Deep Q-learning of Prices in Oligopolies: The Number of Competitors Matters

Herbert Dawid, Philipp Harting and Michal Neugart
Additional contact information
Herbert Dawid: Bielefeld University, Germany
Philipp Harting: Université Côte d'Azur, CNRS, GREDEG, France
Michal Neugart: Technical University of Darmstadt, Germany

No 2024-32, GREDEG Working Papers from Groupe de REcherche en Droit, Economie, Gestion (GREDEG CNRS), Université Côte d'Azur, France

Abstract: Artificial intelligence algorithms are increasingly used for online pricing and are seen as a major threat to competitive markets. We show that if firms use a deep Q-network (DQN) as an example of a state-of-the-art machine learning algorithm, prices are supra-competitive in duopoly but quickly move to competitive prices as the number of competitors in an oligopoly increases. This finding is very robust concerning variations of the exploration and learning rate used in the DQN algorithm.

Keywords: algorithmic price setting; deep Q-network; oligopoly; supracompetitive prices (search for similar items in EconPapers)
Pages: 20 pages
Date: 2024-12
New Economics Papers: this item is included in nep-ain, nep-cmp, nep-com, nep-ind and nep-reg
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://195.220.190.85/GREDEG-WP-2024-32.pdf First version (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gre:wpaper:2024-32

Access Statistics for this paper

More papers in GREDEG Working Papers from Groupe de REcherche en Droit, Economie, Gestion (GREDEG CNRS), Université Côte d'Azur, France Contact information at EDIRC.
Bibliographic data for series maintained by Patrice Bougette ().