واسطه گران منطقی و نوسانات نرخ ارز

This paper suggests a plausible microstructural connection between rational speculative activity and exchange rate volatility. When Friedman (Essays in Positive Economics, University of Chicago Press, 1953) claimed that rational speculators must smooth exchange rates, he excluded interest rate differentials from his interpretation of speculator behavior. Informed, rational speculators who consider interest differentials will magnify the exchange rate effects of interest shocks and could increase overall exchange rate volatility. This connection between speculators and volatility, which does not rely on asymmetric information, is structural because speculators affect the exchange rate's generating process. Rational speculation is stabilizing at low levels of speculative activity and destabilizing at high levels.

A strong correlation seems to exist between trading volume and price volatility in major currency markets (Baillie and Bollerslev, 1991; Dacorogna et al., 1993; Ito et al., 1996). Evidence for such a correlation is also abundant for major equity and bond markets (Cornell, 1981; Gallant et al., 1992). Many observers would argue that the high trading volume reflects high speculative activity which, in turn, induces the high price volatility. In fact, over 90% of foreign exchange market participants in Japan, Hong Kong, and Singapore believe that speculation increases volatility (Cheung and Wong, 1996). Others claim that rational speculation must reduce exchange rate volatility. The classic statement of the latter position comes from Milton Friedman (1953, p. 175): `People who have argued that speculation can be destabilizing seldom realize that this is largely equivalent to saying that speculators lose money, since speculation can be destabilizing in general only if speculators sell when the currency is low in price and buy when it is high'. He also points out that speculators who regularly lose money this way will be driven out of the market by speculators with more successful strategies. In sum, Friedman's position is that only rational speculators will survive in the market, and that rational speculation cannot be destabilizing. Important policy issues hinge on the resolution of this debate. The elimination of capital controls in Europe has coincided with a renewal of intra-ERM turbulence which threatens the viability of the EMU. Viewing this as the result of heightened speculative activity, some observers have argued for the reimposition of capital controls, if only on an as-needed basis (Eichengreen et al., 1995). Others with a similar view of speculative activity have argued for the imposition of a foreign-exchange turnover tax (Tobin, 1974; Eichengreen et al., 1994). If Friedman is right, however, a policy-induced reduction of speculative flows would increase foreign exchange volatility, worsening rather than improving the situation. This paper shows that rational speculators can but need not increase exchange rate volatility and that, contrary to Friedman's argument (Friedman, 1953), the circumstances under which they might increase volatility are plausible. An examination of Friedman's line of reasoning reveals that it does not incorporate interest rates or risk, both crucial factors for many speculators when they choose the size and direction of their positions. Changing interest differentials across countries could lead rational speculators to buy currency even when its value is `high', or to sell when its value is `low', thus `destabilizing' the exchange rate. The result is derived in a straightforward model of the foreign exchange market with two types of traders: `speculators' and `current account traders'. The speculators are rational and fully informed. Current account traders are analogous to liquidity traders in standard finance models, and can be interpreted realistically in the foreign exchange context as importers and exporters of goods and services. We find that speculators' effect on exchange rate volatility varies according to the types of shocks hitting the market, and we divide these shocks into two categories. Some shocks, such as changes in liquidity demand, do not affect speculators' preferred portfolio positions directly. An increase in speculation dampens the exchange rate impact of these shocks, consistent with Friedman's view. Other shocks, such as changes in interest rates or risks, do directly change speculators' preferred portfolio positions. As more speculators are introduced into the market, their total reaction to these shocks increases, inducing a rise in the exchange rate's reaction to the shock – an outcome entirely at variance with Friedman's view. These mixed effects of speculators on exchange rate volatility sort themselves out according to the level of speculative activity. At low levels of speculative activity, the Friedman effect dominates and the introduction of more speculators reduces exchange rate volatility; at high levels the reverse is true. The results of the paper support Flood and Taylor's observation (Flood and Taylor, 1996) that there may be `speculative forces at work in the foreign exchange market which are not reflected in the usual menu of macroeconomic fundamentals' (p. 9). In particular, the results suggest that the fundamental determinants of exchange rate dynamics include microstructural factors such as the extent of speculative activity. By pointing to the potential importance of microstructural factors in exchange rate dynamics, our results could potentially help explain the increase in real and nominal exchange rate volatility following the industrial world's 1973 shift to floating exchange rates. Evidence provided in Flood and Rose (1995) and Baxter and Stockman (1989) strongly suggests that this change cannot be explained by increased volatility among underlying macroeconomic variables. The results of the present paper suggest that the enormous increase in speculative activity since 1973 could provide an alternative explanation. Our analysis also suggests a possible explanation for the strong high-frequency connections between the activity of speculators and financial price volatility (such as that highlighted by Ito et al. (1996). French and Roll (1986) divide the possible explanations for these observed connections into three groups, one which relies on public information, one which relies on private information, and a third which relies on pricing errors. Our analysis suggests a fourth explanation, orthogonal to these original three: a rise in speculative activity could fundamentally affect financial price dynamics, even controlling for information availability and assuming rational pricing, by changing the way prices respond to information. The conclusion that speculators can increase financial price volatility is certainly not new. Numerous examples of destabilizing speculation were developed in response to Friedman's claim. Early suggestions came from Baumol (1957), Stein (1961), and Farrell (1966). More recently, Hart and Kreps (1986) show that `speculative activity in an economy in which all agents are rational, have identical priors, and have access to identical information may destabilize prices, under any reasonable definition of stabilization' (p. 927). Likewise, Stein (1987) shows that `introducing a new group of speculators into the spot market for a commodity can destabilize prices' (p. 1124), even when speculators are fully informed and have rational expectations. Hau (1995) shows that speculators could increase exchange rate volatility if individual exchange rate expectations differ. Though the mechanisms highlighted in these earlier papers are all plausible, many rely on a set of narrowly defined circumstances that may not be present in reality. For example, the Hart and Kreps (1986) model requires a very specific relationship between stochastic consumption behavior and signals about that behavior. Stein's (1987) model requires that the new group of speculators introduced to the market have access to an information source unavailable to the original speculators. The mechanism highlighted in the present paper, by contrast, will always operate in foreign exchange markets. Analogous mechanisms can also be found in other financial markets. In stock and commodity markets, for example, changes in the domestic interest rates directly affect speculators' desired positions and could cause speculators to destabilize rather than stabilize prices. Non-rational speculators could also destabilize financial prices, as shown in a group of papers including Frankel and Froot (1990) and DeLong et al. (1990). Though this paper is concerned with rational speculators, its results are not inconsistent with these important papers. This paper does not address the welfare implications of speculative behavior. Though there is a presumption among naive observers that increased volatility reduces welfare, academics have noted repeatedly that the reverse could be true (see, for example, Stein 1987). Since the welfare of non-speculative agents is not modeled explicitly here, the paper focuses exclusively on the implications of speculative activity for exchange rate dynamics, without considering whether changes in those dynamics benefit or harm welfare. Section 2develops and solves the model. Section 3analyzes the dynamics associated with specific types of shocks and Section 4shows how the degree of speculation affects overall exchange rate volatility. Section 5considers a few extensions, and Section 6concludes.

