英语论文网

Kenen and Rodrigues(1986) studied the effects on exchange rate uncertainty with respect to multilateral manufacturing imports for eleven industrialised countries and found a significantly negative effect in only four of them. On the other hand Bailey, Tavlas and Ulan(1987) found no significant effect of exchange rate uncertainty on multilateral exports of industrialised countries.

However most of the early works were inclusive about the relationship between exchange rate risk and trade and this could be attributed to a number of factors. First, the sample data used covered only a short period of exchange rate variability. Second, the equations used were rather basic and consisted only a few macro economic variables. According to the analysis of Mckenzie(1999) recent empirical studies have been more successful at generating statistically significant relationship between volatility and trade. This was due mainly to the emergence of more precise techniques of estimation and measurement of volatility and the use of more appropriate models. Kroner and Lastrapes (1993) and Arrize(1997a) argued that the mixed results obtained from early studies could be attributed to the use of inadequate measures of exchange rate risk. Unconditional measures of exchange rate volatility were mostly used, usually a moving standard deviation of levels or changes in past exchange rate. Recent studies use autoregressive conditional heteroskedasticity (ARCH) or generalised autoregressive conditional heteroskedasticity (GARCH) models instead. Other investigators like Caporale and Doroodian (1994) used a generalized autoregressive conditional heteroskedasticity (GARCH) technique to measure the volatility of exchange rate. They used monthly data for the period of 1974 -92 and found a significant negative effect of volatility on US imports from China. McKenzie and Brooks (1997) and McKenzie (1999) adopted ARCH modeling for measuring exchange rate volatility in their model for export trade for both German-US and Australian trade flows respectively. Even if their results were statistically significant they showed positive impact of volatility on trade, while for McKenzie (1999), the results were mixed.

Some recent studies adopted the co-integration analysis which takes into account trend characteristics of the time-series data. It was found that such studies tend to be more clear-cut and most of them yield a significant negative impact of exchange rate volatility on the trade variables. Examples of such studies include Koray and Lastrapes (1989), Arize (1997, 1998a and b), Fountas and Aristotelous (1999) and Flam and Jansson (2000). Fountas and Aristotelous (1999) used cointegration technique to assess the effect or exchange rate variability on exports among European Union countries and found a significant negative long run effect. Similarly, in a series of studies, Arize(1997, 1998a and 1998b) found a significant negative long run and short run effect. Koray and Lastrapes (1989) arrived at a significant negative relationship between exchange rate uncertainty and bilateral imports for five industrialised countries and a smaller and weaker negative short run relationship.

Finally, another set of empirical research employed the gravity model and has suggested significant evidence of a negative impact of exchange rate volatility on international trade. Basically, the gravity model express bilateral trade flows