Linear rational-expectations models with lagged expectations: A synthetic method

This paper contains a solution and an estimation method for linear rational-expectations models with lagged expectations. The solution method is a synthetic approach, combining state-space and infinite-MA representations with a simple system of linear equations. The advantage lies in the particular combination of methods from the literature, providing faster execution, more general applicability, and more straightforward usage than existing algorithms. Bayesian estimation methods are employed without the Kalman filter using a recursive algorithm to evaluate the likelihood function and are used to compare small-scale sticky-information and sticky-price DSGE models. Standard truncation methods are shown to not generally be innocuous.