A Tractable Framework for Analyzing a Class of Nonstationary Markov Models
Lilia Maliar, Serguei Maliar, John B. Taylor, Inna Tsener
Abstract:
We consider a class of infinite鈥恏orizon dynamic Markov economic models in which the parameters of utility function, production function, and transition equations change over time. In such models, the optimal value and decision functions are time鈥恑nhomogeneous: they depend not only on state but also on time. We propose a quantitative framework, called extended function path (EFP), for calibrating, solving, simulating, and estimating such nonstationary Markov models. The EFP framework relies on the turnpike theorem which implies that the finite鈥恏orizon solutions asymptotically converge to the infinite鈥恏orizon solutions if the time horizon is sufficiently large. The EFP applications include unbalanced stochastic growth models, the entry into and exit from a monetary union, information news, anticipated policy regime switches, deterministic seasonals, among others. Examples of MATLAB code are provided.