Psychological and institutional forces and the determination of exchange rates.

Author:Harvey, John T.

Neoclassical economists have, by their own admission, had a terrible time explaining foreign-currency prices (Taylor 1995a,b). Not only have they been unable to develop models that explain the past time series of foreign exchange rates with any regularity but their premises lead to conclusions about the character of the market that are inconsistent with many of its salient features. Mainstream approaches cannot, for example, account for exchange rate volatility, nonrational expectations, or the popularity of technical analysis.

It is my contention that neoclassicists' failure can be traced in large part to their assumption that economic behavior is independent of social, psychological, and cultural influences and is instead driven by rational and presumably natural free market impulses. But our economic system emerged from the unique historical circumstances of eighteenth and nineteenth century Western Europe. It did not always exist, nor was it inevitable--and it would be arrogant of us to assume that history has arrived at its end point. Understanding behavior within our system requires that we understand the behavioral norms imposed on agents, norms that are specific and evolving.

The purpose of this paper is to show that an explanation of exchange rate determination that places the activity in its psychosocial context yields results superior to those of one based on traditional neoclassical principles. In particular, it will be shown that if agents' behavior is modeled as being a function of their socialization in the capitalist system (and subsequent molding within the subculture of portfolio investors) and if their forecasting and decision-making processes are assumed to be driven by the same factors as theorized by psychologists (rather than economists), then the picture that emerges is one wherein bandwagon effects are common, periods of volatility are to be expected, and predictions can contain persistent bias.

This paper is distinct from earlier institutionalist works on foreign exchange (including my own) in a number of ways. It is, to my knowledge, the first to place currency trading in a cultural context and to show how the latter has nontrivial consequences. Second, it shows not only how asset markets can generate volatility (not something by itself that would come as a surprise to readers of the JEI) but also that the volatility is self-limiting. Third, bandwagon effects are explained as a function of the psychosocial features of trading. Last, the popularity of technical analysis is shown to be a function of more than just self-fulfilling prophecy.

The paper is arranged as follows. First, a brief review of the mainstream approach is offered (along with its weaknesses in terms of explaining the salient features of the foreign currency market). Second, the first piece of the psychosocial approach is put into place by defining the various roles played by the participants in the market. Third, their relative significance in the determination of prices will be discussed. Next, and most important, the psychological and institutional forces molding the behavior of the most influential actors will be explained. Finally, the implications for foreign exchange prices are outlined.

Neoclassical Explanations of Currency Markets: a Brief Overview

On the macro level, neoclassical approaches to exchange rate determination typically assume long-run money neutrality and a natural tendency for the market to settle at full employment. These are related in that if the level of output and employment is guaranteed by the efficient workings of the market, then money and finance have no role to play outside of accommodation. For foreign currency markets these premises translate into the conclusion that it must be trade flows, rather than capital, that dominate price determination. Capital flows (like money and finance) do no more than provide the funds for the dominant activity (even though they may superficially appear to be more important). Furthermore, it is believed that there is a strong tendency for imports and exports to equalize, a conclusion that follows naturally if the primary demand for currency is derived from the desire to undertake these activities. (1) In the long run currency prices are driven by international demand for goods and services in a way that guarantees balanced trade.

At the micro level, agents are thought to be rational and efficient. Their forecasts are unbiased, and all available information is taken into account. Social and cultural factors are unimportant since economic behavior is natural and Homo sapiens all over the world and across time are driven by the same desire for short-term profit. The most direct consequence of all this in foreign currency markets is that it means that forecasts must be unbiased predictors of future rates. Agents will take into account all relevant data in forming their expectations, and, because of the strong motive to profit combined with their rationality, persistent errors will be detected and eliminated rather quickly. Irrational behavior, were it to be exhibited at all, is soon driven out by the smart money, and currency prices adjust only as quickly as the underlying determinants (i.e., factors related to trade flows) change. Rates move smoothly to their new levels.

Unfortunately, the stylized facts of currency trade match with very little of the above. In the real world it appears that capital flows are terribly important in determining currency prices, trade is rarely balanced (nor does there appear to be a tendency toward that level), volatility is the rule rather than the exception, and technical trading analysis--an irrational method if prices already reflect their own history--is widely used. Furthermore, empirical studies of rational expectations and market efficiency have found little support for either one (despite the fact that most of these studies have been undertaken by neoclassical economists themselves).

The facts listed above are so obviously in evidence that neoclassical economists have not been able to ignore them. In fact, in the late 1980s and early 1990s, as study after study brought their theories into question, it appeared as if neoclassicism might begin to search for new analytical tools in their attempt to explain currency markets. However, this did not occur. There was indeed a shift in focus, but it was to the long run, where neoclassical scholars believed that their micro and macro characterizations were rejected the least often (though still more often than not). (2) Short-run currency price determination has been all but abandoned as a phenomenon about which economists, whose purview is rational behavior, have little to say.

That is not the assumption made in this paper. That neoclassicists might feel that they have little to contribute to explanations of short-term (where "short" can imply years) foreign exchange rate movements is not surprising. However, that does not mean that economists cannot explain it, just that we must adopt new tools. That process begins below, starting with definitions of the basic actors in the market.

Structure of the Market

The Participants

The definitions of market participants employed here are by function rather than by example. In other words, players are classified by what they do rather than who they are. I hope using such a method will avoid some of the confusion often encountered in this literature. Having said this, there are three basic groups of exchange market participants: primary price makers, secondary price makers, and price takers (Bishop and Dixon 1992, 125-133). Those in the first are distinguished by the fact that they are willing to make two-way offers on a continuous basis, meaning that they stand ready to buy and sell the currencies in which they are considered primary price makers. Banks and other large financial institutions capable of maintaining currency trading rooms are typically the only ones willing to offer this dealership service. In contrast, secondary price makers' offers are one way. They are willing only to buy or sell a particular money, and consequently they do not need to maintain a staff of active currency dealers. They purchase their funds from a primary price maker and sell them, at a markup, to their customers. Small banks, restaurants, and hotels are especially likely to perform this service)

Price takers are, roughly speaking, the final customers of the price-making entities. Those entering the market at this level fall into three categories: importers, direct investors, and portfolio-capital investors. (4) Importers buy foreign currency in anticipation of purchasing foreign goods or services. Direct investors wish to undertake long-term capital investments (often in the form of multinational subsidiaries) abroad. Portfolio-capital investors enter the foreign-exchange market to purchase the currencies that will allow them to acquire international assets, including interest-earning deposits of the money itself. (5)

It is possible, even likely, that one entity may play all three roles in the market. (6) This could be because it is a primary price maker in one currency, a secondary price maker in another, and a price taker in a third, or its role may change over time and from one transaction to the next.

The Participants' Roles

In a market like that for foreign exchange, prices are determined by the continuous interaction and negotiation among the market participants. Excess supplies or demands can be quickly translated into price movements. The question of what determines prices, then, becomes what determines the relative demands for currencies? That question will be answered here by looking more closely at the primary price makers, secondary price makers, and price takers and then specifying their behavior in the market.

Beginning with secondary price makers, their influence on exchange rates is minimal. Agents performing this...

To continue reading