Persistent dynamics in (in)determinate equilibrium rational expectations models

Equilibrium indeterminacy in rational expectations models is often claimed to produce higher time series persistence relative to determinacy. Proceeding by means of a simple linear stochastic model, I formally show that, for reasonable parameter configurations, there exists an uncountable (continuously infinite) set of indeterminate equilibria in low-order AR(MA) representation, which exhibit strictly lower persistence than their determinate counterpart. Implications for empirical studies concerned with, e.g., testing for indeterminacy and macroeconomic forecasting are discussed.