برخی از دیدگاه ها در پازل قیمت گذاری ارز خارجی : شواهدی از یک اقتصاد کوچک باز

One of the great puzzles of international finance research has been the surprising finding that firm value is only mildly sensitive, if at all, to exchange rate fluctuations. This runs against the conclusions of the standard net discounted cash flow theory. In this article we examine this puzzle, and provide some insights by looking at whether the result is peculiar to large economies like the US, using improved methodology including a residual regression model, the use of individual firm data, and looking at the effect of different exchange rates. We do find clear evidence that exchange rate movements affect the value of listed firms. We also find that the direction and degree of sensitivity is dependant on the currency used.

The effect of exchange rate movements on the value of a firm has become an important field in both academic research and practical investment analysis. Academic researchers and industry investors are faced with important questions to answer: (i) does the volatility of exchange rates affect firm value and (ii) is this exchange rate effect reflected in stock prices? Then, if the answers are yes, the next two questions are: (iii) how sensitive is the value of firms to exchange rate movement and (iv) what factors are important in determining the degree of this sensitivity? Conventional wisdom and economic theory suggest a positive answer to the first question—changes in the relative prices of domestic and foreign goods are widely believed to influence the current and future expected cash flows of the firms with foreign sales and operations abroad. This should change the NPV of the firm, and thus the market capitalisation of the firm will react to movements in exchange rates. Even firms that are mainly domestic in nature may also be affected by exchange rates as their input and output price linkages, or their supply and demand chains, or their competitors' prices, might be influenced by currency movements. The efficient market hypothesis would also argue that this information should be priced into share prices immediately. The size of the reaction will be complex, as it will depend on the various determinants of the NPV, the diversification of the firm's cash flows and assets/liabilities, and the expected duration of the change in the exchange rate. However, the effect should be significant. Note that only unexpected changes in the exchange rate will affect share prices, as expected changes will already be priced in. There is a two-step linkage here: (i) changes in exchange rates should affect firm value and (ii) and investors should recognize this information and price it into shares. A failure to find a relationship between exchange rates and share prices could be caused by the break down of either link. We would argue that, given the unusually exposed nature of the New Zealand (NZ) economy, both these links, in theory, should be stronger in NZ than in the US. The puzzle in this area arises because previous empirical studies have shown only mixed results with regard to the impact of unexpected changes in foreign exchange rate on firm value. Findings vary depending on the firm sample selection and models employed. However, the general research conclusion is that firm value is only mildly sensitive to exchange rate fluctuations, if at all. There have been a limited number of research papers that have documented a small but significant correlation between exchange rate movements and changes in firm value. Choi and Prasad (1995), Booth and Rotenberg (1990) and Frennberg, (1994) did find significant sensitivity. However, the mean values of that sensitivity were far lower than they should be in theory. Conversely Jorion (1990), Amihud (1994), Gao (2000), Doidge et al. (2000), Di Iorio and Faff (2002) and Griffin and Stulz (2001) found only weak or insignificant sensitivity of firm value to exchange rate exposure, Bartov and Bodnar (1994) found only a lagged relation, implying a degree of market inefficiency, and Allayannis (1997), Chow et al. (1997) and Bodnar and Wong (2000) found only exposure over a long horizon. Jorion (1990) and Shin and Soenen (1999) also found that hedging activities exhibit economies of scale and, consequently, the magnitude of exchange rate exposure is less for large firms than for small firms. The failure to find the expected theoretical relationship has normally been explained in two ways: (i) there were methodological issues, or (ii) there is a basic theoretical failure of the fundamental cash flow model to match reality. We, however, would argue that it is also possible that the relative lack of sensitivity found in previous studies is a peculiarity of large markets. Given that the US stock market is large, with domestic factors playing a predominant role in the determination of both share and consumer goods prices, it is possible that investors are not very focused on changes in international factors. There also tends to be a high degree of price stickiness in the goods market, and delayed exchange rate pass through (Bodnar et al., 2002). US multinationals also tend to be fairly diversified, so that it could be difficult to find causal relationships between firm value and any one currency. There is thus a serious data selection problem in US and UK studies, as only firms that have a relatively high degree of foreign influence can be used. Even with those firms, only their net cash flows will be affected by exchange rate movement. These characteristics will substantially cut the degree of exchange rate sensitivity found from any regression analysis. In contrast for small open economies, like NZ, all company cash flows are heavily influenced by international factors, and companies tend to be less diversified. It is thus likely that there is a significant structural difference between the US, and other large markets, and the markets of small open economies. This study using NZ market data to examine foreign exchange exposure is thus unique and makes a special contribution to the literature. While NZ is often regarded a typical small and open economy, it is unusually open to the international economy. Its exchange rate has been volatile ever since the NZ dollar (NZD) was floated in 1985. NZ operates its exchange rate under a clean float, with the Reserve Bank rarely intervening in the foreign exchange market. The Reserve Bank monitors the dollar value through the trade-weighted index (TWI), which weights the currencies of NZ's five main trading partners according to their trade value associated with these countries. The TWI has been volatile since floating (see Fig. 1). This volatility has many causes, but one of the main causes is the small size of the currency market compared to international markets. Full-size image (12 K) Fig. 1. NZ dollar exchange rates 1993–2000. Figure options NZ is thus very useful as a case study for a number of reasons; the main one being that the NZ economy is more vulnerable to international influences than many other countries. NZ firms are also not world market leaders, especially in manufacturing, thus they are normally foreign currency denominated price-takers. Even the price of domestic sales tend to closely follow the world price, as there is high exchange rate pass through, apart from a 6/12-month hedge in some areas. Thus currency depreciation leads to an increase in sales for both exporters and domestic sellers. NZ firms also tend to be less diversified, with a predominantly export or import orientation, with few natural hedges, and a tendency to be highly exposed to single currencies. These factors will mean that there will be a far weaker data selection problem in NZ, as (a) it does not matter as much what proportion of sales/costs are foreign, as all cash flows will tend to follow changes in the exchange rate, and (b) company foreign exposures tend to be predominately related to one currency. The Kiwi dollar has experienced cyclical volatility since its float, which has often had a major impact on firm profitability. Given the high impact that this has had on the mostly small or middle size of firms in NZ, we would thus expect a higher level of market awareness of the impact of foreign exchange rate movements on firm value among investors, and a high level of market information efficiency. Thus stock prices can be expected to be more closely tied, than that of other countries, to exchange rate movements. Our hypothesis is thus that firm values have been significantly affected by the fluctuations of the Kiwi dollar value, or, in other words, the exchange rate risk has been adequately priced into the NZ stock market, unlike the case for the larger markets. Our mean degree of sensitivity should also be higher than comparable US or UK studies. The paper is organized in the following order. Section 2 reviews relative literatures. Section 3 describes the methodology used. The data analysis and test results are provided in Section 4, and Section 5 summarizes the conclusion.

