|
Summary /
Abstract
|
I study how two different monetary policy regimes characterized by their difference in degrees of credibility (a 'commitment' regime, in which the central bank can credibly commit to future policy and a 'discretion' regime, in which it cannot) affect the term structure of interest rates and attempt to evaluate which monetary policy regime seems more consistent with the data on macroeconomic variables and term structure dynamics. To this end, I construct a no-arbitrage affine-term structure model based on a New-Keynesian type micro-foundation. The model is augmented with Epstein-Zin (EZ) preferences, real wage rigidity and a simple central bank optimization problem. A shock structure that exhibits stochastic volatility in long-run risk of TFP growth parsimoniously generates time-varying term premia. The estimation of the model suggests that the assumption of a discretion regime performs better than a commitment regime in terms of quantitatively fitting some salient features of the US data on the term structure and the business cycle during the Volcker-Greenspan-Bernanke era. The lack of policy credibility leads to volatile and persistent inflation, which generates volatile expected long-run inflation that is negatively correlated with future continuation values. This is perceived particularly risky by EZ nominal bond holders and results in upward sloping average nominal yields, long-term yield volatility and excess return predictability closer to the magnitude observed in the data while keeping the unconditional volatilities of consumption growth and inflation realistic. |
|
Keywords
|
Monetary policy regime, commitment, discretion, term structure, recursive utility, Volcker-Greenspan-Bernanke era |