Exchange Rates and the Role of the Trade Balance Account

Abstract

The essence of the modern asset‐market approach to the analysis of exchange rate behavior includes the role of the trade balance account. We examine the relationship between exchange rate changes and US trade balance announcements. Statistically significant exchange rate adjustments to these announcements are documented using for the first time the comparison period approach to testing the significance of trade balance announcements on exchange rates. The evidence is consistent with the predictions of the modern asset‐market exchange rate model. There is also evidence that the foreign exchange market is more sensitive to increasing rather than decreasing trade balance deficit announcements. To date, a number of theoretical papers have investigated the possible sources of the exchange rate determination process (see, Dornbusch [1976,1980], Dornbusch and Fisher [1980], Frenkel [1976, 1981], Kouri [1976], and Mussa [1982], among others). There is no consensus on how exchange rates are determined and why they have exhibited increased volatility lately. The interpretations vary widely among the various theories, ranging from the flow‐market approach to the modern asset‐market view. The asset‐market approach of exchange rates is based on the principle that the current value of the exchange rate (i.e. the relative price of two national currencies) is influenced not only by current economic conditions but also by expectations of its future value and, therefore, by the information that underlies these expectations. The asset‐market literature on the determination of exchange rates establishes a direct relationship between changes in the exchange rate and the current account (or trade balance account). For example, Mussa [1982] shows that the equilibrium exchange rate depends on expectations about the exogenous factors that affect the current account in present and future periods. A central implication of the asset‐market view is that “innovations” in the current account induce unexpected changes in the exchange rate. This is because an innovation in the current account, defined as a deviation of the current account balance from its previously expected level, conveys information about changes in economic conditions relevant for determining the equilibrium exchange rate (see Mussa [1982]). For example, if a country experiences an unexpectedly strong trade balance performance, this might be perceived to imply changes in relative economic efficiency, product demand, or international competitiveness that will improve the current account in future periods leading to an appreciation of the foreign value of the domestic currency. In essence, the asset‐market view argues that information about changes in real economic conditions requiring exchange rate adjustments can be inferred from innovations in the trade balance and/or the current account. Dornbusch and Fischer [1980] also argue that while asset markets determine exchange rates, it is the current account through its effect on net asset positions, and subsequently on asset markets, which influences the path of the foreign exchange rate. Thus, it can be argued that unanticipated current account announcements should be associated with exchange rate movements immediately following such announcements. While the relationship between the current account and the exchange rate has been extensively analyzed, the empirical evidence pertaining to the association between exchange rates and the current account has produced mixed results. Hardouvelis [1988] examines the effects of macroeconomic news, including US trade balance announcements, on three interest rates and seven exchange rates over the October 1979 to August 1984 period. He reports that announcements of the trade deficit have no statistically significant effects on interest rates, with the exception of the three‐month T‐bill rates and the exchange rates. The evidence with respect to the short‐term interest rate reactions may be associated with the fact that the “Federal Reserve Bank throughout the 1977–1984 period was unable to establish full credibility among market participants about its fight against inflation” (see Hardouvelis [1988]). Deravi et al [1988] have also investigated the financial market's response to US balance of trade announcements. They find similar results to those reported in Hardouvelis [1988] for the February 1980 to February 1985 period, but they report a significant exchange rate response to trade deficit announcements over the March 1985 to July 1987 period. Irwin [1989], however, uncovered a significant breakdown in the relationship between trade balance announcements and dollar exchange rates during the month of June 1984; that is, larger trade deficits were found to be associated with the dollar's depreciations only in the post‐June 1984 period. Contrary to previous studies, Hogan et al [1991] find larger US trade balance deficits to have a significant effect on exchange rates throughout the 1980s. Because expected trade balance figures are available from the Money Market Service Inc. and since the trade balance figures according to Crystal and Wood [1980] represent 85 