The foregoing analysis has revealed relatively little in the way of directly measureable adverse effects of exchange rate variability on trade. This might seem surprising in view of the widely accepted proposition that uncertainty is bad for economic activity, and the obvious fact that, for businessmen, exchange rate variability is a source of uncertainty for which they must allow in their business decisions.
Part of the explanation for the lack of positive findings may lie in the inadequacy of the statistical techniques employed. There is always room to refine the empirical measures of exchange rate variability, and the lags with which changes in the economic environment affect behavioral relationships may be too long and variable to be easily captured by standard analytical techniques. Nevertheless, given the wide variety of empirical testing that has been performed, it seems unlikely that, as far as data up until the early 1980s are concerned, more intensive or sophisticated tests would show a greatly different result.
The reason for this probably lies in the difficulty of trying to separate the independent effect of exchange rate variability from the impact of other changes in the economic environment (Williamson, 1983). Moreover, as McCulloch (1983) points out, of the large number of risks involved in commercial activity, exchange rate variability may be a relatively minor one. These other risks arise not only from the microeconomic uncertainties of business activity but also from uncertainties concerning the level and variability of inflation, the amplitude of the trade cycle, the need for balance of payments adjustment, and so on. These factors can have direct and substantial effects on the growth of world trade, and also affect exchange rate relationships. This impact on exchange rate relationships may magnify or offset the direct impact on trade flows, but in any event it will not be related in a simple manner to any straightforward measure of exchange rate variability. In principle, it would be desirable to measure that part of exchange rate variability that was independent of, and in addition to, the uncertainty already present as a result of the volatility of other variables. No satisfactory way of making such a measure has yet been proposed.
In this connection, it is worth recalling that the exchange rate is a price, and like any other price it appears as exogenous to any individual market participant but endogenous to the aggregated system of market demands and supplies. A stable price reflects stability in the underlying system, and an unstable price reflects instability in supply and demand. It should be recognized that, although varying, the price is nevertheless fulfilling a stabilizing function, equalizing supply and demand. If the price were somehow prevented from moving, it would be unable to fulfill the function of calling forth additional supply and restraining demand (or vice versa), and there would be a disequilibrium between supply and demand and an inappropriate level of transactions. Translated to the exchange market, the implications of this are that exchange rate variability is a symptom of unexpected shifts in demand relative to supply, and an absorber of part of the impact of such shifts (Mussa, 1982). It is not in itself independent of the transactions that are influenced by it.
The foregoing argument suggests the response that if the exchange rate system is supposed to promote trade between countries but actually discourages useful exchange because of uncertainty, should there not be an alternative way of ensuring continuous equilibrium between supply and demand? Furthermore, when exchange rate movements are reversible, and when they are clearly inconsistent with current account equilibrium, would it not be better to smooth such movement rather than permit temporary false price signals to distort the allocation of resources?
These questions approach, and in some respects step outside, the limitations that have been placed on the scope of this paper. Nevertheless, it is worth noting that, while the case for smoothing exchange rate fluctuations is strong in principle, there are considerable complications in practice. These complications are numerous, but can be grouped under two main headings:
(1) The difficulty of determining whether a given exchange rate movement represents a shortly to be reversed shift or is part of a movement to a new equilibrium.
(2) The fact that exchange rates determine, and respond to, capital account developments as well as to factors that influence the current account.
In practical terms, it is not possible to identify within narrow bounds what exchange rate would secure a current account position of a given size. This reflects in part weakness of data and estimating procedures but more important the dynamic evolution of economic structures in different countries, unexpected exogenous disturbances, and more or less foreseen economic policy reactions. When exchange rates move in response to market forces, it is even more difficult to say that they are “wrong” in a longer-term sense.
Sometimes, of course, it may be possible to say with greater confidence that an exchange rate has proceeded beyond the point that would produce a given balance on current account. In this connection, it should be remembered that the function of international exchange is not just to exploit comparative advantage under the constraint of a given current account position but at the same time to permit shifts of real resources among countries in response to changing savings and investment preferences. As savings and investment flows change, reflecting shifting opportunities and preferences, a movement in the exchange rate is needed to call forth a counter-balancing shift in the current account. Savings and investment flows can change for a wide variety of reasons, some of a fundamental character and others of shorter-term significance.
One of the most important insights of recent theoretical developments in international economics is the recognition that the exchange rate is a price that balances the willingness to hold claims (or maintain liabilities) in different currencies, as well as the desire to exchange one currency for another in order to effect trade in goods. Since the stock of assets denominated in different currencies is large relative to the flow of transactions over exchange markets, a change in the attractiveness of assets denominated in one currency relative to another can have significant short-term effects on the desire to acquire or dispose of foreign exchange.
In a world of integrated capital markets, interest rate differentials could give rise to sizable flows of funds across the exchanges, unless there were some mechanism to dampen this effect. One such mechanism is the freedom of exchange rates to vary in response to autonomous shifts in supply and demand. An increase in the desire to hold, say, U.S. dollars, results in the foreign exchange price of the dollar being bid up to a point where asset holders are just willing to hold the available stock of assets at the new price. If the price of the dollar were somehow prevented from moving, there would be a large excess demand for dollars which, if the authorities accommodated it, would generate a big increase in the money supply.
The upshot of this discussion is that it is not really meaningful to pose the question what would be the effect on trade if exchange rates were stable and all other conditions remained the same. The determination of exchange rates is part of a complex policy nexus and cannot be separated from the effect of other influences that interact with exchange rates. They have as well a separate and independent effect on economic activity and on trade flows.