Printer Friendly

Foreign exchange exposure elasticity and financial distress.

Financially distressed firms have limited ability to manage exchange rate exposure over time which could cause their fundamental value to be sensitive to the cash flow volatility related to currency movements. Accordingly, we hypothesize that the likelihood and costs of financial distress help explain cross-sectional variations in return sensitivity to currency movements. We find that the level of exchange rate exposure elasticity is related to proxies for the likelihood of financial distress, growth opportunities, and product uniqueness. Further, firms with a greater likelihood and higher costs of financial distress exhibit greater abnormal returns in response to large exchange rate shocks.

**********

Movements in exchange rates can have a dramatic effect on a firm's financial performance. For example, for the fiscal year ending March 31, 2009, Honda announced that exchange rate changes hurt the firm's profits to a larger extent than did falling sales over the same period (Greimel, 2009). Due to the magnitude of the potential effects of exchange rate exposure, a large body of work examines the scale, as well as sources, of this exposure. Specifically, prior research explores the cross-sectional variation in exchange rate exposure due to differences in firms' competition, hedging, liquidity, leverage, and foreign sales (Allayannis and Ihrig, 2001; Allayannis and Ofek, 2001; Bartram and Karolyi, 2006; Doidge, Griffin, and Williamson, 2006; Francis, Hasan, and Hunter, 2008; He and Ng, 1998; Jorion, 1990; Williamson, 2001). In this paper, we argue that another important factor affecting variation in firms' exchange rate exposure is financial distress.

A key finding of previous research is that although exchange rate movements have significant effects on firms' cash flows and operations, there exist relatively weak correlations between exchange rate movements and stock prices (Griffin and Stulz, 2001; Williamson, 2001). Evidence also suggests that this disparity results as many firms are able to manage their foreign exchange exposure by passing through its effects to customers or by engaging in financial or operational hedging (Carter, Pantzalis, and Simkins, 2003; Bartram, Brown, and Minton, 2010). (1) Thus, for the average firm with sufficient opportunity to manage its foreign exchange risk over time, the long-term valuation effects are likely to be less severe than the short-term effects.

Alternatively, for firms in the left tail of the distribution (i.e., financially distressed firms) currency movements are more likely to significantly affect their fundamental values. We hypothesize this would occur due to the limited ability, or even inability, of these firms to access external capital markets, hedge foreign exchange risks through operational or financial hedging, or to pass through increased costs to their customers.

Specifically, although the average firm may be able to smooth out the effects of unfavorable currency movements on cash flows by accessing the external capital market, financially distressed firms would have more difficulty doing so because of their higher costs of capital. (2) Consequently, these firms would face increased cash flow volatility. Moreover, this inability to raise money externally would ultimately cause financially distressed firms to bypass attractive investment opportunities (Campello, Graham, and Harvey, 2010). Thus, the direct cash flow effect of foreign exchange movements is more likely to lead to an impact on the fundamental value of financially distressed firms. Consistent with this argument, Eisdorfer (2007) shows that for financially distressed firms, shocks to cash flows have a stronger impact on current stock prices than shocks to discount rates.

Second, according to Minton and Schrand (1999), increased cash flow volatility may lead to lower S&P bond ratings, higher yields to maturity, and lower analyst following. Thus, financially distressed firms may have limited ability to hedge the foreign exchange risk in financial markets. For example, a distressed firm would find it harder to hedge currency risk using foreign exchange (FX) derivatives as hedging costs depend upon a firm's creditworthiness. Specifically, a firm with a poor credit rating may have difficulty finding banks willing to enter into multiyear forward contracts. As a result, it may have to resort to option contracts that would cost significantly more (Schoenberger, 2011). Bergbrant and Hunter (2011) find that exchange rate exposure is highly sensitive to credit constraints. Further, financially distressed firms may also have difficulty engaging in financial hedging through the issuance of foreign currency denominated debt due to their higher default risk. (3) This increased difficulty of financial hedging implies that the firm's fundamental value is more likely to be negatively affected by foreign exchange risk. Indeed, in 2002, Xerox announced that its reduced credit rating prevented the firm from hedging the currency risk of its foreign operations resulting in substantially larger losses than expected.

Third, operational hedging, such as moving plants overseas, lending internationally, or engaging in foreign direct investment, may not be feasible for financially distressed firms as this distress often leads to reduced investments. In addition, since these firms may have already started to lose customers (Titman and Wessels, 1988; Molina and Preve, 2009b), they would have fewer opportunities to pass through the negative effects of foreign exchange movements to their customers.

Since financial distress prevents firms from effectively managing and reducing the effects of exchange rate movements on firm value, we conjecture that firms that have greater probabilities and costs of financial distress are likely to have greater exposure to exchange rate risk. Using a sample of US manufacturing firms, we test this central hypothesis in several ways. We begin by examining whether a simple correlation exists between a firm's likelihood of default and its foreign exchange exposure. Using default probabilities derived from Merton's (1974) bond pricing model, we find that firms that are more likely to enter into distress have greater absolute foreign exchange exposure elasticities over the subsequent 36 months, controlling for market returns and prevailing interest rates.

Next, we examine the determinants of firms' currency exposure through a panel regression analysis. Consistent with our hypothesis, we find that a firm's currency exposure elasticity is increasing in proxies for the probability and costs of financial distress, even after controlling for previous determinants such as foreign sales, leverage, and size. In particular, we find that a firm's exchange rate exposure elasticity is positively related to the firm's default probability, market-to-book ratio, capital expenditures, and product uniqueness. These findings remain robust when we use the accounting-based O-Score of Ohlson (1980) as our measure of financial distress instead of the option-based default probability measure.

A further implication of our central hypothesis is that firms with greater probabilities and costs of financial distress and, consequently, greater exposure to exchange rate risk, should observe more pronounced stock price effects around large currency shocks. Consistent with this implication, we find that firms with higher default probabilities, greater growth opportunities, or more unique products have larger absolute abnormal returns at the time of large exchange rate changes.

In an effort to control for the effect of unknown firm characteristics on a firm's foreign exchange exposure, we also examine whether firms that have larger increases in their default probabilities experience greater increases in subsequent exchange rate exposure. Our results indicate that the difference in the change in future absolute exposure elasticities between firms with the highest and lowest increases in current default probability is statistically and economically significant.

Finally, because a higher probability of financial distress may reflect a greater incentive for firms to hedge, we gauge the effects of financial distress when hedging is limited or not possible, as is the case with currencies from developing countries. If the probability and costs of financial distress are simply proxies for hedging, we should observe less of an association between these variables and currency exposure. Alternatively, if our empirical tests detect an exposure net of hedging, we should expect to see a stronger relationship. We find, consistent with our hypothesis, that financial distress is significantly more important to a firm's exposure to developing countries' currencies (in contrast to the currencies of developed countries). This finding also suggests that hedging can potentially add value, especially for those firms with higher probabilities or greater costs of financial distress. (4)

Prior studies identify foreign sales and assets, pass-through, and hedging as important determinants of firms' exchange rate exposures. The primary finding of our paper is that a particular capital market imperfection, namely financial distress, amplifies firms' exposure to exchange rate risks. As a result, our paper adds to the understanding of the sources of, and variations in, exchange rate exposure. Moreover, our paper expands the large body of literature that explores the indirect costs of financial distress (see, for example, Opler and Titman, 1994; Andrade and Kaplan, 1998).

Our paper focuses on the effects of financial distress on foreign exchange exposure (as opposed to other types of risks) for two main reasons. First, as explained above, financial distress may limit a firm's ability to smooth out the effects of foreign exchange risk over time. Consequently, financial distress would be more detrimental to managing a firm's currency risk than to managing other types of risk. Second, currency risk may be highly important to firms as reflected in our anecdotal evidence regarding its effects on Honda and as noted in the hedging that firms undertake to manage that risk. Compared to other types of financial risks, such as interest rate or market risks, evidence suggests that foreign exchange risk has consistently been the most commonly hedged risk among nonfinancial firms. (5) For example, prior studies have found that foreign currency derivatives are the most commonly used class of derivatives. (6) Therefore, our paper echoes the importance of foreign exchange risk management by highlighting the heightened effects of currency risk on financially distressed firms.

The rest of the paper proceeds as follows. In Section I, we describe our data and variables, and present summary statistics. Section II examines the relations between foreign exchange exposure and the probability and costs of financial distress. In Section III, we examine stock price reactions to foreign exchange shocks, while in Section IV, we examine whether changes in currency exposure are driven by changes in firms' default probability. Section V explores the effect of hedging on currency exposure. Finally, we provide our conclusions in Section VI.

I. Data and Methodology

Given the multiple sources of direct and indirect exposure to changes in exchange rates, instead of restricting our analysis to firms with foreign operations, we examine all US manufacturing firms with stock return data available from Center for Research in Security Prices (CRSP) and with financial information available from Compustat from 1980 to 2003. These firms come from 27 industry groups according to the Fama and French (1997) industry classification. To reduce potential effects from outliers, for each year, we only include those firms with assets greater than $50 million dollars and nonnegative book values of equity.

Unless otherwise stated, in this paper, we measure the changes in foreign exchange rates as the percentage changes in the real, trade-weighted USD (US dollar) exchange rate index against major currencies as published by the Federal Reserve. Since this index measures the value of foreign currencies per dollar, an increase in the value of this index indicates an appreciation of the USD. Further, we check the sensitivity of our tests with respect to a number of alternative exchange rate indices and find little difference in the results. In addition, the choice of a nominal or real index is not crucial to the results as inflation is generally only a small component of the total exchange rate variation over a short period of time (Jorion, 1990).

