This book develops the idea of rational expectations and surveys its use in economics today.

Assumptions about how people form expectations for the future shape the properties of any dynamic economic model. To make economic decisions in an uncertain environment people must forecast such variables as future rates of inflation, tax rates, governme.

A theoretical assessment of the Rational Expectations Hypothesis through subjecting a collection of economic models to an "eductive stability" test. The rational expectations hypothesis (REH) dominates economic modeling in areas ranging from monetary theory, macroeconomics, and general equilibrium to finance. In this book, Roger Guesnerie continues the critical analysis of the REH begun in his Assessing Rational Expectations: Sunspot Multiplicity and Economic Fluctuations, which dealt with the questions raised by multiplicity and its implications for a theory of endogenous fluctuations. This second volume emphasizes "eductive" learning: relying on careful reasoning, agents must deduce what other agents guess, a process that differs from the standard evolutionary learning experience in which agents make decisions about the future based on past experiences. A broad "eductive" stability test is proposed that includes common knowledge and results in a unique "rationalizable expectations equilibrium." This test provides the basis for Guesnerie's theoretical assessment of the plausibility of the REH's expectational coordination, emphasizing, for different categories of economic models, conditions for the REH's success or failure. Guesnerie begins by presenting the concepts and methods of the eductive stability analysis in selected partial equilibrium models. He then explores to what extent general equilibrium strategic complementarities interfere with partial equilibrium considerations in the formation of stable expectations. Guesnerie next examines two issues relating to eductive stability in financial market models, speculation and asymmetric price information. The dynamic settings of an infinite horizon model are then taken up, and particular standard and generalized saddle-path solutions are scrutinized. Guesnerie concludes with a review of general questions and some "cautious" remarks on the policy implications of his analysis.

A Rational Expectations Approach to Macroeconometrics pursues a rational expectations approach to the estimation of a class of models widely discussed in the macroeconomics and finance literature: those which emphasize the effects from unanticipated, rather than anticipated, movements in variables. In this volume, Fredrick S. Mishkin first theoretically develops and discusses a unified econometric treatment of these models and then shows how to estimate them with an annotated computer program.

Do traders in futures markets make use of all relevant information and is this reflected in prices? This collection of original essays by a team of international economists considers these and other questions central to futures markets.

One of the major controversies in macroeconomics over the last 30 years has been that on the effectiveness of stabilization policies. However, this debate, between those who believe that this kind of policies is useless if not harmful and those who argue in favor of it, has been mainly theoretical so far. The Rational Expectation Hypothesis, Time-Varying Parameters and Adaptive Control wants to represent a step toward the construction of a common ground on which to empirically compare the two "beliefs" and to do this three strands of literature are brought together. The first strand is the research on time-varying parameters (TVP), the second strand is the work on adaptive control and the third one is the literature on linear stationary models with rational expectations (RE). The material presented in The Rational Expectation Hypothesis, Time-Varying Parameters and Adaptive Control is divided into two parts. Part 1 combines the strand of literature on adaptive control with that on TVP. It generalizes the approach pioneered by Tse and Bar-Shalom (1973) and Kendrick (1981) and one recently used in Amman and Kendrick (2002), where the law of motion of the TVP and the hyperstructural parameters are assumed known, to the case where the hyperstructural parameters are assumed unknown. Part 2 is devoted to the linear single-equation stationary RE model estimated with the error-in-variables (EV) method. It presents a new formulation of this problem based on the use of TVP in an EV model. This new formulation opens the door to a very promising development. All the theory developed in the first part to control a model with TVP can sic et simpliciter be applied to control a model with RE.

3 in philosophy, and therefore in metaphilosophy, cannot be based on rules that avoid spending time on pseudo-problems. Of course, this implies that, if one succeeds in demonstrating convincingly the pseudo-character of a problem by giving its 'solution', the time spent on it need not be seen as wasted. We conclude this section with a brief statement of the criteria for concept explication as they have been formulated in several places by Carnap, Hempel and Stegmiiller. Hempel's account ([13J, Chapter 1) is still very adequate for a detailed introduction. The process of explication starts with the identification of one or more vague and, perhaps, ambiguous concepts, the so-called explicanda. Next, one tries to disentangle the ambiguities. This, however, need not be possible at once. Ultimately the explicanda are to be replaced (not necessarily one by one) by certain counterparts, the so-called explicata, which have to conform to four requirements. They have to be as precise as possible and as simple as possible. In addition, they have to be useful in the sense that they give rise to the formulation of theories and the solution of problems. The three requirements of preciseness, simplicity and usefulness. have of course to be pursued in all concept formation.

A comprehensive exposition of rational expectations models is provided here, working up from simple univariate models to more sophisticated multivariate and non-linear models.