Title: Extended function path perturbation methods for nonstationary and unbalanced growth models
Authors: Serguei Maliar - Santa Clara University (United States) [presenting]
Vadym Lepetyuk - Bank of Canada (Canada)
Lilia Maliar - Stanford University (United States)
Abstract: Dynamic stochastic economic models are built on the assumption that the economy's fundamentals such as preferences, technologies and laws of motions for exogenous variables do not change over time. Such models have stationary solutions in which optimal value and decision functions depend on the current state but not on time. At the same time, real-world economies constantly evolve over time, experiencing population growth, technological progress, trends in tastes and habits, policy regime changes, evolution of social and political institutions, etc. There are global solution methods for solving such nonstationary and unbalanced growth models. We apply the previous methodologies to develop local perturbation methods for solving nonstationary and unbalanced growth models. Our numerical examples show that the proposed perturbation methods can deliver solutions that are comparable in accuracy to the global solutions at considerable lower cost.