Speculation and Destabilisation Kim F. Radalj a , Michael McAleer b a Department of Economics, University of Western Australia (kradalj@ecel.uwa.edu.au ) b Department of Economics, University of Western Australia Abstract: In the context of flexible exchange rates, Milton Friedman proposed that speculation must exert a stabilising influence upon prices to remain profitable. This generated a wealth of predominantly theoretical research into the behaviour of speculators, for which the results seem to depend critically upon the assumptions. Such theoretical models need to be tested against empirical evidence to determine whether speculators behave in a destabilising manner. Using recent theoretical developments in the literature on modelling financial volatility, this paper tests the significance of speculators and their contributions to describing weekly volatilities across a series of currency, metals and commodity markets. As the time- varying GARCH volatility model and its variants have been criticised for lacking economic content, incorporating speculators into such models contributes to an accommodation of this criticism. The economic implications from establishing the importance of speculators are far-reaching. Policymakers often discuss the imposition of a Tobin tax to curb speculation, so it must be established that speculators behave in economically destructive ways. The inclusion of speculators is also likely to yield superior forecasting models of volatility, and hence more efficient pricing of derivative instruments. Keywords: Volatility; speculation; destabilisation; Tobin tax; GARCH models. 1. INTRODUCTION Whether speculators cause financial markets to be more volatile than warranted by fundamentals has long been a topic of intense intellectual debate. The matter has important consequences for regulators of the international capital market framework. Capital can now travel great distances very rapidly, facilitated by advances in communications and the relaxation of capital account controls. However, whether such rapid redeployment of capital is in the best interests of world economic prosperity has been queried. For example, Tobin (1978) proposed taxing foreign exchange transactions, or ‘sand in the wheels’ to reduce the level of noise in currency markets. In his case for floating exchange rates, Friedman (1953) argued that speculators must exert a stabilising influence upon markets. This generated a wealth of counter-examples, most of which were viewed with scepticism (see Hart and Kreps (1986)). Since that time, major advancements in financial market and information theory have led to the development of models that show speculators need not act to stabilise prices, and indeed may even survive, contrary to Friedman’s assertion. An array of market models exists, deviating in different ways from the perfect competition, representative agent paradigm that has been the workhorse of economic analysis. For example, De Long et al. (1990) present a model whereby speculators may trade profitably by anticipating the direction in which a category of traders, referred to as noise traders, will trade. The actions of speculators will pre-empt noise trader demand and serves to push prices further away from their fundamental values. Other models, such as in Shalen (1993) and Harris and Raviv (1993), rely upon dispersion of beliefs among agents. These models lead to greater trading than is predicted by a simple model, and differences in beliefs cause prices to fluctuate more heavily before settling at the equilibrium price. By comparison, the amount of empirical research into how speculation affects asset markets is relatively sparse. The major reason for such absence is that researchers are limited by data availability as traders typically do not have to disclose their motivations for trading. Moreover, it is not entirely clear as to what constitutes speculation. As observed by Hart and Kreps (1986, p. 928), a satisfactory definition is unlikely to become available. Thus, the concept of speculation used is that it is associated with traders who have no underlying business interest in the commodity and are simply trading to beat the market estimate of its future value. Such a definition is similar to that used by the