مواجهه با نرخ ارز بدون قید و شرط و مشروط
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|8943||2013||28 صفحه PDF||سفارش دهید||15982 کلمه|
هزینه ترجمه مقاله بر اساس تعداد کلمات مقاله انگلیسی محاسبه می شود.
این مقاله شامل 15982 کلمه می باشد.
نسخه انگلیسی مقاله همین الان قابل دانلود است.
هزینه ترجمه مقاله توسط مترجمان با تجربه، طبق جدول زیر محاسبه می شود:
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of International Money and Finance, Volume 32, February 2013, Pages 781–808
We re-examine the relationship between exchange rate movements and firm value. We estimate the exchange rate exposure of U.S. firms to two currency indices. Firms are clustered into eleven industries. The sample includes exporters and non-exporters. Using a panel approach, we uncover statistically significant and sizable unconditional exposure. We also examine the dynamics of exchange rate exposure modeled as a function of business cycle indicators and firm characteristics. We find that exposure varies over time with macroeconomic and financial variables and increases during economic contractions. Deviations from the unconditional measure of exposure driven by the macroeconomic variables are economically meaningful.
Adler and Dumas (1984) define foreign exchange (FX) economic exposure as the sensitivity of the firm's returns to unexpected changes in real exchange rates. The extant literature finds a puzzling weak relationship between exchange rates and returns.2 There are several reasons that render identifying and estimating the FX exposure difficult. First, methodology matters in how exposure is measured.3 Second, exposure is temporally unstable.4 Third, exposure is measured net of operational and financial hedging.5 Ignoring any of these issues could understate the statistical significance or the economic importance of exposure. Our contribution is twofold. First, using a panel approach in an unconditional setup, we find evidence of statistically significant and economically important exposure for U.S. firms. Second, in a conditional setup, we show that currency exposure varies over time with financial business cycle indicators and macroeconomic variables and increases in periods of contractions. We measure industry level exposure and its dynamics using firm level data as opposed to industry indices. We cluster individual firms into industry panels and explicitly allow for heterogeneous exposure of firms within the same industry. Our approach is meaningful economically and statistically. Economically, an unexpected change in the exchange rate should affect industry's competitiveness but not equally affect each firm within the same industry.6 Statistically, the use of a panel model takes advantage of expanded observations to yield greater testing power and higher precision in estimation. It also overcomes the potential loss of information and bias induced by grouping firms. The results of the methodology augment the traditional firm-by-firm approach. They provide a manager with industry exposure benchmarks useful to better manage operations and possibly design more effective hedging strategies. We study the universe of US firms from COMPUSTAT over the period 1973–2005 using quarterly data. The sample includes firms with international involvement and purely domestic firms.7 The latter could be indirectly exposed through for instance import-competing. We measure exposure to two trade-weighted real currency indexes: the major index (MJ) and the other important trading partners (EM) index. In a static framework, we find that exposure is statistically significant and economically important. We first replicate the well known puzzling finding of a low proportion of firms significantly exposed as in Jorion (1990).8 However, panel regressions show that the results from the individual firms regressions do not imply that exposure is unimportant or insignificant. Taking into account the joint evidence from the cross-section of firms we find statistically significant and sizable unconditional exposure in most industries. We also provide a detailed statistical analysis of the exposures by introducing additional controls, examining exporters and non-exporters, and by looking at two sub-periods. We then relate our findings to some stylized facts and statistics about the industry trade balance of US industries by region. We uncover significant changes of the exposure over time. These results motivate the analysis of the dynamics of exposure that we discuss next. Importantly, the changes in exposures are overall consistent with the changes in the trade balance for some industries. In addition, we analyze the cross sectional determinants of exposure. Exporters and non-exporters show noticeable qualitative similarities both in their level of exposure and the determinants of the exposures. Our findings show that firms with higher international involvement, that are smaller, that are more levered, or that have lower growth opportunities, are also more exposed. Our second contribution is to relate the dynamics of exposure to the business cycle. Several studies examine time variation in exposure using subperiod dummies (see e.g. Williamson, 2001; Parsley and Popper, 2006), or using rolling regressions (see e.g. Glaum et al., 2000; Starks and Wei, 2006).9 A few other studies allow exposure to vary as a function of industry or firm level variables. Allayannis (1997) finds that the foreign exchange exposure of U.S. manufacturing industry varies with changes in the imports and exports. Gao (2000) discovers that exposure of US multinationals vary with firms' foreign sales and foreign production. Allayannis and Ihrig (2001) find that allowing exposure to vary with industry's markup, export share, and imported import share, helps uncovering exposure for U.S. manufacturing industries. Priestley and Ødegaard (2007) relate exposure dynamics to exchange rate regimes. Therefore, there is evidence that FX exposure changes over time. However, what drives the dynamics of exposure is still an open question. In particular, very little is known about how exposure varies over the business cycle. A seminal paper by Bodnar