Alexander Rasch, DICE Dusseldorff

"Demand forecasting and collusive pricing with hidden actions"

Abstract

We analyze the effects of better algorithmic demand forecasting on collusive profits. In a general framework, we show that the comparative statics crucially depend on whether actions are observable or hidden. Thus, the optimal antitrust policy must pay attention to the institutional settings of the industry in question. Our analysis reveals a dual role of improving forecasting ability when actions are hidden. Deviations become more tempting, reducing profits, but also uncertainty concerning deviations is increasingly eliminated. In an application with promotional activities, this results in a u- shaped relationship between profits and prediction ability.

Contact person: Egor Starkov