A. Measuring Foreign Exchange Exposure

Previous studies measure firms' foreign exchange exposure by regressing stock returns on the percentage change in exchange rate, while controlling for the market return (Adler and Dumas, 1984; Jorion, 1990; Bodnar and Gentry, 1993; Amihud, 1994; Griffin and Stulz, 2001). Since our principal interest is in the correlation between foreign exchange exposure and the likelihood of financial distress, we need to control for exposure to other types of macroeconomic risk, particularly since firms that are closer to financial distress may also be more sensitive to other types of macroeconomic risk, especially those related to the cost of borrowing. To capture these risks, we employ both a stock market index and a real interest rate measure as controls (Choi and Prasad, 1995). Consequently, we estimate firms' foreign exchange exposure by regressing monthly stock returns on the percentage change in the foreign exchange rate index, the return on the equally weighted CRSP index, and the return on the one-month T-bill less the monthly inflation rate:

[R.sub.j,t] = [[alpha].sub.j] + [[beta].sub.j], [R.sup.fx.sub.t] + [[gamma].sub.j] [EWRET.sub.t] + [[delta].sub.j] [INTEREST.sub.t] + [[epsilon].sub.j,t] for j = 1 ... N, (1)

where [R.sub.j,t] is the stock return for firm j in month t, [R.sup.fx.sub.t] is the percentage change in the exchange rate index for month t, [EWRET.sub.t], is the equally weighted CRSP index return for month t, and [INTEREST.sub.t] is the real interest rate for month t. As Bodnar and Wong (2003) note, since value-weighted market index returns are dominated by large firms that are more likely to conduct foreign business and be net exporters, using value-weighted CRSP index returns to control for macroeconomic risk may bias tests toward a finding of no exposure. Therefore, following Dominguez and Tesar (2001), among others, we control for equal-weighted CRSP index returns when estimating an individual firm's foreign exchange exposure. However, in unreported analyses, we report that our findings on the impact of default risk on firms' foreign exchange exposure are not affected by the choice of CRSP index returns used to control for macroeconomic effects in Equation (1).

The regression coefficient of interest, [[beta].sub.j], can be interpreted as the elasticity of firm j's equity valuation to exchange rate risk, controlling for other macroeconomic risks captured by [EWRET.sub.t], and [INTEREST.sub.t]. Note that [beta] is essentially the foreign exchange exposure elasticity (i.e., the cash flow exposure of foreign currency of a firm divided by its market value). Throughout the paper, we use the term foreign exchange exposure and foreign exchange exposure elasticity interchangeably for brevity. Panel A of Table I presents the summary statistics of the estimated exposure elasticities. In unreported analysis, we find that in each period, about 10.14% of the sample firms are significantly exposed to foreign exchange risk (at the 10% level or higher), with around 48% of these firms having significantly positive exposures. These results are roughly comparable to those in prior studies (see, for example, Jorion, 1990). The average [R.sup.2] for these regressions is about 0.256.

B. Measuring the Probability of Financial Distress

To test our hypothesis that the effects of exchange rate movements on a firm's stock price are increasing with the firm's potential financial distress, we consider accounting-based models, as well as the distance to default model based on Merton's (1974) bond pricing model with equity as a call option on firm value whose strike price is equal to the value of debt (Hillegeist et al., 2004; Vassalou and Xing, 2004; Duffie, Saita, and Wang, 2007; Bharath and Shumway, 2008; Da and Gao, 2010). While both types of models have advantages and disadvantages, the option-based default model has the empirical convenience of allowing us to capture important aspects of default probability without having to rely on specific estimates of the weights on relevant variables. (7) Furthermore, Hillegeist et al. (2004) show that the option-based default probability measure outperforms both the Z-score and the O-score models in predicting bankruptcies during their 1980-2000 sample period. Given its advantages, we employ the Merton (1974) model as our primary model and later conduct robustness checks using Ohlson's (1980) model. Our results are not sensitive to the choice of model.

To derive our measure of default probability, we use the approach followed by prior studies (see, for example, Vassalou and Xing, 2004; Bharath and Shumway, 2008). Specifically, based upon Merton (1974), we consider equity to be a call option on the firm's assets. We derive the market value of equity as follows:

[V.sub.E] = VN([d.sub.1]) - [Fe.sup.-rT]N([d.sub.2]), (2)

where

[d.sub.1] = ln(V/F) + (r + 0.5[[sigma].sup.2.sub.V])T/[[sigma].sub.V] [square root of T], [d.sub.2] = [d.sub.1] - [[sigma].sub.V] [square root of T]. (3)

In Equations (2) and (3), F is the face value of the firm's debt, V and [[sigma].sub.V] are the total values of the firm and its volatility, respectively, r is the risk-free rate, and T is the time to maturity. N(.) is the cumulative density function of the standard normal distribution.

Under the assumption that the value of the firm's assets follows a geometric Brownian motion, the expected default probability (DP) under a normal distribution is given by

DP = N (-(ln(V/F) + ([mu] - 0.5[[sigma].sup.2.sub.V])T/[[sigma].sub.V] [square root of T])), (4)

where [mu] is the expected return of V. To implement this model, we follow the metrics suggested by Bharath and Shumway (2008) and Vassalou and Xing (2004). Following Vassalou and Xing (2004) and Bharath and Shumway (2008), we use daily data to estimate the value of assets as well as their volatility through time. We use an iterative procedure where the initial value of asset volatility is set to the firm's equity volatility adjusted for the proportion of equity in the firm's assets, [mu] is computed as the mean of the change in ln (V). F is equal to "Debt in One Year" plus half of "Long-Term Debt." T is set to 1 year. We measure the risk-free rate using the 1-year Treasury bill rate obtained from the Federal Reserve. Also following Bharath and Shumway (2008), we delete those observations for which Debt in One Year or Long-Term Debt is negative or zero. In addition, when calculating the volatility of a firm's asset returns, we require that the firm have at least 50 daily return observations. It should be noted that the default probability measure does not merely proxy for accounting ratios or firms' capital structure. For example, even if two firms have the same debt to equity ratio, their default probabilities may differ if they have different asset volatilities.

As demonstrated in Panel A of Table I, from 1980 to 2003, the average US manufacturing firm in our sample has a mean monthly default probability of approximately 2.09% with a standard deviation of 7.56% and a median close to zero. These estimates are slightly smaller than those of previous studies due to our sample restriction on minimum firm asset value. Our estimates would be much closer to those in prior studies if we did not delete firms with assets of less than $50 million (see, for example, Vassalou and Xing, 2004; Da and Gao, 2010). Since we eliminate the smaller firms that, on average, have higher probabilities of default, this sample restriction would bias against our results.

To examine the distribution of our sample firms' default probabilities, each year, we rank firms into deciles according to their DP measure and we calculate each decile's average default probability. We then take an average over time for each decile. Figure 1 illustrates the average default probabilities for each decile. Since the difference in average default probability between firms in the top decile (18.17%) and the bottom decile (close to zero) is quite dramatic, we chart the natural log of average default probability for each decile. The figure indicates that the distribution of DP is skewed in that firms in the top decile have substantially higher default probabilities. In Section III, we take this skewness into account by examining whether small to moderate default probabilities also matter for firms' foreign exchange exposure.

C. Other Firm and Industry Characteristics

If exchange rate movement affects fundamental values primarily when the resulting short-term cash flow fluctuations force financially distressed firms to forgo valuable investment opportunities, then stock prices of firms with greater costs of financial distress should exhibit added sensitivities with exchange rate movements. We employ several measures of the costs of financial distress. We first consider a firm's growth opportunities. The reason is that cash flow volatility caused by exchange rate movement may be more costly for firms with greater investment opportunities due to their larger underinvestment costs. We employ two alternative measures of a firm's growth opportunities: 1) the firm's market-to-book (M/B) ratio, calculated as the sum of market equity plus book debt over the book value of assets, and 2) the firm's capital expenditures as a percentage of net property, plant, and equipment. (8) Both M/B and capital expenditures are winsorized at the 1st and 99th percentiles to minimize the impact of outliers.

The cost of financial distress may also depend upon the degree to which the firm has specialized products. As discussed in Titman and Wessels (1988), financial distress can be more costly for firms with specialized products since the consequence of disrupted long run relationships with customers and business partners is more severe. Following Titman and Wessels (1988), we use the ratio of research and development (R&D) expenses to total assets to proxy for the degree of product specialization. (9)

Since previous studies indicate that firms' foreign sales significantly affect the sensitivity of stock returns to exchange rate movement (Jorion, 1990; Bodnar and Wong, 2003; Doidge et al., 2006), we control for foreign sales in our analysis. We measure a firm's foreign activities as the sum of foreign sales and export sales normalized by the firm's total sales using information from the Compustat Geographic Segment file. Firms without foreign sales records are considered to have zero foreign sales. Given the lack of detailed information on individual firms' trade balances, the effect of foreign sales on a firm's absolute exposure is unclear. We try to alleviate this issue by breaking down foreign sales into two series. Specifically, we calculate the difference between individual firms' percentage foreign sales and the average value across all sample firms during the same year, setting FSales_Hi to a firm's excess foreign sales if it has above average foreign sales during the year (and zero otherwise), and setting FSales_Lo to a firm's excess foreign sales if it has below average foreign sales during the year (and zero otherwise). (10, 11) Since firms with above (below) average foreign sales are more likely to be net exporters (importers), we expect the absolute exposure' to be significantly increasing in FSales_Hi and significantly decreasing in FSales_Lo.

Next, we control for a firm's capital structure, sales to market value ratio (SP), and size. A firm's foreign exchange exposure may be related to the firm's leverage through a magnification effect. Since foreign exchange risk may have a direct effect on a firm's cash flow, the more levered the firm, the greater an effect we may observe from its equity value. We control for the market leverage ratio, measured as the book value of debt over the market value of total assets. (11) Since our tests measure the effect of exchange rate changes on a firm's market value instead of cash flow, a firm with lower market values of equity relative to the size of cash flow exposures is likely to have greater exposure elasticity. To account for this scaling effect, we also control for the ratio of sales to the market value of equity. Given that firm size is related to the extent of foreign business activities, as well as the likelihood of hedging, we control for firm size, measured as the logarithm of the firm's market capitalization. In addition, since we analyze the correlation between foreign exchange exposure elasticity and financial distress, controlling for the market value of the firm may also alleviate the concern that the actual cash flow exposures of distressed firms are overestimated as they often have lower market values and greater exposure elasticities, all else being equal.

Finally, we include industry dummies in multivariate analyses. These dummies control for factors related to industry structure and competition, such as markup and pass-through effects, that are important determinants of foreign exchange exposure (Allayannis and Ihrig, 2001; Bodnar, Dumas, and Marston, 2002; Bartram et al., 2010). Moreover, these industry dummies may also capture differences in firms' hedging incentives related to the elasticity of demand and the convexity of production costs (Adam, Dasgupta, and Titman, 2007).

