بررسی مدل سه عاملی مواجهه با نرخ ارز خارجی

We investigate the likelihood of extreme foreign exchange-rate exposure (FXE), conditioning upon key firm factors and an expanded view of hedging. Our investigation incorporates the Fama and French (1993) three-factor (FF three-factor) model terms in reconciling equity returns vis-à-vis exchange-rate exposure. Our results suggest the following conclusions. First, consistent with effective hedging, non-hedging firms tend to have greater FXE than hedging firms. Second, all key factors that explain the likelihood of high FXE are economically and statistically significant using the more complete FF three-factor model. Third, we note that firm size is important in explaining FXE. Fourth, we find more FXE coefficients that are significant using the FF three-factor model compared to the traditional market model

Although Shapiro (1975) suggests that firm value is sensitive to changes in currency exchange rates, the empirical evidence supporting this theory varies (e.g., Vygodina, 2006). Researchers have relied primarily on returns-based estimates of foreign exchange-rate exposure (FXE), using a market-model approach (e.g., Adler and Dumas, 1984, Jorion, 1990, Choi and Prasad, 1995, Miller and Reuer, 1998, Ihrig, 2001, Bodnar and Wong, 2003 and Dominguez and Tesar, 2006). The extant evidence suggests that a firm's FXE depends in large part on its size, foreign sales ratio, and the effective use of hedging. In this study, we focus on the role of these 3 key factors in understanding the likelihood that a firm will face extreme levels of FXE, as measured by the Fama–French three-factor model (Fama & French, 1993). Recent research in the asset-pricing literature indicates that the Fama–French three-factor (FF three-factor) model often outperforms the traditional market model (e.g., Lawrence et al., 2007, Hung, 2008 and Simpson and Ramchander, 2008). To date, however, studies have not integrated this asset-pricing model into returns-based estimates of FXE. We augment the traditional market-model approach by including the Fama and French book-to-market and size factors. Size is of particular interest to understanding the likelihood that a firm will have FXE. Using returns-based estimates of FXE, prior research has provided mixed evidence in this area. Bodnar and Wong, 2003 and He and Ng, 1998 show that large firms have more exchange-rate exposure than small firms. In contrast, Dukas, Fatemi, and Tavakkol (1996) report an inverse relationship between firm size and FXE. Dominguez and Tesar (2006) suggest that exchange-rate exposure varies little with size. Faff and Marshall (2005) review 3 specific global regions and conclude that size is not universally consistent. In all these studies, the estimation of FXE does not control for the relationship between size and returns. Our paper makes a notable contribution to the current literature by merging the asset-pricing literature with the FXE literature. We investigate the likelihood that firms will face extreme levels of FXE, conditioning upon the key firm factors identified above and an expanded view of hedging instruments, and incorporating the FF three-factor model terms in reconciling equity returns vis-à-vis exchange-rate exposure of multinational corporations (MNCs). To date, no study has investigated exchange-rate exposure in the context of all these conditions. Utilizing the FF three-factor model produces more FXE coefficients that are statistically significant, compared to the traditional market-model approach, which does not control for the relationship between firm size and returns. A logit regression analysis, based on the absolute values of FXE estimates using the FF three-factor model, indicates that small firms with higher foreign sales that do not use financial hedges will have greater exchange-rate exposure. Further analysis of the logit probabilities suggests that firm size remains a primary determinant of FXE despite the use of the FF three-factor model. Finally, consistent with effective hedging, we find that the use of financial hedging is important in understanding the likelihood of extreme FXE. The remainder of the paper is organized as follows. Section 2 presents the data and methodology used to estimate and analyze FXE. The results are detailed in 3 and 4 concludes the paper.

This paper contributes to the current literature by merging the asset-pricing literature with the exchange risk literature. We investigate the likelihood that firms will face extreme levels of foreign exchange-rate exposure (FXE) by conditioning upon firm size (SIZE) and the foreign sales ratio (FSTS), using an expanded view of hedging instruments (HEDGE), and incorporating the Fama–French three-factor (FF three-factor) model terms in reconciling equity returns vis-à-vis exchange-rate exposure of multinationals (MNCs). To date, no study has investigated exchange-rate exposure in the context of all these conditions. We sample 185 US MNCs with ex ante exposure to changes in exchange rates. Utilizing the FF three-factor model produces more FXE coefficients that are statistically significant, compared to the traditional market-model (FX market model) approach introduced by Adler and Dumas (1984). Logit analysis using the FF three-factor model suggests that small firms with higher FSTS that do not use financial hedges have significantly greater absolute levels of FXE (|FXE|) than predicted by the FX market model. Focusing on firms with the greatest likelihood of high |FXE|, our results suggest that firm size remains important in such analyses. Finally, the use of financial hedging affects the likelihood of firms facing high |FXE| for our sample, which is consistent with effective hedging. Our results suggest the following conclusions. First, non-hedging firms tend to have greater FXE than hedging firms. Second, all key factors that explain the likelihood of high |FXE| (i.e., SIZE, FSTS, HEDGE) are economically and statistically significant using the more complete FF three-factor model. Third, we note size is important in explaining FXE, despite the SMB variable in the three-factor model. This result suggests that, in addition to the well-documented relationship between size and returns, there is a relationship between size and FXE. Fourth, we find more FXE coefficients that are statistically significant using the FF three-factor model compared to the traditional FX market model.