Predictable Biases in Macroeconomic Forecasts and Their Impact Across Asset Classes

43 Pages Posted: 27 Jul 2017 Last revised: 21 Sep 2020

See all articles by Luiz F. F. Felix

Luiz F. F. Felix

APG Asset Management

Roman Kräussl

Luxembourg School of Finance; Hoover Institution, Stanford University

Philip A. Stork

Vrije Universiteit Amsterdam, School of Business and Economics; Tinbergen Institute

Date Written: August 19, 2020

Abstract

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

Felix, Luiz F. F. and Kraeussl, Roman and Stork, Philip A., Predictable Biases in Macroeconomic Forecasts and Their Impact Across Asset Classes (August 19, 2020). Available at SSRN: https://ssrn.com/abstract=3008976 or http://dx.doi.org/10.2139/ssrn.3008976

Luiz F. F. Felix (Contact Author)

APG Asset Management ( email )

Gustav Mahlerplein 3
Amsterdam, 1082 MS
Netherlands

Roman Kraeussl

Luxembourg School of Finance ( email )

4, rue Albert Borschette
Luxembourg, 1246
Luxembourg
+3524666445442 (Phone)

HOME PAGE: http://www.art-finance.com

Hoover Institution, Stanford University ( email )

Stanford, CA 94305
United States

Philip A. Stork

Vrije Universiteit Amsterdam, School of Business and Economics ( email )

De Boelelaan 1105
Amsterdam, 1081HV
Netherlands

Tinbergen Institute ( email )

Gustav Mahlerplein 117
Amsterdam, 1082 MS
Netherlands

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
137
Abstract Views
916
rank
236,462
PlumX Metrics