Artificial Collusion: Examining Supra-competitive Pricing by Autonomous Q-learning Algorithms

Open Access
Authors
Publication date 2022
Series Amsterdam Law School Legal Studies Research Paper
Number of pages 48
Publisher Amsterdam: University of Amsterdam
Organisations
  • Faculty of Law (FdR)
  • Faculty of Science (FNWI) - Korteweg-de Vries Institute for Mathematics (KdVI)
  • Faculty of Economics and Business (FEB) - Amsterdam School of Economics Research Institute (ASE-RI)
  • Faculty of Economics and Business (FEB)
  • Interfacultary Research - Amsterdam Center for Law & Economics (ACLE)
Abstract We examine recent claims that a particular Q-learning algorithm used by competitors `autonomously' and systematically learns to collude, resulting in supracompetitive prices and extra profits for the firms sustained by collusive equilibria. A detailed analysis of the inner workings of this algorithm reveals that there is no immediate reason for alarm. We set out what is needed to demonstrate the existence of a colluding price algorithm that does form a threat to competition.
Document type Working paper
Language English
Published at https://doi.org/10.2139/ssrn.4213600
Downloads
Artificial Collusion (Submitted manuscript)
Permalink to this page
Back