Real-time forecasting in a data-rich environment

This paper assesses the ability of different models to forecast key real and nominal U.S. monthly macroeconomic variables in a data-rich environment and from the perspective of a real-time forecaster, i.e. taking into account the real-time data revisions process and data flow. We find that for the real variables predictability is confined over the recent recession/crisis period. This is in line with the findings of D’Agostino and Giannone (2012) that gains in relative performance of models using large datasets over univariate models are driven by downturn periods which are characterized by higher comovements. Regarding inflation, results are stable across time, but predictability is mainly found at the very short-term horizons. Inflation is known to be hard to forecast, but by exploiting timely information one obtains gains at nowcasting and forecasting one-month ahead, especially with Bayesian VARs. Furthermore, for both real and nominal variables, the direct pooling of information using a high dimensional model (dynamic factor model or Bayesian VAR) which takes into account the cross-correlation between the variables and efficiently deals with the “ragged edge”structure of the dataset, yields more accurate forecasts than the indirect pooling of bi-variate forecasts/models.