Panel B of Table I reports the average cross-sectional correlations between firm characteristics and the absolute value of the estimated future exposures over the sample period. As expected, firms' absolute exposures are positively correlated with their default probability, M/B ratio, capital expenditures, and R&D expenses. Firms that have greater leverage, with a greater sales to market value ratio, or smaller market capitalization, also tend to have greater absolute foreign exchange exposure.

II. Foreign Exchange Exposure and Financial Distress

A. Industry Level Foreign Exchange Exposure

Our first analysis explores the relation between a firm's foreign exchange exposure and its default probability. We conduct this test at the industry level as previous studies find that industry characteristics are an important component of firms' currency exposure. In addition, industry level analyses may help control for differences in firms' growth opportunities or asset tangibility, which may be related to the cost of financial distress. Specifically, using the Fama-French (1997) industry classifications, we divide our sample firms into 22 manufacturing industry sectors. To ensure test accuracy, in this analysis, we only include those industries that have at least three firms within each of the three default probability (DP) groups and that exist in the sample for at least 10 years. These filters result in the elimination of five industries from the original 27 manufacturing industries used in other analyses.

We then employ rolling regressions to shed light on the relationship between the current level of financial distress and future foreign exchange exposure. Within each industry sector, for each sample year, we sort firms into three equally sized groups according to their previous year's average monthly DPs. For each of the three DP groups, we estimate the individual firms' foreign exchange exposures according to Equation (1), where we use the firms' monthly stock returns over the subsequent 36 months. A firm is required to have at least 30 monthly observations during the estimation period to be included in the analysis. We then calculate the average absolute exposure for the group. We repeat this process for each year.

Table II presents the time series averages of the absolute foreign exchange exposures for low, medium, and high default probability groups in each of the 22 industries, along with the average default probability for each group. As previously mentioned, since we cannot differentiate between firms that are net importers versus net exporters, we report absolute exposures. (12) The results in Table II provide strong support for our hypothesis that the equity valuations of firms with higher probabilities of default are more sensitive to foreign exchange movement. In every industry, the magnitude of the absolute exchange rate exposure is larger for the high DP group than for the low DP group with the difference being statistically significant for 18 of the 22 industries. Given the overlapping nature of the observations in the annual rolling regressions, the Mest statistics are calculated based upon Newey-West (1987) autocorrelation and heteroskedasticity-consistent standard errors with two lags.

B. Determinants of Foreign Exchange Exposures

Table II establishes that there is a significant relationship between a firm's probability of financial distress and its exchange rate exposure. Next, we examine the question of whether this relationship continues to exist after controlling for other possible determinants of foreign exchange exposure including those firm and industry characteristics that contribute to a firm's cost of financial distress.

To control for time variation in exposure, we employ a two-step approach for our examination of whether the current probability and costs of financial distress are associated with greater foreign exchange exposure in the future. In the first step, we estimate firms' foreign exchange exposure over a series of three-year windows according to Equation (1). We move this three-year estimation window ahead one year at a time to create a time series of estimated exposures for each firm. In the second step, we match the estimated exposures to the beginning-of-period firm and industry characteristics. We then run a panel regression of estimated absolute exposures on the probability of financial distress and firm characteristics over the 1980-2001 sample period. The panel approach takes into account the fact that foreign exchange exposures vary over time, as well as in the cross-section (see, e.g., Allayannis, 1997; Francis et al., 2008).

More specifically, we run the following regression:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII], (5)

where Default Prob, M/B, CapExp, and R&D denote firm i's default probability, market-to-book ratio, capital expenditures over net property, plant, and equipment, and R&D over assets at the beginning of the three-year period t. FSales denotes the sum of foreign sales and export sales normalized by the firm's total sales. Leverage is defined as the ratio of the book value of debt over total assets, SP is the ratio of annual sales over the market value of equity, and Size denotes the logarithm of the firm's market value at the beginning of period t. Lastly, [Ind.sub.it,n] are 26 industry dummies that control for potential industry effects. Since the exposures are estimated from the first-stage regression and may be serially correlated, we calculate the t-statistics of the panel regression using heteroskedasticity robust standard errors adjusted for both firm and year clustering.

The results of estimating Equation (5) are presented in Table III, column 1. Consistent with a positive relation between foreign exchange exposure and the likelihood of financial distress, we find that the magnitude of firms' exchange rate exposure is significantly positively related to their default probabilities indicating that firms that have a greater probability of financial distress are more affected by exchange rate movements. This relationship is not only statistically significant, it is also economically significant. A one standard deviation increase in the likelihood of financial distress increases foreign exchange exposure by 0.093. In addition, we find that consistent with a positive relation between firms' foreign exchange exposure and the costs of financial distress, firms with higher M/B ratios, greater capital expenditures, or higher R&D expenses appear to have significantly larger absolute foreign exchange rate exposures. The table also shows that more levered firms have greater exposure, consistent with the magnifying effect of leverage. Alternatively, larger firms appear to have smaller exchange rate exposure. This may be due to the fact that larger firms have more resources to hedge against currency risk. In addition, this result is potentially attributed to the fact that low (high) market value firms (holding their cash flow exposure constant) have mechanically higher (lower) exposure elasticities given that we measure foreign exchange exposure elasticity, as opposed to foreign exchange exposure. Finally, we do not find firms with high cash flow exposure relative to equity to have larger exposure elasticity as indicated by the negative coefficient on the sales to market value ratio.

In column 2 of Table III, we rerun Equation (5) replacing Foreign Sales with FSales_Hi and FSales_Lo. Consistent with our expectations, a firm's excess foreign sales significantly increases its absolute exposure if the firm is estimated to be a net exporter (i.e., above average level of foreign sales), whereas excess foreign sales significantly reduces the firm's absolute exposure if it is estimated to be a net importer (i.e., a below average level of foreign sales).

Since the default probability is a nonuniform measure that becomes significantly larger for those firms in severe financial distress, it is important to investigate whether our main results in columns 1 and 2 are driven by a few extreme DP observations. To address this issue, we split DP into two variables: 1) Default_Top, which is equal to DP if a firm's default probability is ranked among the top decile during the year and zero otherwise, and 2) Default_Other that is set to DP if a firm's default probability is ranked among the other nine deciles during the year and zero otherwise. To facilitate the comparison of these variables' relative effects on foreign exchange exposure, we standardize these variables such that they have a mean equal to zero and a standard deviation equal to one across the entire sample. The results, reported in column 3 of Table III, suggest that both extremely high levels of default probability and more moderate to mild levels of default probability have a significant effect on firms' foreign exchange exposure. However, the effect of Default_Top is stronger than that of Default_Other suggesting that, as expected, exceptionally high default probability has the most pronounced impact on firms' foreign exchange exposure.

To examine whether financial distress has a differential effect on net importers versus net exporters, we re-estimate the effects separately for firms with positive versus negative exposure. In this analysis, we include Foreign Sales rather than FSales_Hi and FSales_Lo, as the single variable allows us to examine the relationship between signed exposures and foreign sales. The results of this analysis are reported in Table IV. According to our hypothesis, higher probability and the cost of financial distress should cause positive exposures to be more positive and negative exposures to be more negative. As expected, Table IV indicates that although the financial distress measures have slightly different effects for firms with positive versus negative exposure, they tend to enhance the exchange rate exposure for both groups of firms. That is, among firms with positive exposure, those with higher probability or costs of financial distress have significantly larger exposure. An analogous effect is observed among firms with negative exposure. Also worth noting is the finding that while foreign sales significantly increase the magnitude of foreign exchange exposure for net exporters, it has an insignificant effect on the currency exposure for net importers. This result may be due to the possibility that among net importers, those firms with more foreign sales are multinational firms that also have a greater proportion of foreign-produced inputs or foreign liabilities.

Our baseline hypothesis regarding the relation between firms' foreign exchange exposure and financial distress could potentially apply to exposures resulting from both favorable and unfavorable foreign exchange movement. For example, while an appreciation of the US dollar may lead to cash flow shortfalls for a net exporter and, as such, have a more pronounced impact on firm fundamentals if the firm is in financial distress, a depreciation of the US dollar may generate valuable financial slack thereby loosening the firm's financial constraints. This positive effect of favorable exchange rate movement may be particularly helpful at the margin for firms in financial distress. To test whether the effect of financial distress is symmetric between favorable and unfavorable foreign exchange exposure, we condition the effect of default probability on the nature of exchange rate movement. Specifically, we interact default probability with a dummy variable indicating significant dollar depreciation (Depreciation) for net importers or a dummy variable representing significant dollar appreciation (Appreciation) for net exporters during the period exchange rate exposure is measured and repeat the analysis in Table IV. Appreciation (Depreciation) is set to one if the average monthly exchange rate movement during the 36-month period foreign exchange exposure is ranked among the top one-third in magnitude among all periods in which the US dollar has appreciated (depreciated), and zero otherwise.

The results in the last two columns of the table suggest that the probability of financial distress tends to have a greater impact on net exporters' foreign exchange exposure during periods when the US dollar appreciates significantly. That is, financially distressed exporters suffer particularly from the adverse effects of exchange rate movement. In contrast, we do not observe the same asymmetry for net importers. This difference between exporters and importers is possibly due to the fact that product prices are often sticky in the buyer's currency (Goldberg and Knetter, 1997; Betts and Devereux, 2000). When the US dollar appreciates relative to a foreign currency that the firm is exposed to, US exporters are likely to suffer more relative to those US importers facing a dollar depreciation of a similar magnitude.

C. Alternative Measures of Financial Distress

To verify whether our main results above (Table III) are sensitive to the choice of financial distress measure, in this section, we consider an alternative measure of financial distress, namely Ohlson's (1980) O-Score, which is a composite accounting measure for the probability of financial distress. (13) Based on our sample of firms from 1980 to 2003, the O-Score has a mean of -1.7662 and a standard deviation of 1.8834.

In Table V, we re-examine the determinants of foreign exchange exposure by replacing the default probabilities in Equation (5) with Ohlson's (1980) beginning-of-period O-Scores. Consistent with our results using default probabilities, we find that firms with larger O-Scores (i.e., greater probabilities of financial distress) tend to be more sensitive to foreign exchange risk. A one standard deviation increase in the average firm's O-Score results in a 0.1006 increase in its exposure. The results regarding the costs of financial distress are also similar to our earlier findings. One notable departure from our earlier results is that the leverage ratio becomes marginally significant. A possible explanation for this result is that the effect of leverage is being picked up by the O-Score, as total liability is an important component of the O-Score. This observation makes the option-based default probability measure a more attractive measure of the likelihood of financial distress as its correlation with the leverage ratio is only 0.35 as compared to the much higher correlation of 0.68 between the O-Score and the leverage ratio.

