. of obs. 1005 312No. (%) of obs. with positive FTA 493 (49.05%) 112 (35.90%)Adjusted R-square (%) 12.21 10.94

Table 5FTA and changes in the major currency index.

Mean Median Std. dev. Q1 Q3

Years with a negative change in the index 0.0022 0.0000 0.0139 −0.0003 0.0033Years with a positive change in the index −0.0023 −0.0001 0.0113 −0.0034 0.0000


We validate the summary index measure by examining the descriptive statistics for FTAin the years with negative and positive changes in the index. If a negative change in theindex indicates that foreign currencies are strengthening, then FTA should be positive,and vice versa. The descriptive statistics for FTA in the years with negative and positivechanges in the index are presented in Table 5.

Table 5 shows that in the years in which the change in the index is negative (positive)FTA is on average positive (negative). The mean FTA in the negative (positive) years is0.0022 (−0.0023). The Pearson (Spearman) correlation between FTA and the changes inthe index is −0.26 (−0.30) and is statistically significant. Thus, the assumption that on av-erage all the multinational firms in the sample have the same distribution of business acrosscountries as indicated by U.S. trade flows is reasonable.

In the first sub-sample for conditioning on relative growth, we include years in whichthe U.S. economy was growing and the change in the index was negative; that is, boththe U.S. and foreign economies were growing, and the foreign economy was growing ata faster rate on average. This sub-sample includes the following years: 1986, 1987,1988, 1992, 1994, 1995, 1999, 2002, 2003, 2004, 2005, and 2006. We expect the hypoth-esis to be supported for this sub-sample because the foreign country was growing at a fasterrate than the U.S. economy. Our hypothesis is based on this assumption.

Table 6, Panel A provides the estimates of Eq. (1) for the sub-sample when both theU.S. and the foreign economies were growing and the foreign economy was growing at afaster rate on average. The estimated coefficient on FTA∗RDLeader is 2.20 with t=2.17(1.33 with t=2.90) for manufacturing (service) firms. The estimated coefficient onFTA∗HighAsset is 0.22 with t=1.66 (1.30 with t=2.38) for manufacturing (service)firms. These results are consistent with the hypothesis.

Notes to Table 4:

Notes: Eq. (1):

SARETit ¼ β0 þ β1RDLeader þ β2HighAsset þ β3HighLaborþβ4FTAit þ β5FTAit∗RDLeader þ β6FTAit∗HighAsset þ β7FTAit∗HighLaborþβ8NIADJit þ β9ΔNIADJit þ β10FTAXit þ β11TADJit þ error:

Panel A contains the results of estimating Eq. (1) for the U.S. expansionary years (i.e., years excluding 1990, 1991, and2001). Panel B contains the results of estimating Eq. (1) for the U.S. recessionary years (i.e., years 1990, 1991, and 2001).The U.S. expansionary and recessionary years are based on the recession years identified by the Business Cycle DatingCommittee of the National Bureau of Economic Research (NBER). All independent variables in the abovemodel, includ-ing the intercept, are deflated by the firm's beginning market value of equity. The t-statistics are based on standard errorsusing the Huberman–White–Sandwich procedure (see Petersen, 2009) with firm-level clustering. Manufacturing (ser-vice) firms are firms operating in SIC codes 2000–3999 (7200–8799).Variable definitions are contained in Appendix 1.



Table 6Relative growth and foreign translation adjustment.



