50 Pages Posted: 3 Feb 2017
Date Written: January 2, 2017
In classical perfect and complete markets, prices form a Martingale and stock returns (or equivalently, successive price changes) are serially uncorrelated. However, there is considerable evidence in the finance literature showing that stock returns are serially correlated both in the short and the long-term. This empirical phenomenon has been viewed as a violation of semistrong efficiency and has resulted in considerable discussion in the literature. In this paper we demonstrate that within a multi-period noisy rational expectations equilibrium framework, a first order autoregressive (AR-1) liquidity trading process, by itself, suffices to give rise to systematic correlations in price changes, either positive or negative, depending on the specific parameters of the process, even if the (unknown) underlying liquidation value is fixed. That is, unsystematic random fluctuations in observed prices arising from factors such as liquidity trading affect Bayesian belief formation, and thereby trading strategies, in such a way that equilibrium price changes can manifest both momentum and reversals.
Keywords: Noisy Rational Expectations, Multi-period model, Price Momentum
JEL Classification: C61, C73, D47, D53
Suggested Citation: Suggested Citation
Dontoh, Alex and Ronen, Joshua and Sarath, Bharat, The Dynamics of Belief Formation and Price Momentum (January 2, 2017). Available at SSRN: https://ssrn.com/abstract=2910570