Switching between Chartists and Fundamentalists: A Markov Regime-Switching Approach

Since the early 1980s, models based on economic fundamentals have been poor at explaining the movements in the exchange rate (Messe 1990). In response to this problem, Frankel and Froot (1988) developed a model that uses two approaches to forecast the exchange rate: the fundamentalist approach, which bases the forecast on economic fundamentals, and the chartist approach, which bases the forecast on the past behaviour of the exchange rate. This was an innovation, as only the fundamentalist approach had been used before. A feature of the chartist-and-fundamentalist (c&f) model is that these two approaches' relative importance varies over time. Because this weighting is unobserved, the c&f model can not be estimated or tested using standard techniques. To overcome these difficulties and to test the model, the author uses Markov regime-switching techniques. He defines the two groups' different methods of forecasting as regimes and rewrites the c&f model as a regime-switching model. The model is then used to test for c&f behaviour in the Canada-U.S. daily exchange rate between 1983 and 1992. The author finds favourable though inconclusive evidence for the c&f model and accordingly makes suggestions for further research. French Version Depuis le debut des annees 80, les modeles fondes sur les facteurs economiques fondamentaux n'ont guere contribue a expliquer les variations du taux de change (Messe 1990). Devant ce probleme, Frankel et Froot (1988) ont elabore un modele dans lequel ils ont utilise deux approches de prevision du taux de change : l'approche fondamentaliste, dans laquelle la prevision repose sur des facteurs economiques fondamentaux, et l'approche technique, ou la prevision s'appuie sur le comportement passe du taux de change. Il s'agissait la d'une innovation car, auparavant, seule l'approche fondamentaliste etait utilisee. Une caracteristique propre au modele fondamentaliste et technique (f et t) est que le poids relatif des deux composantes varie avec le temps. Comme cette ponderation n'est pas observable, le modele ne peut etre estime ni teste a l'aide des techniques standard. Pour surmonter ces difficultes et tester le modele, l'auteur fait appel aux techniques de changement de regime de Markov. Il definit chacune des deux approches comme un different regime de generation d'informations et transforme le modele f et t en un modele avec changement de regime. Le modele est ensuite utilise pour tester le comportement du taux de change au jour le jour Canada-Etats-Unis pour la periode 1983-1992. Les resultats favorables quoique non concluants du modele amenent l'auteur a suggerer la poursuite des recherches.

[1]  R. Meese,et al.  Currency Fluctuations in the Post-Bretton Woods Era , 1990 .

[2]  James D. Hamilton Specification testing in Markov-switching time-series models , 1996 .

[3]  Simon van Norden,et al.  Terms of trade and real exchange rates: the Canadian evidence , 1995 .

[4]  Stephen L Taylor,et al.  Rewards Available to Currency Futures Speculators: Compensation for Risk or Evidence of Inefficient Pricing? , 1992 .

[5]  B. LeBaron,et al.  Simple Technical Trading Rules and the Stochastic Properties of Stock Returns , 1992 .

[6]  L. Summers,et al.  Positive Feedback Investment Strategies and Destabilizing Rational Speculation , 1989 .

[7]  James D. Hamilton,et al.  Long Swings in the Dollar: Are They in the Data and Do Markets Know It? The American Economic Review , 1990 .

[8]  William L. Goffe,et al.  Simulated annealing: An initial application in econometrics , 1992 .

[9]  K. Pilbeam THE PROFITABILITY OF TRADING IN THE FOREIGN EXCHANGE MARKET: CHARTISTS, FUNDAMENTALISTS, AND SIMPLETONS , 1995 .

[10]  Tad Hogg,et al.  Bubbles and Market Crashes , 1994, adap-org/9409001.

[11]  The Empirical Performance of Alternative Monetary and Liquidity Aggregates , 1995 .

[12]  Bruce E. Hansen,et al.  Erratum: The likelihood ratio test under nonstandard conditions: Testing the Markov switching model of GNP , 1996 .

[13]  Charles Goodhart,et al.  The Foreign Exchange Market: A Random Walk with a Dragging Anchor , 1988 .

[14]  Hans Dewachter,et al.  A chaotic model of the exchange rate: The role of fundamentalists and chartists , 1993 .

[15]  C. Engel Can the Markov Switching Model Forecast Exchange Rates? , 1992 .

[16]  Thomas H. McCurdy,et al.  Duration-Dependent Transitions in a Markov Model of U.S. GNP Growth , 1994 .

[17]  G. Kaminsky Is There a Peso Problem? Evidence from the Dollar/Pound Exchange Rate, 1976-1987 , 1993 .

[18]  John B. Moore,et al.  Hidden Markov Models: Estimation and Control , 1994 .

[19]  Robert J. Vigfusson,et al.  Analytical Derivatives for Markov Switching Models , 1997 .

[20]  Kenneth A. Froot,et al.  Chartists, Fundamentalists and the Demand for Dollars , 1991 .

[21]  James D. Hamilton 9 Estimation, inference and forecasting of time series subject to changes in regime , 1993 .

[22]  J. Poterba,et al.  Speculative Dynamics and the Role of Feedback Traders , 1990 .

[23]  Mark P. Taylor,et al.  The use of technical analysis in the foreign exchange market , 1992 .

[24]  P. Grauwe Exchange rates in search of fundamental variables , 1994 .

[25]  L. Summers,et al.  Noise Trader Risk in Financial Markets , 1990, Journal of Political Economy.