Panel A: domestic growth and foreign growth (negative change in foreign currency index)FTA (β4) 0.60 ⁎⁎⁎ 4.32 0.10 0.20FTA∗RDLeader (β5) 2.20 ⁎⁎ 2.17 1.33 ⁎⁎⁎ 2.90FTA∗HighAsset (β6) 0.22 ⁎ 1.66 1.30 ⁎⁎ 2.38FTA∗HighLabor (β7) −0.88 ⁎⁎ −2.09 −0.41 −0.47RDLeader (β1) 0.01 0.53 0.03 ⁎ 1.86HighAsset (β2) −0.01 −0.55 0.03 ⁎⁎ 2.27HighLabor (β3) −0.03 ⁎⁎ −2.02 −0.07 −1.54NIADJ (β8) 0.32 ⁎⁎⁎ 11.79 0.30 ⁎⁎⁎ 6.60ΔNIADJ (β9) 0.72 ⁎⁎⁎ 29.85 0.50 ⁎⁎⁎ 13.54FTAX (β10) 5.59 ⁎⁎⁎ 10.30 8.16 ⁎⁎⁎ 5.45TADJ (β11) −0.43 −0.24 1.59 0.32Fixed year dummies Yes YesFixed industry dummies Yes YesNo. of obs. 4160 1357No. (%) of obs. with positive FTA 2012 (48.37%) 653 (48.12)Adjusted R-square (%) 20.51 15.34

Panel B: domestic growth and foreign growth indeterminate (positive change in foreign currency index)FTA (β4) −0.16 −0.49 −0.86 −1.06FTA∗RDLeader (β5) 4.83 ⁎ 1.82 3.89 ⁎⁎ 2.12FTA∗HighAsset (β6) 1.62 ⁎ 1.64 1.07 0.70FTA∗HighLabor (β7) −0.29 −0.22 −0.23 −1.52RDLeader (β1) 0.05 ⁎⁎⁎ 2.88 0.01 0.15HighAsset (β2) −0.01 −0.40 0.08 ⁎⁎⁎ 3.40HighLabor (β3) 0.01 0.05 −0.06 ⁎⁎⁎ −2.60NIADJ (β8) 0.77 ⁎⁎⁎ 23.64 0.63 ⁎⁎⁎ 10.34ΔNIADJ (β9) 0.32 ⁎⁎⁎ 10.66 0.27 ⁎⁎⁎ 5.08FTAX (β10) 5.23 ⁎⁎⁎ 7.76 7.42 ⁎⁎⁎ 4.01TADJ (β11) 0.33 0.16 4.04 0.61Fixed year dummies Yes YesFixed industry dummies Yes YesNo. of obs. 2924 837No. (%) of obs. with positive FTA 736 (25.17%) 220 (26.28%)Adjusted R-square (%) 16.90 14.84


In the second sub-sample, we include the years in which the U.S. economy wasgrowing and foreign economic growth was indeterminate. This sub-sample includesall the years with a positive change in the index and in which the U.S. economy was grow-ing: 1985, 1989, 1993, 1996, 1997, 1998, and 2000.22 We also expect the results for

22 We do not include 1990 or 1991 in this sub-sample. These are the years in our sample in which the U.S. econ-omy was declining and the index was negative. That is, both the U.S. and foreign economies were declining, butthe foreign economy was declining at a slower rate. As reported in Table 4, Panel B, the results for recessionaryyears in the United States do not support the hypothesis, potentially because of the lack of statistical power. Therewas a positive change in the index in 2001, but we also remove this year from the sub-sample as it was a reces-sionary year in the United States.


this sub-sample to be consistent with our hypothesis because FTA provides information onthe growth potential of the foreign country at a more granular level. Table 6, Panel B pro-vides the estimate of Eq. (1) when the U.S. economy was growing but foreign economicgrowth was indeterminate. Conditioned on U.S. economic growth, our hypothesis isagain supported. For manufacturing (service) firms, the estimated coefficients on FTA∗R-DLeader and FTA∗HighAsset are β5=4.83, t=1.82; and β6=1.62, t=1.64 (β5=3.89,t=2.12; and β6=1.07, t=0.70), respectively. This suggests that FTA is likely to providegranular information on the foreign country's growth.

3.5. Robustness tests

3.5.1. Future foreign incomeAs an alternative research design we consider the relationship between FTA and future for-

eign income similar to that in Louis (2003). In particular, we estimate the following equation.

