What were they thinking? Estimating the quarterly forecasts underlying annual growth projections

Dr. Christian Hepenstrick and Jason Blunier

Issue
2022-05

Pages
31

JEL classification
C53, E37

Keywords
Forecasting, frequency disaggregation, survey expectations

Year
2022

Many prominent forecasters publish their projections at an annual frequency. However, for applied work, an estimate of the underlying quarterly forecasts is often indispensable. We demonstrate that a simple state-space model can be used to obtain good estimates of the quarterly forecasts underlying annual projections. We validate the methodology by aggregating professional forecasts for quarterly GDP growth in the United States to the annual frequency and then applying our imputation methodology. The imputed forecasts perform as well as the original quarterly forecasts. Applying the imputation methodology to Consensus forecasts for other advanced economies provides further evidence of the good performance of our proposed methodology.