On some computational aspects of equilibrium business cycle theory (Q1263502)

The paper considers ``real business cycle theory'' and its practice to approximate, quadratically, the return function of such models in order to generate the economy's equilibrium time series. It analyzes the legitimacy of this procedure by means of a specific one-good stochastic growth model. Its findings are that quadratic approximate methods will provide good proxies for discrete state-spaces and that there is evidence for this fact for continuous state-space quadratic methodology as well.