New Phenomena Identified in a Stochastic Dynamic Macroeconometric Model: A Bifurcation Perspective

In this paper, we consider new bifurcation phenomena in a class of stochastic dynamic macroeconometric models as represented by the stochastic model developed by Leeper and Sims (1994). This model serves as a prototype that could be suitable for monetary policy analysis although the complexity of the model makes any attempt of analytical analysis a difficult task. Leeper and Sims model consists of differential equations with a set of algebraic constraints. Our analysis reveals that singularity occurs within a small neighborhood of estimated parameter values. Singularity boundary is determined. When the parameter values are close to the singularity boundary, one eigenvalue of the linearized part of the model rapidly moves to infinity while others remain bounded, implying nearly instantaneous response of some variables to changes of other variables. On the singularity boundary, the number of differential equations will decrease while the number of algebraic constraints will increase. Such change in the order of dynamics is a new phenomenon in macroeconometric models. We shall determine the singularity-induced bifurcation and its effect on model behavior.