III. Stock Price Reaction to Exchange Rate Shocks

To the extent that exchange rate movement has the potential to cause severe liquidity problems or affect the fundamental value of a firm, more pronounced effects are likely to be observed during periods of large currency movement. Consequently, in this section, we employ event study analysis to examine whether stock price reactions to large exchange rate shocks vary systematically with firms' short-term cash flow sensitivities. One advantage to this approach is that we avoid imposing a specific structure on the relation between currency risk and firm value. It also allows us to focus on the cross-sectional variation in foreign exchange exposure at a point in time without imposing a fixed exposure estimate over time (as is done in the estimation of Equation 1). Using a similar approach to examine the abnormal returns of US firms exposed to a large unexpected change in the currencies of Mexico and Thailand, Dewenter, Higgins, and Simin (2005) suggest that the failure of prior studies to find a stronger contemporaneous price response to exchange rate movement can be partially attributed to methodological problems.

Therefore, we examine the relation between stock price reactions to large currency movement and proxies for a firm's probability and costs of financial distress. Stock prices may react to currency movement through the influence of broad macroeconomic factors that affect firms in general, in addition to reactions through their direct exposure to foreign exchange fluctuations. Because of this, we are primarily interested in the cross-sectional variation in stock price reactions around large currency movements rather than the magnitude of the stock price reactions.

To identify days with large currency movements, we begin by calculating the mean and standard deviation of the daily percentage changes in real exchange rates from 1980 to 2003. The sample mean daily percentage change in exchange rate is -0.0011 %. The standard deviation is 0.4235%. We define large currency movements as those daily real exchange rate changes that are more than three standard deviations, in absolute value, from the sample mean daily movement. A total of 67 days fall outside the three standard deviations boundary. For each of the 67 days, we calculate individual firms' cumulative abnormal returns (CAR) over the three-day window (t-1, t+1). Since we use a three-day window to calculate the firms' cumulative abnormal returns, we drop three large currency movement days whose event windows overlap with those of other event days. Further, since a large shock in one direction may be accompanied by a reversal in subsequent days during periods when exchange rate volatility is high, we calculate the average abnormal returns across all events within a month of any given event to account for the offsetting effect of opposing shocks. This procedure leaves us with 46 days associated with large currency movements.

For each event day, we estimate the market model using a 200-day window that ends 40 days before the event, with the return on the CRSP equally weighted market index as the proxy for the market return. Next, we calculate the abnormal returns for the three days around the event day. Over the 46 event days, on average, sample firms observe a three-day cumulative abnormal return of 3.69% in absolute value. Again, we focus on the absolute value of CAR as exchange rate movement should have an opposite effect across net importers and exporters, but our data do not allow us to distinguish between these two types.

For each large foreign exchange movement event, we regress the absolute value of CAR (in percent) on our proxies for a firm's probability and costs of financial distress, controlling for previous year-end firm characteristics and industry dummies:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]. (6)

The independent variables in Equation (6) are default probability, M/B ratio, capital expenditures, R&D over assets, excess foreign sales, the leverage ratio, sales to market value ratio, and firm size. We also include industry dummies (although we do not report their coefficients for brevity). When estimating Equation (6), the dependent and independent variables are scaled by the standard deviation of the residuals from the pre-event estimation to account for potential heteroskedasticity across firms.

The time series mean estimates and Fama-MacBeth (1973) /-statistics from this weighted least squares analysis are presented in Table VI. The results in column 1 indicate that abnormal returns around large exchange rate movements are significantly higher for firms that are more likely to default, consistent with our hypothesis that the more financially distressed a firm is, the more its value is affected by exchange rate movements. Further, as also predicted by our hypothesis, firms with more growth opportunities and greater capital expenditures and R&D expenses are more affected by large exchange rate movements. The economic importance of financial distress on firms' foreign exchange exposure is more apparent following large foreign exchange movements. Finally, we find similar results using a longer 11-day window of (t-5, t+5) as reported in column 2 of Table VI.

To avoid the potential look-ahead bias associated with defining large exchange rate shocks according to sample standard deviation, we employ an alternative approach whereby we define large changes as changes in the exchange rate greater than 1.5%. The results of this robustness analysis are reported in columns 3 and 4 of Table VI for three-day and 11-day windows, respectively. Although the significance of some control variables is reduced due to the much smaller number of events (18) in the analysis, our main findings remain largely unchanged.

IV. The Effect of Changes of Default Probability

In earlier analyses, we attempt to avoid the spurious effect of financial distress by controlling for firm characteristics that are correlated with financial distress, but that also affect foreign exchange exposure, such as firm size and leverage. To further alleviate concerns about the effect of omitted firm characteristics that may drive our results, in this section, we attempt to identify the causal effect of financial distress on foreign exchange exposure by relating changes in a firm's default probability to its subsequent foreign exchange exposure. To do so, each year, we calculate the change in a firm's default probability relative to three years prior. Then, we examine whether changes in the firm's foreign exchange exposure over the subsequent three years are in line with changes in its current default probability.

Specifically, each year, we rank firms into quintiles based on changes in their default probabilities over the previous three years (i.e., differences between year t-4 and year t-1). Within each quintile group, we calculate each firm's change in the absolute foreign exchange exposure estimated for the period t to t+2 from that estimated for the period t-3 to t-1. Finally, within each DP change quintile, we examine whether average changes in firms' foreign exchange exposures over the subsequent three years are in line with changes in their current default probability. Note that we measure the change in absolute exposure rather than signed exposure as a firm may switch from being a net importer to a net exporter for exogenous reasons. However, if the firm's default probability has not changed significantly, its current negative exposure should be comparable to its earlier positive exposure in magnitude, everything else being equal.

Table VII indicates that during our sample period, firms' absolute exposures have exhibited an upward trend over time. As a result, the change in an average firm's absolute exposure relative to three years ago is significantly positive. This is likely due to the overall increase in foreign trade for US firms as the global market has become more integrated. However, although firms' absolute exposures have increased slightly on average, those firms that have experienced significant increases in their default probabilities observe much greater increases in absolute foreign exchange exposure. The difference in the changes in absolute exposures between Quintile Five and Quintile One firms is as large as 12.55%, and is statistically significant at the 1% level based on the paired /-test. Since this analysis effectively controls for the impact of unknown firm characteristics, the results lend strong support to our main hypothesis that default probability has a direct effect on the size of firms' foreign exchange exposure.

V. Hedging and Firms' Exchange Rate Exposures

Some studies in the literature on corporate hedging argue that firms with low short-term liquidity, greater growth opportunities, and tighter financial constraints are more likely to use currency derivatives (see, for example, Nance, Smith, and Smithson, 1993; Geczy, Minton, and Schrand, 1997). Therefore, it is possible that firm characteristics related to the probability and the costs of financial distress may correlate with incentives for firms to hedge foreign exchange risk. However, Mian (1996), Brown (2001) and Bartram, Brown, and Fehle (2009) find little evidence that the use of currency derivatives is related to financial distress cost. Further, even with hedging, some studies find that firms cannot eliminate their foreign exchange exposure as the potential effects of hedging on firm risk and firm value are small and most firms engage in currency hedging selectively (see, for example, Guay and Kothari, 2003; Hentschel and Kothari, 2001; Bodnar, Hayt, and Marston, 1998).

In this paper, we implicitly assume that hedging cannot fully insulate firms from currency risk. Further, firms that have a higher probability of financial distress may have greater difficulty entering into hedged positions as illustrated by the case of Xerox (see the Introduction section). Alternatively, to the extent that firms with greater costs of financial distress are more likely to hedge their currency exposure, our empirical findings understate the effect of financial distress as we only examine its correlation with firms' exposure net of potential operational or financial hedging activities.

To gauge the effects of financial distress in the relative absence of hedging, we examine the relationship between the probability of distress and firms' exposure to currencies that are more difficult to hedge. To the extent that the design of this test allows us to detect a firm's exposure net of hedging, we should expect to see a stronger relation between firms' probability and costs of financial distress and their exposure to these currencies (in contrast to major currencies, which are relatively easier to hedge). Following Francis et al. (2008), we proxy for currencies that are more difficult to hedge using a real trade-weighted USD exchange rate index against the currencies of 19 developing economies defined by the Board of Governors of the Federal Reserve Bank as "other important trading partners" (OITP). (14,15)

Using the same two-step procedure employed earlier in Section III, we estimate foreign exchange exposure by regressing firms' monthly stock returns on both the percentage change of the real USD exchange rate index against the OITP currencies and the percentage change of the real USD exchange rate index against major currencies during each three-year period, controlling for returns on the equally weighted CRSP index and the real interest rate. We find that firms' average absolute exposure to OITP currencies is about 1.85, slightly higher than the average estimated exposure to major currencies (1.01). This finding is consistent with Francis et al. (2008) who find that the overall average absolute OITP betas are approximately double those of the major index. The OITP index includes currencies of developing economies such as China, Mexico, and India, whose weights in overall US international trade have climbed significantly in recent years. (16) In addition to these currencies' growing importance in foreign trade, US firms may be more exposed to these currencies due to greater difficulty when hedging against the risk (through either financial or natural hedging). (17,18)

Column 1 of Table VIII presents the estimates from Equation (5) when the absolute value of estimated exposures to OITP currencies is regressed on current firm characteristics. Again, we report t-statistics calculated with robust standard errors adjusted for both firm and year clustering. In general, we continue to find a positive relation between the absolute exposure to OITP and the probability and costs of financial distress. More importantly, this relation appears to be even stronger than what we observe from firms' exposures to major currencies as reported in Table III.

For comparison, in column 2 of Table VIII, we report the estimation results for the determinants of firms' absolute exposures to major currencies (estimated simultaneously with exposures to the OITP index) and formally test the differences in coefficients for default probability, market-to-book ratio, capital expenditures, and R&D expenses across these two estimations. The relation between default probability and firms' exposure to more difficult-to-hedge currencies is more than three times as strong as that for firms' exposure to major currencies. The difference is significant at the 1% level using a [chi square] test that corrects for firm-level clustering. We also find stronger relations between firms' growth opportunities and product uniqueness and their exposure to OITP. The differences in the coefficients for M/B ratio, capital expenditures, and R&D expenses are all statistically significant across the two models.

