Reading the minds of investors: an empirical term structure model for policy analysis

Building on the recent macro finance literature, this paper develops an empirical term structure model in which investors' judgmental forecasts of macro variables play an important role. The model allows for a limited form of time-variation in the dynamics describing the behavior of short-term interest rates and macro variables. As a result, changes in economic forecasts over time reflect the influence of both economic shocks and perceived changes in economic structure. The latter, in particular, are shown to be important in explaining the evolution of the yield curve over time. An interest rate accounting framework based on the model is applied in parsing changes in long-term interest rates into portions associated with changes in term premiums and changes in expected future short-rates. The changes in expected future short rates are then further decomposed into portions attributable to changes in the expected future paths for inflation, the unemployment rate, and GDP growth and also to a fourth factor interpreted as changes in the "stance of monetary policy." The model results indicate that changes in long-term interest rates, on average, have been about equal parts changes in term premia and changes in expected future short rates. Changes in expected future short rates seem to be driven largely by changes in the stance of monetary policy and in the outlook for inflation while the estimated influence of changes in the outlook for the unemployment rate and GDP growth is more muted.