Publications

Consensus forecasts by professionals are highly accurate, yet hide large heterogeneity. We develop a framework to extract the judgement component from survey forecasts and analyse the extent to which it contributes to respondents’ disagreement. For the average respondent, we find a substantial contribution of judgement about the current quarter, which often steers unconditional forecasts towards the realisation, thereby improving accuracy. We identify the structural components of judgement by exploiting stochastic volatility and give an economic interpretation to expected future shocks. For individual respondents, just over one-third of the disagreement is due to differences in the coefficients or models used, and the remainder is due to different assessments of future shocks; the latter mostly concerns the size of the shocks, while there is general agreement on their source.

Keywords: Expectations Formation, Identification via Stochastic Volatility, Judgement, Survey of Professional Forecasters

JEL classification: C32, C33, C51, D84, E37