VI. Conclusion

In this paper, we examine the hypothesis that market frictions caused by financial distress can affect the exchange rate exposure of US manufacturing firms. Specifically, the cash flow effects of exchange rate movement are more likely to affect the fundamental value of distressed firms given other firms' ability to access external capital markets or to conduct operational and financial hedging in the longer term. Based on this insight, we conjecture that the sensitivity of a firm's stock price to currency risk is greater for firms that are more likely to face financial distress or that have higher costs of financial distress.

Controlling for macroeconomic risks, we provide evidence that firms with higher default probabilities, greater growth opportunities, and more unique products exhibit higher foreign exchange exposure. This finding continues to hold when we replace our primary measure of financial distress, namely the option-based measure based on Merton (1974), with the accounting-based O-Score of Ohlson (1980). We further find that these firms demonstrate greater exposure to large, unexpected exchange rate movement as indicated by their event-period abnormal returns. In addition, we find that greater changes in firms' current default probability are associated with greater changes in their foreign exchange exposure in the subsequent period. Finally, by comparing firms' exposure to currencies of developed and developing countries, we find that firms with higher probability or costs of financial distress are more vulnerable to foreign exchange risk when their ability to hedge this risk is limited. As such, we provide evidence supporting the value of foreign exchange risk management.

Appendix

Currency Weights in the Broad Dollar Index

Country                       1997     2003

Euro Area                    17.49     18.80
Canada                       16.92     16.43
Japan                        14.27     10.58
United Kingdom                5.73      5.17
Switzerland                   1.43      1.44
Australia                     1.31      1.25
Sweden                        1.22      1.16
Major Currencies Subtotal    58.37     54.84
China                         6.58     11.35
Mexico                        8.50     10.04
Korea                         3.68      3.86
Taiwan                        3.77      2.87
Hong Kong                     2.65      2.33
Malaysia                      2.25      2.24
Singapore                     2.87      2.12
Brazil                        1.82      1.79
Thailand                      1.59      1.43
India                         0.88      1.14
Philippines                   1.18      1.06
Israel                        0.84      1.00
Indonesia                     1.25      0.95
Russia                        0.78      0.74
Saudi Arabia                  0.80      0.61
Chile                         0.53      0.49
Argentina                     0.61      0.44
Colombia                      0.49      0.41
Venezuela                     0.58      0.30
OITP Currencies Subtotal     41.63     45.16

Source: Federal Reserve, 2008. H10 Foreign Exchange Rates Previous
Currency Weights. At http://www/federalreserve.gov/Releases/H10/
Weights/previousweights.htm.


The authors would like to thank Bill Christie (Editor), two anonymous referees, Andres Almazan, Warren Bailey, Sohnke Bartram, Greg Brown, Alex Butler, Fang Cai, Ty Callahan, Katheryn Dewenter, Li Gan, Jane Ihrig, Stephen Magee, Bernadette Minton, Sridhar Sundaram, Sheridan Titman, Hong Yan, and the participants at seminars at the University of Texas at Austin, the FMA meetings, and the WFA meetings for helpful comments.

References

Adam, T., S. Dasgupta, and S. Titman, 2007, "Financial Constraints, Competition and Hedging in Industry Equilibrium," Journal of Finance 62, 2445-2473.

Adler, M. and B. Dumas, 1984, "Exposure of Currency Risk: Definition and Measurement," Financial Management 13,41-50.

Allayannis, G., 1997, "The Time-Variation of the Exchange Rate Exposure: An Industry Analysis," Darden Graduate School of Business, University of Virginia Working Paper.

Allayannis, G. and J. Ihrig, 2001, "Exposures and Markups," Review of Financial Studies 14, 805-835.

Allayannis, G. and E. Ofek, 2001, "Exchange Rate Exposure, Hedging, and the Use of Foreign Currency Derivatives," Journal of International Money and Finance 20, 273-296.

Allayannis, G. and J.P. Weston, 2001, "The Use of Foreign Currency Derivatives and Firm Market Value," Review of Financial Studies 14, 243-276.

Amihud, Y., 1994, "Exchange Rates and the Valuation of Equity Shares," in Y. Amihud and R.M. Levich, Eds. Exchange Rates and Corporate Performance, New York, NY, Irwin.

Andrade, G. and S.N. Kaplan, 1998, "How Costly is Financial (not Economic) Distress? Evidence from Highly Leveraged Transactions that Became Distressed," Journal of Finance 53, 1443-1494.

Atanasova, C., 2007, "Access to Institutional Finance and the Use of Trade Credit," Financial Management 36, 49-67.

Bartram, S.M., G. Brown, and F.R. Fehle, 2009, "International Evidence on Financial Derivatives Usage," Financial Management 38, 185-206.

Bartram, S.M., G. Brown, and B. Minton, 2010, "Resolving the Exposure Puzzle: The Many Facets of Exchange Rate Exposure," Journal of Financial Economics 95, 148-173.

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

Bergbrant, M. and D.M. Hunter, 2011, "Credit Market Constraints and Firms' Exchange Rate Exposure," University of South Florida Working Paper.

Berrospide, J., A. Purnanandam, and U. Rajan, 2011, "Corporate Hedging, Investment and Value," University of Michigan Working Paper.

Betts, C. and M.B. Devereux, 2000, "The Exchange Rate Dynamics in a Model of Pricing-to-Market," Journal of International Economics 50, 215-244.

Bharath, S.T. and T. Shumway, 2008, "Forecasting Default with the Merton Distance-to-Default Model," Review of Financial Studies 21, 1339-1369.

Bodnar, G.M., B. Dumas, and R. Marston, 2002, "Pass-Through and Exposure," Journal of Finance 57, 199-231.

Bodnar, G.M. and W.M. Gentry, 1993, "Exchange Rate Exposure and Industry Characteristics: Evidence from Canada, Japan, and the USA Journal of International Money and Finance 12, 29-45.

Bodnar, G.M., G. Hayt, and R.C. Marston, 1996, "1995 Wharton Survey of Derivative Usage by US Non-Financial Firms," Financial Management 25, 113-133.

Bodnar, G.M., G. Hayt, and R. Marston, 1998, "1998 Survey of Risk Management by US Non-financial Firms," Financial Management 27, 70-91.

Bodnar, G.M, and M.H. Wong, 2003, "Estimating Exchange Rate Exposures: Issues in Model Structure," Financial Management 32, 35-67.

Brown, G., 2001, "Managing Foreign Exchange Risk with Derivatives," Journal of Financial Economics 60, 401-448.

Campello, M., J.R. Graham, and C.R. Harvey, 2010, "The Real Effects of Financial Constraints: Evidence from a Financial Crisis," Journal of Financial Economics 97, 470-487.

Carter, A.D., D.A. Rogers, and B. Simkins, 2006, "Does Hedging Affect Firm Value? Evidence from the US Airline Industry," Financial Management 35, 53-86.

Carter, D., C. Pantzalis, and B. Simkins, 2003, "Asymmetric Exposure to Foreign-Exchange Risk: Financial and Real Options Implemented by US Multinationals," University of Oklahoma Working Paper.

Choi, J.J. and A.M. Prasad, 1995, "Exchange Risk Sensitivity and Its Determinants: A Firm and Industry Analysis of US Multinationals," Financial Management 24, 77-88.

Covitz, D.M. and S.A. Sharpe, 2005, "Do Nonfinancial Firms Use Interest Rate Derivatives to Hedge," Finance and Economics Discussion Series Paper, Federal Reserve Board.

Da, Z. and P. Gao, 2010, "Clientele Change, Liquidity Shock, and the Return on Financially Distressed Stocks," Journal of Financial and Quantitative Analysis 45, 27-48.

Dewenter, K.L., R.C. Higgins, and T.T. Simin, 2005, "Can Event Study Methods Solve the Currency Exposure Puzzle?" Pacific-Basin Finance Journal 13, 119-144.

Doidge, C., J.M. Griffin, and R. Williamson, 2006, "Measuring the Economic Importance of Foreign Exchange Exposure," Journal of Empirical Finance 13, 550-576.

Dominguez, K.M. and L.L. Tesar, 2001, "A Reexamination of Exchange Rate Exposure," American Economic Review 91, 396-399.

Duffie, D., L. Saita, and K. Wang, 2007, "Multi-Period Corporate Failure Prediction with Stochastic Covariates," Journal of Financial Economics 83, 635-665.

Eisdorfer, A., 2007, "The Importance of Cash-Flow News for Financially Distressed Firms," Financial Management 36, 33-48.

Fama, E.F. and K.R. French, 1997, "Industry Costs of Equity ''Journal of Financial Economics 43, 153-194.

Fama, E.F. and J. MacBeth, 1973, "Risk, Return, and Equilibrium: Empirical Tests," Journal of Political Economy 71, 607-636.

Faulkender, M., 2005, "Hedging or Market Timing? Selecting the Interest Rate Exposure of Corporate Debt," Journal of Finance 60, 931-962.

Federal Reserve, 2008, "H.10 Foreign Exchange Rates: Broad Index of the Foreign Exchange Value of the Dollar," http://www.federalreserve.gOv/Releases/H 10/Weights/previousweights.htm.

Francis, B., I. Hasan, and D. Hunter, 2008, "Can Hedging Tell the Full Story? Reconciling Differences in US Aggregate- and Industry-Level Exchange Rate Risk Premium," Journal of Financial Economics 90, 169-196.

Geczy, C., B.A. Minton, and C. Schrand, 1997, "Why Firms Use Currency Derivatives," Journal of Finance 52, 1323-1354.

Goldberg, P.K. and M. Knetter, 1997, "Goods Prices and Exchange Rates: What Have We Learned?" Journal of Economic Literature 35, 1243-1272.

Greimel, H., 2009, "Honda: First Quarterly Loss in 15 Years," Automotive News 83, 30B.

Griffin, J.M. and R.M. Stulz, 2001, "International Competition and Exchange Rate Shocks: A Cross-County Industry Analysis of Stock Returns," Review of Financial Studies 14, 215-241.

