مواجهه با نرخ ارز خارجی در شرکت های چند ملیتی ایالات متحده: یک رویکرد شرکت خاص

We examine the relationship between changes in foreign exchange rates and stock prices of US MNCs. Using firm-specific foreign exchange indices, we find more firms with significant exposure than when a common foreign exchange rate index is used as in comparable studies. We find that the number of firms found to have significant foreign exchange exposure, as well as whether or not particular regions of a firm’s geographic network structure is associated with any exposure, is dependent upon the type of foreign exchange rate index used to capture exposure. The findings in this study highlight the need for caution in the interpretation of previous studies of foreign exchange rate exposure.

Economic theory presumes changes in foreign exchange rates impact the stock prices of domestic as well as multinational corporations (MNCs). Changes in exchange rates drive changes in cash flows and ultimately the value of the firm. Shapiro (1975) models the theoretical effects of changes in exchange rates on the value of import and export firms alike. Export firms will gain from a devaluation of the host currency as their goods become more competitive while importers stand to lose. However, surprisingly there is little empirical evidence to support these theoretical predictions at the firm level. Previous studies have found little or no relation between changes in exchange rates and the value of the firm, leaving the question of why stock prices do not appear to be influenced by unexpected changes in exchange rates unanswered.2 In this study, we re-examine the influence of changes of foreign exchange rates on equity shares of US MNCs. However, where previous efforts measured foreign exchange exposure by using a common foreign exchange index, in this study we form and use a firm-specific foreign exchange rate index based on the structure of each company’s geographic network of foreign subsidiaries (FS). The empirical research on foreign exchange rate risk has produced mixed results. Jorion (1990) finds only little significance for individual firm exchange rate exposures of US MNCs over the period 1971–1987. Amihud (1994) finds no evidence of a significant exchange rate exposure for the 32 largest US exporting firms over the period 1982–1988. And Bodnar and Gentry (1993) find only 11 out of 39 industries with significant exchange rate exposures for the period 1979–1988. One key characteristic of these studies is the reliance on a common exchange rate index that is applied to all of the companies in the sample. Because the companies operate in different, distinct international locations, it is perhaps not surprising that the foreign exchange rate exposure variable is found to have little significance when a common index is applied. Consider the case of two firms with substantially different operational networks. Firm A has two foreign subsidiaries located in two different countries while firm B has 15 foreign subsidiaries in 15 different countries. It is unlikely that the same common foreign exchange rate index depicts the foreign currency environment of both firms. In this study, we use a firm-specific exchange rate proxy to examine whether the stock returns of US multinational corporations are influenced by changes in foreign exchange rates. The only other study we are aware of that implements such an approach is Ihrig (2001). Her study finds higher exposure for firms using firm-specific exchange rates. Here, we expand on this analysis by examining the relationship between a firm’s operational network and its exposure. For example, we examine whether a firm’s operations in Central America are associated with a firm’s exposure in the same manner as its operations in the region defined by the North American Free Trade Agreement (NAFTA). This study contributes to the field’s better understanding of exchange rate risk exposure in several ways. First, we provide complimentary evidence on the nature of foreign exchange rate exposure through the use of firm-specific measures. Secondly, when we compare firm-specific results to previous approaches, we show that not only is the magnitude of exposure dependent on the selection of a foreign exchange index, but so too is the sign of that exposure. This finding has direct implications on a firm’s potential hedging and capital expenditure decisions.3 Finally, our findings from examining the determinants of the foreign exchange rate exposure using different firm-specific and common foreign exchange indices indicates that we must use caution when interpreting the findings of existing studies. The remainder of the paper is organized as follows: Section 2 discusses the current literature studying foreign exchange rate exposure. Section 3 outlines the methodology and describes the sample, Section 4 reports the results, and Section 5 concludes. 2. Issues in foreign exchange rate exposure 2.1. Measuring foreign exchange rate exposure Much of the research in exchange rate exposure employs variations of a model in which an individual stock return is regressed on a foreign exchange rate variable. In each case, the coefficient on the foreign exchange variable is interpreted as the firm’s exchange rate exposure, see for example Dumas (1978), Hodder (1982), and Adler and Dumas (1984). Adler and Dumas (1984) illustrate that exposure to exchange rate risk is similar to that of market risk measured in the traditional sense.4 Therefore, the average exposure to exchange risk can be obtained by regressing the stock returns on the changes in the prices of currencies (exchange rates). While this regression does not imply any causal relationship between foreign exchange rates and stock prices, following Jorion (1990) and He and Ng (1998) we treat changes in the prices of foreign exchange as exogenous to individual stock returns.

We introduced this study with the simple question as to whether or not the stock returns of US multinational corporations are influenced by changes in foreign exchange rates if we use a more representative, firm-specific exchange rate proxy than the traditional common index variable approach used by many studies. The short answer is yes. Consistent with Ihrig (2001), we find a higher number of firms with significant exposure than when a comparable common index is used. 8.7% (27 out of 310 firms) are found to have significant exposure with a firm-specific index versus 5.5% (17 out of 311 firms) with the common MAJCUR index. However, neither of these measures results in the level of significance when the FRB’s BROAD index is used (12.6% or 39 of the 310 firms). Furthermore, we find evidence that the percentage of a firm’s foreign subsidiaries is positively related to exposure when exposure is measured with a firm-specific index and negatively related to exposure when a common index is used. While the levels of significance appear to increase with firm-specific measures, the results are not drastically different than those found in previous research. It is interesting to note that we examined 310 firms using four different foreign exchange measures and only three firms, or less than 1% of the sample, were found to have significant exposure across all measures. The longer answer to our question involves interpreting the stark differences in results of studying the determinants of exposure depending on which measure of foreign exchange (common or firm-specific) is used. For instance, we find that the number of foreign subsidiaries is associated with reduced exposure when a firm-specific measure is used while the opposite is true when the common BROAD index is used. Similarly, when we examine the relationship between geographic network variables and exposure, the results are equally puzzling. Using exposures derived from models relying on firm-specific measures of foreign exchange rates, our results suggest that a firm operating in Central America is more exposed and one that operates in the Asian crises region is less exposed. Conversely, using exposures derived from models relying on common foreign exchange indices, the results suggest that firms operating in the Asian crisis region are more exposed. The primary difference between any of the indices is the specific currencies included and their respective weights. Certainly, the inclusion (or exclusion) of individual currencies that compose the indices is a factor in the determinants of exposure studied here. When more currencies are included in an index, whether or not the firm even operates in a particular currency’s home country, we obtain more firms with significant exposure. This constitutes indirect evidence in support of the notion that all firms may be subject to foreign exchange exposure, not just those operating in foreign countries. Alternatively, this may indicate that firms are exposed to currencies of countries where they are not operating since, for example, some of their competitors may be incorporated in these countries, or because some of their inputs to production may be denominated in those currencies. The impact of these findings is significant for researchers and managers alike. First, if managers utilize such measures to determine the need and degree of hedging activity, our findings imply the selection of the foreign exchange rate measure significantly impacts the resulting exposure to the extent it may even suggest different currencies or regions to hedge. For researchers, our results suggest we should be cautious in our interpretation of pervious studies. From the analysis of the determinants of exposure studied here, it is not clear that the operation of a network of foreign subsidiaries increases or decreases exposure. Clearly, the selection of a foreign exchange variable is critical in any analysis of foreign exchange exposure, and more so in the generalization of results across all firms.