The Term Structure of Interest Rates in a DSGE Model

TLDR

The paper evaluates the implications of the Smets and Wouters (2004) DSGE model for the US yield curve. Bond prices are modeled consistently with the macro model, making long‑term risk premiums depend solely on macro parameters. Matching the observed 10‑year term premium requires only modest increases in risk aversion and habit, achieved by lowering the monetary policy parameter; this less aggressive policy raises the term premium by increasing inflation persistence, strengthening the link between consumption substitution and bond prices, and turning the inflation premium positive, thereby reconciling the model with the yield curve.

Abstract

The paper evaluates the implications of the Smets and Wouters (2004) DSGE model for the US yield curve. Bond prices are modelled in a way that is consistent with the macro model and the resulting risk premium in long term bonds is a function of the macro model parameters exclusively. When the model is estimated under the restriction that the implied average 10-year term premium matches the observed premium, it turns out that risk aversion and habit only need to rise slightly, while the increase in the term premium is achieved by a drop in the monetary policy parameter that governs the aggressiveness of the monetary policy rule. A less aggressive policy increases the persistence of the reaction of inflation and the short interest rate to any shock, reinforces the covariance between the marginal rate of substitution of consumption and bond prices, turns positive the contribution of the inflation premium and drives the term premium up. The paper concludes that by generating persistent inflation the presence of nominal rigidities can help in reconciling the macro model with the yield curve data.