Guay, W. and S.R Kothari, 2003, "How Much Do Firms Hedge with Derivatives?" Journal of Financial Economics 70, 423-461.

He, J. and L.K. Ng, 1998, "The Foreign Exchange Exposure of Japanese Multinational Corporations," Journal of Finance 53, 733-753.

Hentschel, L. and S.P. Kothari, 2001, "Are Corporations Reducing or Taking Risks with Derivatives?" Journal of Financial and Quantitative Analysis 36, 93-118.

Hillegeist, S.A., E.K. Keating, D.P. Cram, and K.G. Lundstedt, 2004, "Assessing the Probability of Bankruptcy," Review of Accounting Studies 9, 5-34.

Jin, Y. and P. Jorion, 2006, "Firm Value and Hedging: Evidence from US Oil and Gas Producers," Journal of Finance 61, 893-919.

Jorion, P, 1990, "The Exchange-Rate Exposure of US Multinationals," Journal of Business 63, 331-345.

Merton, R.C., 1974, "On the Pricing of Corporate Debt: The Risk Structure of Interest Rates," Journal of Finance 29, 449-470.

Mian, S., 1996, "Evidence on Corporate Hedging Policy," Journal of Financial and Quantitative Analysis 31,419-439.

Minton, B.A. and C. Schrand, 1999, "The Impact of Cash Flow Volatility on Discretionary Investment and the Costs of Debt and Equity Financing," Journal of Financial Economics 54, 423-460.

Molina, C.A. and L.A. Preve, 2009a, "An Empirical Analysis of the Effect of Financial Distress on Trade Credit," Financial Management 41, 187-205.

Molina, C.A. and L.A. Preve, 2009b, "Trade Receivables Policy of Distressed Firms and its Effect on the Costs of Financial Distress," Financial Management 38, 663-686.

Nance, D.R., C.W. Smith, Jr., and C.W. Smithson, 1993, "On the Determinants of Corporate Hedging," Journal of Finance 48, 267-284.

Newey, W.K. and K.D. West, 1987, "A Simple, Positive Semidefinite Heteroskedasticity and Autocorrelation Consistent Covariance Matrix," Econometrica 55, 703-708.

Ohlson, J., 1980, "Financial Ratios and the Probabilistic Prediction of Bankruptcy," Journal of Accounting Research 18, 109-131.

Opler, T. and S. Titman, 1994, "Financial Distress and Corporate Performance," Journal of Finance 49, 1015-1040.

Schoenberger, C., 2011, "Exposed! As Currency Volatility Rises, Companies Scramble to Avoid Being Caught with Their Hedges Down," Wall Street Journal 02 Mar 2011: C. 10.

Titman, S. and R. Wessels, 1988, "The Determination of Capital Structure Choice," Journal of Finance 43, 1-19.

Vassalou, M. and Y. Xing, 2004, "Default Risk in Equity Returns," Journal of Finance 59, 831-868.

Williamson, R., 2001, "Exchange Rate Exposure and Competition: Evidence from the World Automotive Industry," Journal of Financial Economics 59, 441-475.

(1) Specifically, Bartram et al. (2010) find that pass-through activities and operational hedging may each reduce a firm's foreign exchange exposure by 10% to 15%, and that financial hedging may further reduce foreign exchange exposure by 40%.

(2) For example, Atanasova (2007) and Molina and Preve (2009a) find that financially distressed firms have to replace financial debt and equity with more expensive sources of financing such as trade credit.

(3) Bartram et al. (2010) find that financial hedging with foreign currency denominated debt appears to have an even larger effect on foreign exchange exposure than the use of FX derivatives.

(4) Allayannis and Weston (2001) and Berrospide, Purnanandam, and Rajan (2011) show that the use of foreign currency derivatives increases firm value. Using evidence from the airline industry, Carter, Rogers, and Simkins (2006) find that hedging is positively related to airline firm value. Based upon the hedging activities of US oil and gas producers, Jin and Jorion (2006) indicate that while derivatives reduce a firm's stock price sensitivity to oil and gas prices, they do not affect the sample firms' Q-ratios.

(5) For example, Faulkender (2005) finds that firms' interest rate risk management activities are primarily driven by market timing rather than hedging considerations. Furthermore, Covitz and Sharpe (2005) demonstrate no evidence that nonfinancial firms hedge the interest rate exposure from their operating assets.

(6) See, for example, Bodnar, Hayt, and Marston (1996, 1998) and Bartram, Brown, and Fehle (2009). Further, as argued in Bodnar et al. (1998), as not all surveyed firms face FX risk, the usage of currency derivatives, conditional on having an exposure, will be even higher than the 83% they document for all firms.

(7) While the accounting models are inherently backward looking as they are estimated using financial statement data, they place weaker restrictions on the functional form of default probability. However, the weights on the variables used by the existing accounting-based models may be sample specific, changing over time as accounting systems and firms evolve.

(8) We calculate the market value of equity as common shares outstanding multiplied by price, and the book value of debt as total liabilities plus preferred stock minus deferred taxes and convertible debt.

(9) We assign missing R&D observations a value of zero to preserve the number of observations. There is no material change to our results, however, if we eliminate those firm-years with missing R&D values.

(10) Note that we measure firms' excess foreign sales because it helps differentiate firms with below average foreign sales from firms with zero foreign sales when constructing FSales_Hi and Fsales_Lo.

(11) We calculate the market value of assets as book assets minus book equity plus the market value of equity. Book debt is calculated as indicated above.

(12) Since we present the absolute values of foreign exchange exposures in Table I, the magnitudes appear larger than the average exposure reported in previous studies (Jorion, 1990). Using the average exposure across both net importers and net exporters results in offsetting exposures that would make the average smaller for those studies.

(13) The O-score is defined as:

O-score = -1.32 - 0.4071og(total assets) + 6.03 (total liability/total assets) - 1.43 (working capital/total assets) + 0.076 (current liability/current assets) - 1.72 (1 if total liabilities > total assets, 0 if other wise) -2.37 (net Income/total assets) - 1.83 (funds from operation/total liability) + 0.285 (1 if a net loss for the last two years, 0 otherwise) - 0.521 ([net income.sub.t] - [net income.sub.t-1]/[absolute value of [net income.sub.t]] - [absolute value of [net income.sub.t-1]]).

(14) The list of countries in this index and the major index are provided in the appendix.

(15) During our sample period, the correlation between the real dollar exchange rate indexes against the major currencies and the currencies of the OITP is approximately 0.20. The standard deviation of changes in the major index is 0.018, while that of changes in the OITP index is 0.012.

(16) By 2003, the 19 OITP currencies accounted for 45.16% of the broad US dollar trade-weighted index that incorporates all currencies in the major currencies index and the OITP index (Federal Reserve, 200S).

(17) For example, Mexican peso options and futures did not start trading at CME until 1995, and while Chinese RMB contracts were not available to trade until 2006.

(18) For example, US foreign direct investments into OITP countries are significantly smaller than those into its major trade partners. See http://stats.oecd.org/Index.aspx?DataSetCode=FDI_FLOW_PARTNER for more details.

Kelsey D. Wei and Laura T. Starks *

* Kelsey D. Wei is an Assistant Professor in the Naveen Jindal School of Management at the University of Texas at Dallas in Richardson, TX. Laura T. Starks is the Charles E. and Sarah M. Seay Regents Chair in the McCombs School of Business at the University of Texas at Austin in Austin, TX.

Table I. Summary Statistics of Firms' Foreign Exchange Exposures and
Characteristics

Panel A of the table reports the time series average of the
cross-sectional mean, median, standard deviation, the 25th and 75th
percentile values, and the minimum and maximum values of the foreign
exchange exposure estimated using Equation (1), its absolute value
during the series of three-year estimation periods, and the beginning
of period firm characteristics over the 1980-2003 sample period.
Default Prob. is the percentage default probability estimated using
Equation (4). M/B is the market-to-book ratio. CapExp is the firm's
capital expenditures normalized by property, plant, and equipment.
R&D/Assets is the firm's R&D expenses normalized by the total assets.
Foreign Sales is measured as the percentage of foreign sales plus
export sales over total sales. Leverage is defined as the book value
of debt over the market value of total assets. SP is the ratio of
sales to market capitalization. Size is measured as the firm's market
value of equity in millions of dollars. Panel B presents the
correlation matrix between these variables. The correlations reported
are time series averages of the Pearson correlations calculated for
each three-year estimation period.

Panel A. Summary Statistics of Main Variables

                 Mean     Median     STDEV      25h

Exposure        -0.0221   -0.0387    1.2937   -0.7200
abs(Exposure)    0.9511    0.7114    0.8941    0.3295
Default Prob.    2.0930    0.0022    7.5639    0.0000
M/B              1.5731    1.2645    0.9349    1.0073
CapExp           0.2452    0.2074    0.1481    0.1477
R&D/Assets       0.0368    0.0157    0.0575    0.0000
Foreign Sales    0.1977    0.1527    0.2002    0.0000
Leverage         0.3820    0.3653    0.2089    0.2126
SP               2.3607    1.6504    2.3320    0.9015
Size             1589      235       5515      85

                 75th       Min       Max

Exposure        0.6639    -6.1588     7.5189
abs(Exposure)   1.3039     0.0013     8.1120
Default Prob.   0.3447     0.0000    80.9920
M/B             1.7872     0.5974     7.0497
CapExp          0.3021     0.0332     0.8637
R&D/Assets      0.0503     0.0000     0.6622
Foreign Sales   0.3412     0.0000     0.9589
Leverage        0.5327     0.0145     0.9446
SP              2.9638     0.0454    16.0317
Size              828        6        77,242

Panel B. Correlations between Main Variables

                   abs       Exposure    Default     M/B
                (Exposure)                Prob.

abs(Exposure)    1.0000
Exposure         0.0317       1.0000
Default Prob.    0.1326       0.0262      1.0000
M/B              0.0050      -0.0485     -0.146     1.0000
CapExp           0.1182      -0.0337     -0.0933    0.3738
R&D/Assets       0.1244      -0.0461     -0.0599    0.3008
Foreign Sales    0.0151      -0.1076     -0.0588    0.0882
Leverage         0.0483       0.0713      0.3528   -0.6448
SP               0.0666       0.0651      0.3977   -0.4604
Size            -0.1027      -0.04       -0.077     0.2118

                  CapExp       R&D/      Foreign   Leverage
                              Assets      Sales

abs(Exposure)
Exposure
Default Prob.
M/B
CapExp           1.0000
R&D/Assets       0.3489       1.0000
Foreign Sales    0.1059       0.3036      1.0000
Leverage        -0.3734      -0.3331     -0.1065    1.0000
SP              -0.2421      -0.2545     -0.1966    0.7544
Size            -0.0029       0.0715      0.2195   -0.1160

                    SP         Size

abs(Exposure)
Exposure
Default Prob.
M/B
CapExp
R&D/Assets
Foreign Sales
Leverage
SP               1.0000
Size            -0.1471      1.0000

Table II. The Relation between Absolute Exchange Rate Exposures and
Default Probabilities by Industry

We group manufacturing firms by their Fama and French (1997) industry
classifications. At the beginning of each year, we sort the firms
within each industry sector into terciles according to their default
probabilities. Within each default probability group, we estimate each
firm's foreign exchange exposure using monthly stock returns in the
subsequent 36 months according to Equation (1). This process is
repeated for every year covering the total time period from 1980 to
2003. The table presents the time series averages of the absolute
foreign exchange exposures for the low, medium, and high default
probability groups along with their average default probability (in
italics) for each of the 22 manufacturing industries. Wald test
statistics, calculated with Newey-West (1987) robust standard errors
for the difference in exchange exposure coefficients between the low
and high groups, are reported in parentheses.

