Csóka, Péter
ORCID: https://orcid.org/0000-0003-1703-5835 and Herings, P. Jean-Jacques
ORCID: https://orcid.org/0000-0002-1100-8601
(2022)
Centralized clearing mechanisms: a programming approach.
Journal of Mechanism and Institution Design, 7
(1).
pp. 45-69.
DOI https://doi.org/10.22574/jmid.2022.12.002
|
PDF
- Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
251kB |
Official URL: https://doi.org/10.22574/jmid.2022.12.002
Abstract
We consider financial networks where agents are linked to each other by financial contracts. A centralized clearing mechanism collects the initial endowments, the liabilities and the division rules of the agents and determines the payments to be made. A division rule specifies how the assets of the agents should be rationed. Since payments made depend on payments received, we are looking for solutions to a system of equations. The set of solutions is known to have a lattice structure, leading to the existence of a least and a greatest clearing payment matrix. Previous research has shown how decentralized clearing selects the least clearing payment matrix. We present a centralized approach towards clearing in order to select the greatest clearing payment matrix. To do so, we formulate the determination of the greatest clearing payment matrix as a programming problem. When agents use proportional division rules, this programming problem corresponds to a linear programming problem. We show that for other common division rules, it can be written as an integer linear programming problem.
| Item Type: | Article |
|---|---|
| Uncontrolled Keywords: | integer linear programming; systemic risk; Bankruptcy rules |
| JEL classification: | C71 - Cooperative Games G10 - General Financial Markets: General (includes Measurement and Data) |
| Divisions: | Institute of Finance |
| Subjects: | Finance |
| DOI: | https://doi.org/10.22574/jmid.2022.12.002 |
| ID Code: | 13446 |
| Deposited By: | MTMT SWORD |
| Deposited On: | 09 Nov 2026 14:12 |
| Last Modified: | 09 Nov 2026 14:12 |
Repository Staff Only: item control page


Download Statistics
Download Statistics