GRESHAM’S LAW OF MODEL AVERAGING

A decision maker doubts the stationarity of his environment. In response, he uses two models, one with time-varying parameters, and another with constant parameters. Forecasts are then based on a Bayesian Model Averaging strategy, which mixes forecasts from the two models. In reality, structural parameters are constant, but the (unknown) true model features expectational feedback, which the reduced form models neglect. This feedback permits fears of parameter instability to become self-confirming. Within the context of a standard linear present value asset pricing model, we use the tools of large deviations theory to show that even though the constant parameter model would converge to the (constant parameter) Rational Expectations Equilibrium if considered in isolation, the mere presence of an unstable alternative drives it out of consideration. JEL Classification Numbers: C63, D84

[1]  Ignacio Esponda,et al.  Berk-Nash Equilibrium: A Framework for Modeling Agents with Misspecified Models , 2014, 1411.1152.

[2]  L. Hansen,et al.  Uncertainty Outside and Inside Economic Models , 2014 .

[3]  T. Sargent,et al.  Bayesian Model Averaging, Learning and Model Selection , 2012 .

[4]  J. Benhabib,et al.  Learning, Large Deviations and Rare Events , 2011 .

[5]  V. Borkar.  Stochastic Approximation: A Dynamical Systems Viewpoint , 2008 .

[6]  Daron Acemoglu,et al.  Fragility of Asymptotic Agreement Under Bayesian Learning , 2008 .

[7]  Kenneth Kasa,et al.  Learning and Model Validation , 2007 .

[8]  Kenneth Kasa,et al.  Learning, Large Deviations, and Recurrent Currency Crises , 2004 .

[9]  Noah Williams,et al.  Impacts of Priors on Convergence and Escapes from Nash Inflation , 2003 .

[10]  Bruce McGough,et al.  STATISTICAL LEARNING WITH TIME-VARYING PARAMETERS , 2003, Macroeconomic Dynamics.

[11]  Ravi Bansal,et al.  Risks for the Long Run: A Potential Resolution of Asset Pricing Puzzles , 2000 .

[12]  A. Marcet,et al.  Recurrent Hyperinflations and Learning , 2003 .

[13]  A. Timmermann.  Excess Volatility and Predictability of Stock Prices in Autoregressive Dividend Models with Learning , 1996 .

[14]  H. Kushner,et al.  Analysis of adaptive step size SA algorithms for parameter tracking , 1994, Proceedings of 1994 33rd IEEE Conference on Decision and Control.

[15]  Mordecai Kurz.  On the structure and diversity of rational beliefs , 1994 .

[16]  E. Kalai,et al.  Rational Learning Leads to Nash Equilibrium , 1993 .

[17]  T. Sargent.  Bounded rationality in macroeconomics , 1993 .

[18]  G. Evans,et al.  Adaptive forecasts, hysteresis, and endogenous fluctuations , 1993 .

[19]  J. Bullard.  Time-varying parameters and nonconvergence to rational expectations under least squares learning , 1992 .

[20]  T. Sargent,et al.  Convergence of Least Squares Learning Mechanisms in Self- Referential Linear Stochastic Models* , 1989 .

[21]  P. Dupuis,et al.  Asymptotic behavior of constrained stochastic approximations via the theory of large deviations , 1987 .

[22]  M. Bray,et al.  Rational Expectations Equilibria, Learning, and Model Specification , 1986 .

[23]  Warren E. Weber,et al.  Gresham's Law or Gresham's Fallacy? , 1983, Journal of Political Economy.

[24]  R. Lucas.  Econometric policy evaluation: A critique , 1976 .

[25]  Raman K. Mehra,et al.  Approaches to adaptive filtering , 1970 .

[26]  J. Muth.  Optimal Properties of Exponentially Weighted Forecasts , 1960 .