Collusive networks in market-sharing agreements: An agent-based simulation

Abstract. Market-sharing agreements remain a central concern for competition policy, yet the incentives and stability conditions behind such arrangements have received comparatively less attention than other forms of collusion. This paper develops an agent-based dynamic network model to study their formation and stability. The model incorporates heterogeneity across markets, firm asymmetries, and the presence of entry and transport costs. Firms endogenously decide whether to compete in each other’s markets or form agreements to avoid mutual entry. Whereas most of the literature analyzes such networks under the standard Cournot assumption, we also consider Stackelberg–Cournot competition, where incumbents can commit to higher production. We show that network stability critically depends on the competitive setting: under Stackelberg–Cournot competition, complete networks are more resilient to asymmetries, while under pure Cournot competition the empty network is typically stable. Introducing entry and transport costs alters these dynamics by generating richer network structures and, in some cases, unstable formation processes. Finally, the simulation allows us to examine how network characteristics affect market outcomes, particularly prices.

Recommended citation:
Download Paper