How Memory Shape Drives Stability: Gamma Distributed Delays in a Heterogeneous Cournot Duopoly
When firms adjust output with a lag, does the shape of their memory over past observations matter for market stability, or only the average delay? This paper answers the question within a continuous-time heterogeneous Cournot duopoly in which one firm follows a gradient rule and the other a local monopolistic approximation. We replace the conventional fixed point delay with weak and strong Gamma distributed delays and convert the resulting integro-differential system into an exact finite-dimensional representation via the linear chain trick. In the homogeneous-cost benchmark the local system decouples, yielding a complete analytical comparison: the fixed delay is the most destabilizing, the strong Gamma kernel postpones the first Hopf crossing, and the weak Gamma kernel eliminates it entirely. In the heterogeneous-cost case, cost asymmetry activates cross-feedback between the firms, and Hopf thresholds become parameter-dependent. The central conclusion is that stability is governed not only by the mean delay but also by the way past information is distributed over time: the same average lag can be stabilizing or destabilizing depending on whether memory is concentrated at a single date, exponentially decaying, or hump-shaped. JEL classification: C62, C63, D43, L13
Authors
- Stefania Ragni (ORCID: https://orcid.org/0000-0001-5864-8406)
- Luca Guerrini (ORCID: https://orcid.org/0000-0001-8489-5531)
Institutions
- Marche Polytechnic University (IT)
- University of Ferrara (IT)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-18
- DOI
- https://doi.org/10.21203/rs.3.rs-10213336/v1
- Primary Topic
- Merger and Competition Analysis
- Type
- preprint