Predictable Biases in Macroeconomic Forecasts and Their Impact Across Asset Classes
43 Pages Posted: 27 Jul 2017 Last revised: 21 Sep 2020
Date Written: August 19, 2020
Professional forecasters of economic data are remunerated based on accuracy and positive publicity generated for their firms. This remuneration structure incentivizes them to stick to the consensus but also to make bold forecasts when they perceive to have private information. We find that bold forecasts create skewness in the distribution of expectations and predict economic surprises across a wide range of US economic indicators, confirming our hypothesis that forecasters behave strategically and possess private information. The prevalence of this rational bias depends on the attention given to the economic indicator being released. Further, expected surprises are predictive of returns across asset classes around data announcements but this relation is strongly conditioned on the stage of the economic cycle and on the economic release being inflation- or growth-related. The forecasters' rational bias in US macroeconomic forecasts is also exhibited in individual forecasters' data as well as in Continental Europe, the UK and Japan.
Keywords: Anchoring; Rational Bias; Economic Surprises; Predictability; Stocks, Bonds; Currencies; Commodities; Machine learning
JEL Classification: G14; F47; E44
Suggested Citation: Suggested Citation