COMMENTS ON SPECULATION AND THE VOLATILITY OF FOREIGN CURRENCY EXCHANGE RATES ROBERT P. FLOOD International Monetary Fund and Northwestern University Singleton's paper raises a number of empirical and theoretical issues involving asset-pricing in general and exchange-rate determination in par- ticular. My comments, though, will almost all be directed toward what I believe to be the issue providing the basic motivation for the paper. This, I think, is the recent attempt to salvage (rejected) exchange-rate models by injecting the models' solutions with "bubbles." I will also devote attention to Singleton's interesting ideas concerning revamping the way we think about exchange-rate models. My comments will center on what we have learned from three recent papers by Richard Meese and Kenneth Rogoff (1983a,b, 1985). While many of my points bear directly on Singleton's work, I see these comments largely as my views of Singleton's topic. EXCHANGE-RATE MODELS AND ATTEMPTSTO SAVE THEM Following major exchange-rate movements there are often discussions of under- or over-valuation of currencies. I usually find these discussions difficult to follow. Often the discussions involve strong but differing views among economists concerning what the exchange value of some currency "should" be. Stated normatively I understand such a proposition to mean that the proponent of the proposition thinks that a change in the relative values of some set of currencies will produce an allocation that he prefers to the existing one. When stated positively, however, these statements are confusing to me. Consider, for example, the statement: "The dollar is x% over/under- valued." In positive terms, I interpret the statement to mean that the 0 167 - 22311871303.50 6 1987 Elsrwer Sc~ence Publishers B.V. (North-Holland)