This paper has developed a model highlighting a plausible positive connection between rational speculative activity and exchange rate volatility. Speculators can increase volatility because they magnify the exchange rate influence of shocks to which they respond directly. This class of shocks, illustrated in the model by changes in interest differentials, would also include changes in taxes and transaction costs, changes in perceived risk, or changes in risk aversion. Speculators do not magnify the exchange rate effects of all shocks: on the contrary, they tend to smooth the effects of shocks which do not directly affect their desired positions, such as shocks to international trade. When these are the only shocks, speculators' behavior will correspond directly to that described by Friedman in considering the merits of flexible exchange rates: they will sell when the exchange rate is high (relative to their long-run expected value) and buy when the exchange rate is low, thereby smoothing its path. Despite the varied effects of speculators on the exchange rate's response to shocks, their net effect on volatility has a clear pattern. At low levels of speculation, a rise in speculative activity reduces exchange rate volatility. However, beyond some point, any further rise in speculation increases exchange rate volatility. Unlike the speculation-volatility connections suggested by French and Roll (1986), the one suggested here does not rely on informational asymmetries between speculators and other agents or on irrationality. Instead, the model discussed here suggests that the presence of speculators changes the exchange rate's response to given shocks. In this sense it indicates that an increase in speculative activity can cause changes in foreign exchange market microstructure. One implication of this research is that the dramatic rise in exchange rate volatility since the advent of floating exchange rates could be at least partly explained by a concurrent increase in currency speculation. Other possible causes of the increased exchange rate volatility, such as increased volatility of fundamentals themselves, have not succeeded in explaining the phenomenon (Flood and Rose, 1995; Baxter and Stockman, 1989). The possibility that speculation itself has been a source of increased exchange rate volatility since 1973 may be a worthy topic of future research.