FNIitþ1 ¼ α0 þ α1RDLeader þ α2HighAsset þ α3HighLaborþα4FTAit þ α5FTAit∗RDLeader þ α6FTAit∗HighAsset þ α7FTAit∗HighLaborþα8FNIit þ α9ΔFNIit þ error ð2Þ

where FNI (ΔFNI) is foreign net income (change of FNI from year t−1 to year t) scaled bybeginning market value and all of the other variables are as defined in Eq. (1).

The results are reported in Table 7. Consistent with the rigidity of wages argument andthe results presented in Table 3, the coefficients on the interaction term FTA∗HighLabor(α7) for both the manufacturing and service firms are negative (α7=−0.05, t=−1.68 formanufacturing firms; and α7=−0.21, t=−4.13 for service firms), indicating that foreigntranslation adjustments are negatively associated with next-year foreign earnings forlabor-intensive firms. Consistent with our hypothesis, the coefficient estimates on the inter-action terms FTA∗RDLeader (α5) and FTA∗HighAsset (α6) are both positive: α5=0.13,t=2.97 and α6=0.23, t=5.68 for manufacturing firms, and α5=0.14, t=2.90 andα6=0.20, t=5.56 for services firms. The results are therefore consistent with our

Notes to Table 6:

Notes: Eq. (1):


Panel A contains the results of estimating Eq. (1) for the U.S. expansionary years with negative change in foreigncurrency index (i.e., years 1986, 1987, 1988, 1992, 1994, 1995, 1999, 2002, 2003, 2004, 2005, and 2006). PanelB contains the results of estimating Eq. (1) for the U.S. expansionary years with positive change in foreign cur-rency index, (i.e., years 1985, 1989, 1993, 1996, 1997, 1998, and 2000).The foreign currency index obtained fromthe Federal Reserve is provided in Appendix 2.All independent variables in the above model, including the intercept, aredeflated by the firm's beginningmarket value of equity. The t-statistics are based on standard errors using the Huberman–White–Sandwich procedure (see Petersen, 2009) with firm-level clustering. Manufacturing (service) firms are firms op-erating in SIC codes 2000–3999 (7200–8799).Variable definitions are contained in Appendix 1.



Table 7Foreign translation adjustment and future foreign income.



FTA (α4) −0.16 ⁎ −1.86 0.01 0.01FTA∗RDLeader (α5) 0.13 ⁎⁎⁎ 2.97 0.14 ⁎⁎⁎ 2.90FTA∗HighAsset (α6) 0.23 ⁎⁎⁎ 5.68 0.20 ⁎⁎⁎ 5.56FTA∗HighLabor (α7) −0.05 ⁎ −1.68 −0.21 ⁎⁎⁎ −4.13Intercept (α0) 0.01 ⁎⁎⁎ 11.41 0.01 ⁎⁎⁎ 7.82RDLeader (α1) 0.01 1.51 −0.01 −1.43HighAsset (α2) −0.02 ⁎⁎ −2.25 0.02 1.49HighLabor (α3) −0.01 −0.45 0.01 0.75FNI (α8) 0.60 ⁎⁎⁎ 106.79 0.41 ⁎⁎⁎ 33.67ΔFNI (α9) −0.15 ⁎⁎⁎ −2.63 −0.06 ⁎⁎⁎ −5.00No. of obs. 8089 2506Adjusted R-square (%) 51.43 41.12

Notes: Eq. (2):

FNIitþ1 ¼ α0 þ α1RDLeader þ α2HighAsset þ α3HighLaborþα4FTAit þ α5FTAit∗RDLeader þ α6FTAit∗HighAsset þ α7FTAit∗HighLaborþα8FNIit þ α9ΔFNIit þ error:

The dependent variable is foreign net income in year t+1 scaled by the firm's beginning market value of equity.All independent variables in the above model, including the intercept, are deflated by the firm's beginning marketvalue of equity. The t-statistics are based on standard errors using the Huberman–White–Sandwich procedure (seePetersen, 2009) with firm-level clustering. Manufacturing (service) firms are firms operating in SIC codes 2000–3999 (7200–8799). The manufacturing (service) firms sample contains 8089 (2506) firm-year observations span-ning from 1985 to 2006.




hypothesis that for firms with barriers to entry, foreign translation adjustments are positive-ly associated with next-year foreign earnings.