Based on our two-factor model of firm valuation, this research examined the sensitivity of firm value to foreign exchange movements, using a sample of 161 NZSE listed firms. Comparing with most of the prior studies, this research investigated the firm value sensitivity to exchange rate fluctuation by focusing mainly on individual firms, instead of forming industry or market portfolio, (which often causes under-estimation of existing FX sensitivity of firm value due to risk reduction effect of portfolio formation), and also looked at the differing rate of sensitivity between currencies. The estimation results of this research report are encouraging—we do find strong evidence that exchange rate movements affect the value of the listed NZ firms. More specifically, the regression results show that when measured in semiannual return more than 27% of the total sample firms have statistically significant exchange rate risk exposure and about 85% of these significant firms are positively related. We also find significantly higher mean sensitivity. A special feature of NZ market performance is that firms are on average positively related to the movement of the US dollar and negatively related to the movement of the Australian dollar, which means that firms, on average, gain in value when the NZ dollar appreciates against the US dollar and depreciates against the Australian dollar. This, which was hidden by the TWI results, shows that it is important for research to look at a range of currencies when looking at foreign exchange pricing. If a company or a sample is very diversified across currencies, then the results for any one currency may be quite different from the overall result. This research also found that, consistent with prior research, extending return horizons has significant effects on the correlation between firm value and exchange rate fluctuations. The statistical significance of exchange rate exposure of the firm value does increase with the length of the return horizon in this research. The most sensitive horizon is in semiannual return rates (TWI and USD) or quarterly return rates (AUD). Results were more useful when we divided our sample firms into three groups. We found that the difference in performance relative to US dollar and Australian dollars happened most sharply in the energy and resource industry. These firms were using local resource and exporting part of their products to overseas markets. Their significant positive percentages in most cases are more than double the negative percentages. The USD rate negative percentages are in the range of 40–70%, more than triple the corresponding positive percentages. The result indicates that researchers have to be very careful about the particular characteristics of the firms being analyzed. Thus we may have confirmed our hypothesis that small open markets have a higher level of pricing efficiency with regard to foreign exchange changes than the US market. We have also confirmed that the low levels of sensitivity of firm value to changes in exchange rates that was found in prior US and UK studies may have been a peculiarity of the data problems that exist in those markets, due to the low level of exchange rate pass though. The fundamental cash flow model of firm value is thus well supported. We have also correspondingly found a higher level of market efficiency with regard to exchange rate information in smaller markets. This is further confirmed as cross-sectional regression reveals that FX exposure variation of individual firms only shows a very weak negative relationship to firm's foreign operations, foreign sales revenue and foreign assets. Thus study suggests that the immediate focus of further research in the larger economies could be on overcoming the data selection issues, as the basic theory that firm value will respond to foreign exchange changes is secure. It is important to note that our results only looked at linear exposures, as this enabled comparison to past research. We believe that including nonlinear relationships, as done recently by Bartram (2002) and Priestley and Ødegaard (2002), would only strengthen these results.