摘要：This report analyses some of the reasons mentioned in the literature as to why central banks change interest rates at discrete intervals in the face of a continuously changing environment (interest rate stepping) and why they seem to prefer to implement changes in a series of small steps (interest rate smoothing). Despite the fact that both seem difficult to explain within the certainty equivalent optimal control framework frequently used to study monetary policy, it appears that modifying some of the assumptions of these models go a long way in explaining these phenomena. In particular, replacing the assumption that all uncertainty takes the form of additive shocks with known probability distributions with perhaps more realistic ways to model uncertainty will often yield stepping and smoothing as features of an optimal interest rate rule