Industry                            Low     Medium     High

Food Products                     0.5453#   0.7460    1.0075
                                  0.0000#   0.0002#   0.0814#
Recreation                        0.9708    1.1969    1.4376
                                  0.0001#   0.0091#   0.1349#
Printing and Publishing           0.5508    0.6892    0.9663
                                  0.0000#   0.0000#   0.0503#
Consumer Goods                    0.6709    0.8827    1.1723
                                  0.0000#   0.0003#   0.0946#
Apparel                           0.7870    0.9492    1.3399
                                  0.0000#   0.0019#   0.1163#
Medical Equipment                 0.8205    1.0099    1.2238
                                  0.0000#   0.0001#   0.0709#
Pharmaceutical Products           0.7579    1.1906    1.2923
                                  0.0000#   0.0000#   0.0329#
Chemicals                         0.5675    0.6869    0.9647
                                  0.0000#   0.0001#   0.0599#
Rubber and Plastic Products       0.7667    0.8973    1.1267
                                  0.0001#   0.0042#   0.1349#
Textile                           0.8242    0.9718    1.0454
                                  0.0004#   0.0176#   0.1418#
Construction Materials            0.6959    0.8104    1.0077
                                  0.0000#   0.0007#   0.0945#
Steel Works Etc                   0.6515    0.8322    1.1081
                                  0.0000#   0.0062#   0.1203#
Fabricated Products               0.8821    1.0996    1.3625
                                  0.0005#   0.0057#   0.1117#
Machinery                         0.7190    0.9158    1.2313
                                  0.0000#   0.0002#   0.0738#
Electrical Equipment              0.6032    0.8010    1.1010
                                  0.0000#   0.0004#   0.0487#
Automobiles and Trucks            0.6648    0.8151    1.0184
                                  0.0000#   0.0028#   0.1007#
Aircraft                          0.8123    0.7018    0.9954
                                  0.0000#   0.0021#   0.0946#
Computers                         1.0806    1.3220    1.4339
                                  0.0000#   0.0006#   0.0847#
Electronic Equipment              1.0025    1.2435    1.4067
                                  0.0000#   0.0004#   0.0706#
Measuring and Control Equipment   0.8189    1.0134    1.2290
                                  0.0000#   0.0001#   0.0419#
Business Supplies                 0.6796    0.6582    0.8391
                                  0.0000#   0.0001#   0.0498#

Shipping Containers               0.5951    1.2617    1.1009
                                  0.0001#   0.0039#   0.0742#

Industry                            Low-High
                                  (Wald stat.)

Food Products                      -0.4622 ***
                                  (13.90)
Recreation                         -0.4669 **
                                   (4.52)
Printing and Publishing            -0.4156 ***
                                  (13.94)
Consumer Goods                     -0.5014 ***
                                   (9.77)
Apparel                            -0.5529 ***
                                  (10.76)
Medical Equipment                  -0.4034 *
                                   (3.68)
Pharmaceutical Products            -0.5344 **
                                   (3.96)
Chemicals                          -0.3972 ***
                                  (39.03)
Rubber and Plastic Products        -0.3600 **
                                   (4.01)
Textile                            -0.2212
                                   (0.82)
Construction Materials             -0.3118 ***
                                  (12.27)
Steel Works Etc                    -0.4566 ***
                                  (12.59)
Fabricated Products                -0.4804
                                   (2.16)
Machinery                          -0.5123 ***
                                  (15.15)
Electrical Equipment               -0.4979 ***
                                  (25.09)
Automobiles and Trucks             -0.3536 **
                                   (3.93)
Aircraft                           -0.1269
                                   (1.05)
Computers                          -0.3532 *
                                   (3.03)
Electronic Equipment               -0.4043 *
                                   (3.27)
Measuring and Control Equipment    -0.4101' **
                                   (6.85)
Business Supplies                  -0.1596
                                   (2.10)
Shipping Containers                -0.5058 ***
                                  (11.78)

*** Significant at the 0.01 level.

** Significant at the 0.05 level.

* Significant at the 0.10 level.

Note: The table presents the time series averages of the absolute
foreign exchange exposures for the low, medium, and high default
probability groups along with their average default probability (in
italics) for each of the 22 manufacturing industries indicated with #.

Table III. Determinants of Foreign Exchange Exposures

During each three-year estimation period, firms' monthly stock returns
are regressed on the percentage changes of the real foreign exchange
rate index, returns on the equally weighted CRSP index, and returns on
the one-month T-bill rate minus the monthly inflation rates. The
absolute value of the estimated exposure is then regressed on the
firms' beginning of period default probability, market-to-book ratio,
capital expenditures normalized by property, plant, and equipment, R&D
expense-assets, foreign sales as a percentage of total sales, market
leverage ratio, the ratio of sales to market capitalization, the
logarithm of firm market value, and industry dummy variables. In model
(2), we break foreign sales into two series: 1) FSales_Hi and 2)
FSales_Lo. FSales_Hi is set to a firm's excess foreign sales (relative
to the sample average) if it has an above average level of foreign
sales during the year, and zero if otherwise. FSales_Lo is set to a
firm's excess foreign sales if it has a below average level of foreign
sales during the year, and zero if otherwise. In model (3), we split
DP into two variables: 1) Default_Top and 2) Default_Other.
Default_Top is equal to DP if a firm's default probability is ranked
among the top decile during the year, and zero otherwise.
DefaultjDther is set to DP if a firm's default probability is ranked
among the other nine deciles during the year, and zero otherwise. Both
DefaultJTop and DefaultjDther are standardized to have a mean of zero
and a standard deviation of one. f-statistics for coefficient
estimates, calculated with robust standard errors adjusted for both
firm and year clustering, are reported in parentheses. Number of
observations and S-square are reported at the bottom of the table.

Model                     (1)            (2)            (3)

Intercept               0.8842 ***     0.6685 ***     0.7048 ***
                       (7.58)         (5.17)         (5.44)
Default Probability     1.2258 ***     1.1940 ***
                       (6.09)         (6.26)
Default_Top                                           0.0958 ***
                                                     (6.39)
Default_Other                                         0.0471 ***
                                                     (3.33)
M/B                     0.1223 ***     0.1134 ***     0.1127 ***
                       (5.59)         (5.45)         (5.43)
CapExp                  0.5808 ***     0.5708 ***     0.5711 ***
                       (4.19)         (4.29)         (4.34)
R&D/Assets              1.6897 ***     1.6866 ***     1.6675 ***
                       (4.72)         (4.59)         (4.55)
Foreign Sales           0.2542 ***
                       (4.04)
Fsales_Hi                              0.6082 ***     0.6015 ***
                                      (7.17)         (7.24)
Fsales_Lo                             -1.3758 ***    -1.3480 ***
                                     (-6.63)        (-6.53)
Leverage                0.5473 ***     0.5871 ***     0.5501 ***
                       (5.46)         (6.03)         (5.61)
SP                     -0.0347 ***    -0.0361 ***    -0.0350 ***
                      (-3.10)        (-3.29)        (-3.27)
Ln(Size)               -0.0995 ***    -0.0847 ***    -0.0836 ***
                      (-9.41)        (-7.05)        (-7.04)
Industry dummy            Yes            Yes            Yes
N                        19,770         19,770         19,770
[R.sup.2]               0.098          0.104          0.104

*** Significant at the 0.01 level.

Table IV. Determinants of Signed Foreign Exchange Exposures for Net
Importers and Exporters

During each three-year estimation period, firms' monthly stock returns
are regressed on the percentage changes of the foreign exchange rate
index, returns on the equally weighted CRSP index, and returns on the
one-month T-bill rate minus the monthly inflation rates. The estimated
exposures are then regressed on the firms' beginning of period default
probability, market-to-book ratio, capital expenditures normalized by
property, plant, and equipment, R&D expense/assets, foreign sales as a
percentage of total sales, the ratio of sales to market
capitalization, the logarithm of firm market value, and industry dummy
variables, r-statistics for coefficient estimates, calculated with
robust standard errors adjusted for both firm and year clustering, are
reported in parentheses. Number of observations and T-square are
reported at the bottom of the table.