3.5.2. Other robustness testsWe also conduct several additional robustness tests using the specification in Eq. (1).

We first examine the sample period from 1985 to 2001 to coincide with the period ana-lyzed by Louis (2003). We do so because the coefficient on FTA is positive for our sampleperiod, while Louis (2003) finds it to be negative. We thus examine whether the differencein results is attributable to sample period differences. Table 8, Panel A provides a summaryof the results for this sample period. We find that the coefficient on FTA is negative, indi-cating that sample period differences are likely to have led to the difference in the main ef-fect of FTA, possibly due to a lack of control for domestic and foreign growth. Forexample, in our sub-sample analysis, FTA is negative in the U.S. recessionary years (seeTable 4, Panel B) and is not statistically different from 0 when the U.S. economy is grow-ing but the foreign economy's growth is indeterminate.

Table 8Robustness tests.



Panel A: sample period 1985 to 2001; RDLeader, HighAsset, or HighLabor equals 1 if the industry-adjustedmeasures are above annual median

FTA −0.88 ⁎⁎⁎ −4.31 −0.55 −1.14FTA∗RDLeader 1.28 ⁎⁎⁎ 2.77 1.38 ⁎⁎⁎ 2.97FTA∗HighAsset 2.28 ⁎⁎⁎ 4.97 1.63 ⁎⁎⁎ 3.19FTA∗HighLabor −2.14 ⁎⁎⁎ −3.31 −2.94 ⁎⁎⁎ −2.99

Panel B: RDLeader, HighAsset and HighLabor equals 1 if the industry-adjusted measures greater than 0FTA 0.39 ⁎⁎⁎ 3.00 0.08 0.33FTA∗RDLeader 1.61 ⁎ 1.76 1.53 ⁎⁎⁎ 3.27FTA∗HighAsset 0.22 0.52 1.75 ⁎⁎ 2.11FTA∗HighLabor −1.81 ⁎⁎⁎ −3.45 −0.90 ⁎ −1.87

Panel C: RDLeader, HighAsset and HighLabor equals 1 if intensity levels are above the annual medianFTA 0.23 ⁎⁎ 1.67 1.38 ⁎ 1.91FTA∗RDLeader 2.41 ⁎⁎⁎ 4.33 1.54 ⁎ 1.96FTA∗HighAsset 0.44 ⁎⁎ 2.32 0.42 0.72FTA∗HighLabor −1.79 ⁎⁎⁎ −2.85 −2.77 ⁎⁎ −2.19

Panel D: Fama–MacBeth procedureFTA 0.21 0.44 −0.70 −0.26FTA∗RDLeader 1.56 ⁎⁎ 2.00 3.16 ⁎ 1.88FTA∗HighAsset 1.07 ⁎ 1.94 2.02 1.60FTA∗HighLabor −1.25 ⁎ −1.92 −1.89 ⁎⁎ −2.15

Notes: Eq. (1):


Panel A contains the results of estimating Eq. (1) for the sample period from 1985 to 2001. Panel B contains theresults of estimating Eq. (1) with R&DLeaders, HighAsset, and HighLabor taking on a value of 1, if the corre-sponding industry-adjusted intensity measures are non-negative. Panel C contains the results of estimatingEq. (1) with R&DLeaders, HighAsset, and HighLabor taking on a value of 1, if the corresponding industry-adjust-ed intensity measure is above the median of the annual empirical distribution. Panel D contains the results usingthe Fama–MacBeth procedure.All independent variables in the above model, including the intercept, are deflated by the firm's beginning marketvalue of equity. The t-statistics are based on standard errors using the Huberman–White–Sandwich procedure (seePetersen, 2009) with firm-level clustering.Variable definitions are contained in Appendix 1.