et al. (2002) shows that pass-through and exposure are related to the elasticity of demand.10 The authors assume a two country model with an exporting firm competing with a foreign-importing firm in the export market. They show that pass-through is incomplete because the demand functions permit price elasticities and hence markups to vary as prices change. They also show that exposure changes with markups which is consistent with the empirical evidence in Allayannis and Ihrig (2001). Furthermore, several studies argue for cyclical elasticity of demand and markups. Therefore, it is plausible that changes in elasticity of demand and markups can induce variation in exposure over the business cycle. We use panel regressions and a parametric specification to model the time variation in exposure over the business cycle. We set exposure as a function of financial business cycle indicators and firm-specific variables. As business cycle predictors, we use the default premium and the term premium along with the lagged world market return. The default premium captures the long run effect of business conditions, while the term premium reflects the short run effect (see e.g. Fama and French, 1989; Chow et al., 1997; Avramov, 2002). We also run extensive robustness tests using a wide array of macroeconomic variables. As for firm characteristics, we report results for financial leverage and liquidity. We also include other important variables such as foreign sales, size, profit margin, degree of operating leverage, and proxies for growth opportunities in the robustness section. We find that the exposure to the two currency indexes is significantly time-varying except the Chemicals and Telephone & TV industries. The time variation in the exposure to the two currency indexes is mainly driven by the financial business cycle predictors. Moreover, the exposure increases in periods of recession. Using the case of Non-Durables industry, we illustrate how the unconditional measure of exposure could significantly under or over-state the effect of exchange rate fluctuations on stock returns. The under or over estimation of exposure is economically meaningful both in relative and absolute terms. The robustness results confirm that the exposure's dynamics are mainly driven by the macroeconomic variables. Both the unconditional and conditional evidence indicate that exposure to the EM currency index affects US firms returns in an important way statistically and economically. This is interesting in view of the large currencies volatilities experienced by the emerging markets along with the difficulties to hedge against these currencies. The rest of the paper is organized as follows. Section 2 outlines the model and empirical methodology used in the study. Section 3 describes the data. Section 4 reports the empirical results. Section 5 concludes.
نتیجه گیری انگلیسی
We study industry level exposure to the major and the emerging market currency indexes using firm level data as opposed to industry indices. We measure exposure both in a static and in a conditional framework. Using firm-by-firm regressions we replicate the well known and puzzling result in the literature that only a relatively small number of firms show significant exposure coefficients. However, the Fama MacBeth t-tests of the estimated exposure coefficients suggest that the average of the exposure coefficients are significantly different from zero in most industries. Moreover, formal inferences based on a random coefficient panel approach provide strong evidence that the exposure to the two currency indexes, and particularly to the emerging market index, is statistically and economically significant. Our findings also highlight the important role that emerging markets play in the U.S. economy. We also provide a detailed statistical analysis of the exposures by introducing additional controls, by examining exporters and non-exporters separately, and by looking at two sub-periods. We then relate our findings to trade balances of US industries by region. We uncover two important facts. First, there is sensible time variation in exposures to both currency indices. Second, the results are overall consistent with the changes in the trade balance for some industries vis-à-vis their developed and emerging market partners. In addition, we analyze the cross sectional determinants of exposure. Exporters and non-exporters show noticeable qualitative similarities both in their level of exposure and the determinants of the exposures. Our findings are broadly consistent with the hedging argument. They show that the level of exposure is negatively related to size, and growth opportunities, while positively related to the degree of international involvement and leverage. We also examine the dynamics of exposure. We use a parametric specification to model the time variation in exposure. The advantage of such an approach is that it explicitly links time variation in exposure to macroeconomic state variables and firm characteristics. We use the default premium and the term premium for their ability to capture the macroeconomic conditions and report results about financial leverage and liquidity for firm characteristics. We also run extensive robustness tests using a wide array of both macroeconomic and firm variables. Our conditional results can be summarized as follows. The exposure to the two currency indexes is significantly time-varying except in Chemicals and Telephone & TV industries and the economic importance of exposure varies over time. The dynamics of exposure to the two currency indexes are mainly driven by business-cycle indicators and macroeconomic variables and are such that excess exposure increases during economic contractions. Using the case of Non-Durables industry, we illustrate how the unconditional measure of exposure could significantly under or over-state the effect of exchange rate fluctuations on stock returns. The under or over estimation of exposure is economically meaningful both in relative and absolute terms. The robustness results confirm that the exposure's dynamics are mainly driven by the macroeconomic variables.