Model                          (1) Importer   (2) Exporter

Intercept                        0.8862 ***    -0.8944 ***
                                (7.70)        (-6.09)
Default Probability              1.1431 ***    -1.3192 ***
                                (4.81)        (-5.38)
Default Prob. * Depreciation

Default Prob. * Appreciation

Depreciation

Appreciation

M/B                              0.1257 ***    -0.1161 ***
                                (5.09)        (-4.76)
CapExp                           0.5245 ***    -0.6380 ***
                                (3.38)        (-3.77)
R&D/Assets                       2.2185 ***    -1.2345 ***
                                (4.35)        (-2.99)
Foreign Sales                    0.0027        -0.4660 ***
                                (0.03)        (-4.25)
Leverage                         0.7565 ***    -0.3501 **
                                (6.06)        (-2.49)
SP                              -0.0386 ***     0.0334 **
                               (-2.91)         (2.57)
Ln(Size)                        -0.1125 ***     0.0880 ***
                               (-7.33)         (5.85)
Industry dummy                     Yes            Yes
N                                 9,601          10,169
[R.sup.2]                        0.102          0.101

Model                          (3) Importer   (4) Exporter

Intercept                        0.8890 ***    -0.8743 ***
                                (7.58)        (-6.24)
Default Probability              1.1738 ***    -1.1758 ***
                                (4.04)        (-5.25)
Default Prob. * Depreciation    -0.0327
                               (-0.10)
Default Prob. * Appreciation                   -1.6409 **
                                              (-2.24)
Depreciation                    -0.0771
                               (-0.60)
Appreciation                                   -0.1255
                                              (-0.72)
M/B                              0.1259 ***    -0.1171 ***
                                (5.15)        (-4.87)
CapExp                           0.5296 ***    -0.6006 ***
                                (3.40)        (-3.45)
R&D/Assets                       2.2130 ***    -1.2454 ***
                                (4.35)        (-3.00)
Foreign Sales                   -0.0013        -0.4631 ***
                               (-0.02)        (-4.12)
Leverage                         0.7547 ***    -0.3544 **
                                (6.02)        (-2.51)
SP                              -0.0384 ***     0.0374 ***
                               (-2.88)         (2.75)
Ln(Size)                        -0.1111 ***     0.0876 ***
                               (-7.16)         (5.75)
Industry dummy                     Yes            Yes
N                                 9,601          10,169
[R.sup.2]                        0.106          0.112

*** Significant at the 0.01 level.

** Significant at the 0.05 level.

Table V. The Relation between Foreign Exchange Exposures and O-Scores

During each three-year estimation period, firms' monthly stock returns
are regressed on the percentage change of the foreign exchange rate
index, returns on the equally weighted CRSP index, and returns on the
one-month T-bill rate minus the monthly inflation rates. We then
regress the absolute value of the estimated exposure on a firm's
beginning of period Ohlson's (1980) O-Score, the market-to-book ratio,
capital expenditures normalized by property, plant, and equipment, R&D
expense-assets, two variables measuring above average (FSales_Hi) and
below average (Fsales_Lo) foreign sales, the market leverage ratio,
the ratio of sales to market capitalization, the logarithm of the firm
market value, and industry dummy variables. FSales_Hi is set to a
firm's excess foreign sales (relative to the sample average) if it has
an above average level of foreign sales during the year, and zero
otherwise. FSalesJLo is set to a firm's excess foreign sales if it has
a below average level of foreign sales during the year, and zero
otherwise, t-statistics for coefficient estimates, calculated with
robust standard errors adjusted for both firm and year clustering, are
reported in parentheses. Number of observations and R-square are
reported at the bottom of the table.

Intercept                0.8273 ***
                        (6.14)
O-Score                  0.0534 ***
                        (3.75)
M/B                      0.1059 ***
                        (6.30)
CapExp                   0.3610 ***
                        (2.77)
R&D/Assets               1.4142 ***
                        (3.99)
Fsales_Hi                0.5980 ***
                        (5.85)
Fsales_Lo               -1.1575 ***
                       (-5.48)
Leverage                 0.2367 *
                        (1.70)
SP                      -0.0136
                       (-1.32)
Ln(Size)                -0.0748 ***
                       (-5.10)
Industry dummy              Yes
N                         17,588
[R.sup.2]                0.090

*** Significant at the 0.01 level.

* Significant at the 0.10 level.

Table VI. Reaction of Stock Prices to Large Exchange Rate Shocks

For each of the events over the 1980-2003 sample period with daily
percentage real exchange rate changes at least three standard
deviations from the sample mean or 1.5% in absolute value, the
absolute value of the percentage cumulative abnormal return of
individual firms is regressed against previous year-end firm
characteristics and industry dummies using weighted least squares
regressions. The average coefficients from these regressions are
reported, along with Fama-MacBeth (1973) t-statistics, in parenthesis.
Average R-squares are reported at the bottom of the table.

                           (1)              (2)

                      FX < -3[sigma] or FX > 3 [sigma]

                      (t - 1, t + 1)   (t - 5, t + 5)

Intercept               2.2791 ***       4.9896 ***
                       (9.81)          (12.73)
Default Probability     3.2479 ***       5.8310 ***
                      (10.32)           (7.73)
M/B                     0.1470 ***       0.3388 ***
                       (4.37)           (5.58)
CapExp                  0.7281 ***       1.7626 ***
                       (4.32)           (5.94)
R&D/Assets              3.2523 ***       6.4681 ***
                       (4.98)           (5.69)
Fsales_Hi               0.5691 ***       0.6629 ***
                       (3.36)           (2.75)
Fsales_Lo              -1.2552 ***      -1.4517 ***
                      (-4.63)          (-2.96)
Leverage                0.6697 ***       1.6245 ***
                       (6.41)           (8.19)
SP                      0.0086          -0.0125
                       (0.58)          (-0.53)
Ln(Size)               -0.1276 ***      -0.3299 ***
                      (-5.35)          (-8.44)
Industry dummy             Yes              Yes
[R.sup.2]               0.058            0.071

                           (3)              (4)

                         FX < -1.5%  or FX > 1.5%

                      (t - 1, t + 1)   (t - 5, t + 5)

Intercept               1.7474 ***       4.3784 ***
                       (6.72)           (7.08)
Default Probability     3.0716 ***       5.9043 ***
                       (5.08)           (4.39)
M/B                     0.2226 ***       0.4059 ***
                       (4.67)           (4.50)
CapExp                  0.9131 ***       2.0377 ***
                       (3.30)           (4.85)
R&D/Assets              3.1116 ***       5.5852 ***
                       (3.04)           (2.98)
Fsales_Hi               0.3469           0.6322 *
                       (1.08)           (1.77)
Fsales_Lo              -0.6447          -1.8444 **
                      (-1.65)          (-2.32)
Leverage                0.7280 ***       1.7151 ***
                       (4.13)           (4.99)
SP                      0.0259          -0.0139
                       (1.12)          (-0.31)
Ln(Size)               -0.0854 **       -0.2533 ***
                      (-2.48)          (-3.40)
Industry dummy             Yes              Yes
[R.sup.2]               0.061            0.072

*** Significant at the 0.01 level.

** Significant at the 0.05 level.

* Significant at the 0.10 level.

Table VII. Changes of Default Probability and Changes of Foreign
Exchange Exposure

Each year, we calculate the change in a firm's default probability
relative to that of three years ago. We then rank firms into quintiles
based upon these changes in their default probabilities. Within each
quintile group, we calculate each firm's change of absolute foreign
exchange exposure estimated during the subsequent three-year period
from the absolute exposure estimated during the prior three-year
period. Finally, we calculate the average change of default
probability (in percentages) and the average change of absolute
exposure for each quintile group. The difference in these measures
between the top and bottom quintiles and the associated t-statistics
(in parentheses) are reported in the bottom of the table.

Group             [DELTA]    [DELTA]absfx
                   DP (%)

1                 -5.7302    0.0112
                             (0.22)
2                 -0.0221    0.0555
                             (1.25)
3                  0.0013    0.0946 **
                             (2.46)
4                  0.1105    0.1082 **
                             (2.49)
5                  7.6613    0.1366 ***
                             (3.47)
5 minus 1         13.3915    0.1255 ***
(t-statistics)               (3.57)

*** Significant at the 0.01 level.

** Significant at the 0.05 level.

Table VIII. Determinants of Foreign Exchange Exposures to OITP and
Major Currencies

During each three-year estimation period, firms' monthly stock returns
are regressed on the percentage change of the real foreign exchange
rate index against currencies of "other important trading partners"
(OITP), the percentage change of the real foreign exchange rate index
against major currencies (Major), returns on the equally weighted CRSP
index, and returns on the one-month T-bill rate minus the monthly
inflation rates. We then separately regress the absolute value of the
estimated exposure to OITP and to major currencies on a firm's
beginning of period default probability, market-to-book ratio, capital
expenditures normalized by property, plant, and equipment, R&D
expense-assets, two variables measuring above average (FSales_Hi) and
below average (Fsales_Lo) foreign sales, the market leverage ratio,
the ratio of sales to market capitalization, the logarithm of firm
market value, and industry dummy variables, f-statistics for
coefficient estimates, calculated with robust standard errors adjusted
for both firm and year clustering, are reported in parentheses. Column
3 reports the differences in coefficients for the probability and the
costs of financial distress between the models for OITP and major
currencies, along with the [chi square]-test statistics in
parentheses. Number of observations and R-square are reported at the
bottom of the table.

                           OITP          Major       [chi square]-Test
                                                       of Difference

Intercept                0.9401 *       0.7164 ***
                        (1.80)         (5.35)
Default probability      4.0924 ***     1.1948 ***   2.8976 ***
                        (4.47)         (7.03)        (45.64)
M/B                      0.2771 ***     0.1141 ***   0.1630 ***
                        (3.17)         (5.09)        (18.09)
CapExp                   1.0513 ***     0.5758 ***   0.4755 **
                        (3.09)         (4.91)        (6.32)
R&D/Assets               4.0975 ***     1.6802 ***   2.4173 ***
                        (3.60)         (5.03)        (8.31)
FSales_Hi                0.6671 **      0.6061 ***
                        (2.03)         (7.04)
FSales_Lo               -1.6745        -1.3706 ***
                       (-1.36)        (-6.19)
Leverage                 1.0765 ***     0.5873 ***
                        (3.47)         (6.45)
SP                      -0.0210        -0.0343 ***
                       (-0.58)        (-3.39)
Ln(Size)                -0.1261 ***    -0.0892 ***
                       (-3.49)        (-7.23)
Industry dummy             Yes            Yes
N                         19,770         19,770
[R.sup.2]                0.097          0.102

*** Significant at the 0.01 level.

** Significant at the 0.05 level.

* Significant at the 0.10 level.
COPYRIGHT 2013 Financial Management Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

 
Article Details
Printer friendly Cite/link Email Feedback
Author:Wei, Kelsey D.; Starks, Laura T.
Publication:Financial Management
Geographic Code:1USA
Date:Dec 22, 2013
Words:14419
Previous Article:Competition, efficiency, and stability in banking.
Next Article:Capital structure and debt priority.
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters