Solving linear rational expectations models (Q1863711): Difference between revisions

The paper deals with the following linear rational expectation model \[ \Gamma_0y(t)=\Gamma_1y(t-1)+C+\Phi z(t) + \Pi \eta(t), \] where \(t=1,2,\dots,T,\) \(C\) is a vector of constants, \(z(t)\) is an exogenously evolving, possibly serially correlated, random disturbance, and \(\eta(t)\) is an expectational error, satisfying \(E_t\eta(t+1)=0.\) The methods for solving this general linear expectations model in continuous and discrete timing with or without exogenous variables are described. The analysis of the paper is similar to that of \textit{O. J. Blanchard} and \textit{C. M. Kahn} [Econometrica 48, 1305--1311 (1980; Zbl 0438.90022)] with some differences such as: 1) the paper under consideration covers all linear models with expectational error terms; 2) the approach of the paper handles situations where linear combinations of variables, not individual variables, are predetermined; 3) this paper makes an explicit extension to continuous time; 4) different linear combinations of variables may have different growth rate restrictions.