Information weighting under least squares adaptive learning

This note evaluates how adaptive learning agents weigh different pieces of information when forming expectations with a recursive least squares algorithm. The analysis is based on a new and more general non-recursive representaion of the learning algorithm, namely, a penalized weighted least squares estimator, where a penalty term accounts for the effects of the learning initials. The paper then draws behavioral implications of diferent specifications of the learning mechanism, such as the cases with decreasing-, constant-, regime-switching, and age-dependent gains. The latter is shown to imply the emergence of "dormant memories" as the agents get old.