Optimal Time-Invariant Monetary Policy.

This paper investigates how best to determine time-invariant policy rules in macroeconomic models with forward-looking constraints, where fully optimal policy is known to be time-inconsistent. It proposes a new ‘coefficient optimisation’ approach that improves upon the timeless perspective method of Woodford (2003) in deterministic problems, and on average in stochastic problems, without resorting to asymptotic (‘unconditional’) loss comparisons