Second, we classify R&D leader, asset-intensive, and labor-intensive firms using 0 rath-er than the median as the benchmark; that is, firms with industry-adjusted R&D capital in-tensity of above 0 are classified as R&D leaders. Similarly, we classify asset-intensive andlabor-intensive firms using 0 as the benchmark. Table 8, Panel B provides a summary of


the results for these alternative measures of barriers to entry and rigidity of wages. The re-sults are qualitatively similar to our main results. However, the barriers-to-entry effect thatis captured by asset intensity is not statistically significant, possibly because in this mea-sure of barriers to entry, fewer firms are classified as asset-intensive firms. Thus, it is likelythat we do not classify some firms that have barriers to entry as asset-intensive firms andinstead include them in the benchmark firms. This reduces the statistical power of the test.

Third, we use the medians of R&D intensity, asset intensity, and labor intensity to clas-sify R&D leader, asset-intensive, and labor-intensive firms. The results are summarized inTable 8, Panel C. We find that the coefficients on R&D leader are consistently positive inour robustness tests, while asset intensity sometimes provides weaker results. Overall, theresults suggest that R&D is a consistent measure of barriers to entry, while asset intensitymay not be such a powerful measure.

In our final robustness test reported in Table 8, Panel D, we use the Fama–MacBethprocedure (Fama & MacBeth, 1973) to estimate Eq. (1) on an annual basis for all years.The mean of the annual coefficient estimates and the associated t-statistics based on thestandard errors of the annual means are reported. The results are qualitatively similar tothose discussed along with Table 3, which again supports our main hypothesis.

4. Concluding remarks

We examine the association of foreign currency translation adjustments with multina-tional firms' stock returns by considering the economic effects of barriers to entry. We de-velop arguments grounded in economic theory that support our hypothesis of the linkbetween changes in foreign currency exchange rates and stock returns. Specifically, an in-crease in the exchange rate for a foreign currency is associated with economic growth inthe foreign country concerned. Economic growth can spur competition for two reasons: in-creased demand and cheaper imports. Firms that operate in environments with barriers toentry can stave off increased competition and thus reap the benefits of economic growth.

We consider firms that are R&D leaders in their industries to be strategically positionedto take advantage of the economic growth in foreign countries. We also consider firms withhigh foreign asset intensity to be operating in an environment with barriers to entry. Forboth manufacturing and service firms, we find that foreign currency translation adjust-ments are positively associated with abnormal stock returns for R&D leaders and foreignasset-intensive firms, while foreign translation adjustments are negatively associatedwith abnormal stock returns for firms with high foreign labor intensity. Our results are ro-bust after controlling for relative economic growth and over different specifications.

This finding shows the importance of considering the economic effects of foreign cur-rency movements in assessing the valuation-relevance of foreign currency translation ac-counting rules. The evidence from Louis (2003) raised concerns about a majorcomponent of accounting income – foreign currency translation adjustments – which aredeemed to be income even though the economic effect of such adjustments is quite the op-posite. Furthermore, our finding shows that while the economic argument put forward byLouis (2003) is valid, the positive component of foreign currency translation adjustmentsin accounting income is indeed a form of income in other economic contexts, in that it rep-resents barriers to entry proxied by innovation and asset intensity. Thus, the accounting


rules governing foreign currency translations generally produce results consistent with theeconomic effects of foreign exchange rate changes.

Our study considers the measure of barriers to entry at the firm level. This measure is aresult of the strategic choices of firms within an industry. However, it is also conceivablethat there are both country-level and industry-level barriers to entry that also affect thevalue-relevance of foreign currency translation adjustments. Future work could examinethe effects of industry- and country-level barriers to entry on the value-relevance of foreigncurrency translation adjustment.

Appendix 1. Variable definitions

SARET
Size-adjusted return computed as the difference between the annual raw returns and companionsize portfolio returns ending 3 months after the fiscal year end
NIADJ
Net income before extraordinary items minus TADJ (Compustat item 18 minus Compustat item150)
ΔNIADJ
The change of NIADJ from year t−1 to year t FTA Foreign translation adjustment calculated as the change of the cumulative translation adjustment
(change in Compustat item 230)
FTAX Foreign income tax (Compustat item 64) TADJ Foreign transaction gains or losses (Compustat item 150) R&D Firm's R&D capital intensity computed by capitalizing and amortizing firm's R&D expenses
over 5 years (see Lev & Sougiannis, 1996) divided by net sales (Compustat item 12)
R&D_ADJ Industry-adjusted R&D capital intensity is the difference between R&D capital intensity and the
weighted average R&D capital intensity of the industry (two-digit SIC) with sales as the weights
RDLeader A dummy variable that equals 1 if the firm is classified as an R&D Leader, and 0 otherwise; a
firm is classified as an R&D leader in year t, if R&D_ADJ in year t−1 is above the industrymedian

ForAsset
The firm's foreign asset intensity is foreign assets divided by foreign sales; foreign asset andforeign sales data are obtained from the Compustat segment database
ForAsset_ADJ
The industry-adjusted foreign asset intensity is the difference between the firm's foreign assetintensity and the weighted average foreign asset intensity of the industry (two-digit SIC) withsales as the weights
HighAsset
A dummy variable that equals 1 if the firm has high foreign asset intensity, and 0 otherwise; afirm is classified as a high asset-intensive firm in year t if ForAsset_ADJ in year t−1 is abovethe industry median
ForLabor
The firm's foreign labor intensity is the number of foreign employees divided by foreign salesobtained from Compustat segment data
ForLabor_ADJ
Industry-adjusted foreign labor intensity is the difference between firm's foreign labor intensityand the weighted average foreign labor intensity of the industry (two-digit SIC) with sales as theweights
HighLabor
A dummy variable that equals 1 if the firm has high foreign labor intensity, and 0 otherwise; afirm is classified as a high labor-intensive firm in year t if ForLabor_ADJ in year t−1 isabove median
Market Valueof Equity

Price per share (Compustat item 199) multiplied by total shares outstanding (Compustat item 25)

Market/Book
The market value of equity divided by book value of common equity (Compustat item 60) Raw Return Annual raw return ending 3 months after the fiscal year end Sales Net sales of the firm in millions (Compustat item 12) Sales Growth
Year t
Net sales (Compustat item 12) in year t minus net sales in year t−1 divided by net sales in year t−1; similarly, Sales Growth t+1 and Sales Growth t+2 are defined


Appendix 2. Foreign currency index

Year
Major currency index Increase of index Decrease of index
1985
122.05 3.90 1986 99.72 −22.33 1987 89.21 −10.50 1988 84.19 −5.03 1989 88.52 4.34 1990 85.15 −3.37 1991 83.65 −1.50 1992 82.51 −1.14 1993 85.76 3.25 1994 85.41 −0.35 1995 81.54 −3.88 1996 86.45 4.92 1997 93.76 7.30 1998 98.84 5.08 1999 98.60 −0.24 2000 105.26 6.66 2001 112.73 7.47 2002 111.10 −1.62 2003 98.00 −13.10 2004 91.00 −7.00 2005 90.81 −0.19 2006 90.74 −0.07
The major currency index is obtained from the Federal Reserve (see http://www.federalreserve.gov/releases/H10/Summary/), which is a weighted average of the foreign exchange values of the U.S. dollar against a subset ofmajor currencies. The currency weights used to compute the summary measures are based on annual U.S. tradedata and are updated and revised annually by Federal Reserve. Adjustments to the weights result in changes topast values of the nominal and real indexes. The currency index listed above is obtained from the release of Fed-eral Reserve, which was updated on January 2, 2007. The Federal Reserve provides the daily and monthly curren-cies index. We compute the average index for each year using the monthly currencies index. The change in indexcolumn is the difference in the average index between year t and year t−1.

References

Baginski, S. P., Lorek, K. S., Willinger, G. L., & Branson, B. C. (1999). The relationship between economic char-acteristics and alternative annual earnings persistence measures. The Accounting Review, 74, 105–120.

Bailliu, J., & King, M. R. (2005). What drives movements in exchange rates? Bank of Canada Review, 27–39(Autumn).

Balassa, B. (1964). The purchasing power parity doctrine. Journal of Political Economy, 72, 584–596.Bartov, E. (1997). Foreign currency exposure of multinational firms: Accounting measures and market valuation.

Contemporary Accounting Research, 14, 623–652.Bhagwati, J. (1984). Why are services cheaper in poor countries? The Economic Journal, 94, 279–286.Borensztein, E., DeGregorio, J., & Lee, J. -W. (1998). How does foreign direct investment affect economic

growth? Journal of International Economics, 45, 115–135.Caves, B. E., & Ghemawat, P. (1992). Identifying mobility barriers. Strategic Management Journal, 13, 1–12.Caves, R. E., & Porter, M. E. (1977). From entry barriers to mobility barriers: Conjectural decisions and contrived

deterrence to new competition. Quarterly Journal of Economics, 91, 241–262.Collins, D. W., & Salatka, W. K. (1993). Noisy accounting earnings signals and earnings response coefficients:

The case of foreign currency accounting. Contemporary Accounting Research, 10, 119–160.

http://www.federalreserve.gov/releases/H10/Summary/

http://www.federalreserve.gov/releases/H10/Summary/


Chen, Y., Rogoff, K., & Rossi, B. (2008). Can exchange rates forecast commodity prices? NBER working paper13901. Cambridge, MA: National Bureau of Economic Research.

Cheung, Y., Chinn, M. D., & Pascual, A. G. (2005). Empirical exchange rate models of the nineties: Are any fit tosurvive? Journal of International Money and Finance, 24, 1150–1175.

Chinn, M. D. (1999). Productivity, government spending and the real exchange rate: Evidence from OECDcountries. In R. MacDonald, & J. Stein (Eds.), Equilibrium exchange rates. Boston: Kluwer AcademicPublishers.

Ciftci, M., Lev, B., & Radhakrishnan, S. (2011). Is R&D mispriced or properly risk-adjusted? Journal of Account-ing, Auditing, and Finance, 26, 81.

Clark, T., Masaaki, K., & Rajaratnam, D. (1999). Exchange rate pass-through and international pricing strategy: Aconceptual framework and research propositions. Journal of International Business Studies, 30, 249–268.

Dhaliwal, D., Subramanyam, K. R., & Trezevant, R. (1999). Is comprehensive income superior to net income as ameasure of firm performance? Journal of Accounting and Economics, 26, 43–67.

Eaton, B., & Lipsey, R. (1981). Capital, commitment, and entry equilibrium. Bell Journal of Economics, 12,593–604.

Fama, E. F., & MacBeth, J. D. (1973). Risk, return and equilibrium: Empirical tests. Journal of Political Economy,81, 607–636.

Froot, K., & Rogoff, K. (1995). Perspectives on PPP and long-run real exchange rates. In G. Grossman, & K. Rogoff(Eds.), The handbook of international economics (pp. 1647–1688)..

Froot, K. A., & Stein, J. C. (1989). Exchange rates and foreign direct investment: An imperfect capital marketsapproach. NBER working paper 2914. Cambridge, MA: National Bureau of Economic Research.

Gilbert, L. R. (1989). The effect of SFAS 52 on earnings expectations. Ph.D. Dissertation, Boston University.Giovanni, A. (1988). Exchange rates and traded goods prices. Journal of International Economics, 24, 45–68.Griffin, P. A., & Castanias, R. P. (1987). Accounting for translation of foreign currencies: The effect of statement

52 on equity analyst. Norwalk, CT: FASB.Griffin, J. M., & Stulz, R. M. (2001). International competition and exchange rate shocks: A cross-country indus-

try analysis of stock returns. The Review of Financial Studies, 14, 215–241.Hall, R. E. (2000). E-capital: The link between the stock market and the labor market in the 1990s. Brookings Pa-

pers on Economic Activity(2), 73–118.Haskel, J., & Wolf, H. (2001). The law of one price — A case study. NBER working paper 8112. Cambridge,

MA: National Bureau of Economic Research.Jorgenson, D. W., Ho, M. S., & Stiroh, K. J. (2005). Growth of U.S. industries and investments in information

technology and higher education. In C. Corrado, J. Haltiwanger, & D. Sichel (Eds.), Measuring capital inthe new economy, Vol. 65. (pp. 403–472)Cambridge, MA: National Bureau of Economic Research: Studiesin Income and Wealth.

Kamerschen, D. (1968). The influence of ownership and control on profit rates. American Economic Review, 58,432–447.

Kravis, I., & Lipsey, R. (1983). Towards an explanation of national price levels. Princeton Studies in InternationalFinance, 52, 1–31.

Krugman, P. (1987). Is free trade passe? The Journal of Economic Perspectives, 1(2), 131–144.Lee, J., & Chinn, M. D. (2006). Current account and real exchange rate dynamics in the G7 countries. Journal of

International Money and Finance, 25, 257–274.Lev, B., & Sougiannis, T. (1996). The capitalization, amortization, and value-relevance of R&D. Journal of Ac-

counting and Economics, 21, 107–138.Lev, B., & Radhakrishnan, S. (2005). The valuation of organization capital. In Corrado, C., Haltiwanger, J., &

Sichel, D. (Eds.), Measuring capital in the new economy, Vol. 65. (pp. 403–472). Cambridge, MA: NationalBureau of Economic Research: Studies in Income and Wealth.

Louis, H. (2003). The value relevance of the foreign translation adjustment. The Accounting Review, 78,1027–1047.

Mark, N. C. (1995). Exchange rates and fundamentals: Evidence on long-horizon predictability. American Eco-nomic Review, 85, 201–218.

Mueller, D. C. (1977). The persistence of profits above the norm. Economica, 44, 380–391.Nakamura, L. (2003). The rise in gross private investment in intangible assets since 1978. Federal Reserve Bank

of Philadelphia. Mimeograph.


O'Hanlon, J. F., & Pope, P. F. (1999). The value-relevance of UK dirty surplus accounting flows. The British Ac-counting Review, 31, 459–482.

Petersen, M. A. (2009). Estimating standard errors in finance panel data sets: Comparing approaches. Review ofFinancial Studies, 22(1), 435–480.

Sabac, F., Scott, T. W., & Weir, H. A. (2005). An investigation of the valuation relevance of alternative foreignexchange disclosures. Contemporary Accounting Research, 22, 1027–1061.

Samuelson, P. (1964). Theoretical notes on trade problems. The Review of Economics and Statistics, 46, 145–154.Shumway, T. (1997). The delisting bias in CRSP data. Journal of Finance, 52, 327–340.Soo, B., & Soo, L. (1994). Accounting for the multinational firm: Is the translation process valued by the stock

market? The Accounting Review, 69, 617–637.Stigler, G. J. (1963). Capital and rates of return in manufacturing industries. Princeton, NJ: Princeton University

Press.White, G., Sondhi, A., & Fried, D. (1998). The analysis and use of financial statements (2nd ed). New York: John

Wiley and Sons.

the valuation-relevance of the foreign translation adjustment: the effect of barriers to entry

Documents