Thursday 11 September

Can we improve market forecasts with community predictions?

Cem Peker – NYU Abu Dhabi

Abstract:

Markets aggregate information and produce forecasts/expectations on uncertain quantities. This paper investigates how wisdom of the crowd can be used to improve the accuracy of market predictions. We present evidence from two empirical studies. First study collects data from an online sports betting community, where anonymous users self-report their picks given the odds. Bookmakers’ odds are considered as the market prediction on an event. Second study uses crowdsourced earnings-per-share forecasts from an online financial forecasting platform. Results suggest that crowd wisdom can be used to recalibrate the market predictions. Previous work found evidence for the incremental value of community predictions over the market forecast. Our contribution is to explore forecast transformations that make use of the crowd wisdom to improve accuracy.

Registration, please contact robin@em-lyon.com

Room B3-103, Lyon campus