Forecasting with a Bayesian DSGE model

In monetary policy strategies geared towards maintaining price stability conditional and unconditional forecasts of inflation and output play an important role. This paper illustrates how modern sticky-price dynamic stochastic general equilibrium models, estimated using Bayesian techniques, can become an additional useful tool in the forecasting kit of central banks. First, we show that the forecasting performance of such models compares well with a-theoretical vector autoregressions. Moreover, we illustrate how the posterior distribution of the model can be used to calculate the complete distribution of the forecast, as well as various inflation risk measures that have been proposed in the literature. Finally, the structural nature of the model allows computing forecasts conditional on a policy path. It also allows examining the structural sources of the forecast errors and their implications for monetary policy. Using those tools, we analyse macroeconomic developments in the euro area since the start of EMU.