percent of the US current account, it apears that the trade balance serves as a good proxy for the current account. Therefore, we are able to test more directly the impact of the US trade balance announcements on the exchange rate. The purpose of this paper is to analyze the relationship between exchange rate changes and merchandise balance announcements using a sample of US trade figures spanning the period from August 1986 to April 1989. In the following, we refer to this relationship as the “current account hypothesis”. Unlike previous research, the analysis is based on unanticipated trade balance announcements in order to study the interaction between exchange rates and information contained in the trade balance announced figures as the asset‐market approach to exchange rate determination process predicts. Dornbusch [1980] used the official forecast errors of the Organization for Economic Co‐operation and Development (i.e. biannual six‐month forecasts for current account and exchange rates). In this study, we focus on the major component of the current account‐the trade balance‐to test the current account hypothesis. The trade balance account is by far the best proxy for the current account. Another differentiating aspect of this study from the previous research is that it relies on systematic trade balance announcements. The use of the Commerce Departments' announcements concerning the US merchandise trade balance has also been motivated by the growing financial and non‐financial press coverage of the monthly trade balance reports. Examples of how the financial press covers the monthly trade balance announcements include: 1. “A wider trade deficit jolts a fragile market, shares off 101 points, dollar falls, and interest rates surge as big gap surprises investors, central bankers”, The Wall Street Journal, April 5,1988. 2. “London stocks rise sharply on US trade news; shares close firmer in Tokyo for the second day”, The Wall Street Journal, May 18,1989. 3. “Tricks of the Trade. The huge current‐account imbalances of the 1980s are disappearing fast. Good news? Maybe. But be warned: trade flows are less and less useful as indicators of economic performance” The Economist, March 30, 1991. 4. “Trade deficit grew in April to $6.97 billion… as exports continued to drop and imports jumped. The April deficit was the biggest monthly imbalance since a $9.49 billion deficit in November 1990. The trade gap in March was $5.58 billion. Economists say sluggish economic activity abroad is making it more difficult for US companies to sell their goods.” The Wall Street Journal, June 19, 1992. The different views registered in the financial press about the importance of the current account and trade balance imbalances in influencing exchange rate changes have further motivated the present study. Contrary to the current account hypothesis, it has been argued that because of the increasing integration of world capital markets, it is easier to finance current account deficits and therefore the trade balance or current account figures might be less useful as far as the determination of exchange rates is concerned. In addition, as a result of the increasing foreign investment activity, trade deficits may no longer represent purely national concepts. For example, a significant portion of a country's exports and imports may be accounted for by foreign firms with corporate operations there. Furthermore, US firms may decide to supply an overseas market either by exporting or by locating production abroad. Locally produced sales by US firms overseas, however, do not count as exports, nor do their local purchases of inputs count as imports. But from the firm's point of view, the local sales of a US subsidiary are viewed as being similar to exports. Therefore, it is argued that US trade balance deficits measured on the basis of residency rather than nationality of ownership, which is currently the norm, may mean less than it once did. Consequently, what emerges from the above is that the correlation between exchange rates and the information contained in the trade balance figures may be weaker than predicted by the asset‐market approach. Whether the current account or trade balance figures do matter as far as the determination of exchange rates is concerned is an empirical question. This article presents a first attempt at analyzing the impact of “innovations” in the US trade balance account on the exchange rate. An event study analysis is performed for the first time using trade balance announcement data from August 1986 to April 1989. The event methodology provides an appropriate direct test for the asset‐market model which predicts that unexpected changes in the exchange rate should be related to innovations in the current account (trade balance). The article is arranged as follows. Section II describes the data and methodology used. Section III presents empirical evidence on the relationship between exchange rates and innovations in the trade balance account. The article concludes with Section IV.

Citation

Doukas, J. and Lifland, S. (1994), "Exchange Rates and the Role of the Trade Balance Account", Managerial Finance, Vol. 20 No. 5, pp. 67-78. https://doi.org/10.1108/eb018476