pith. sign in

arxiv: 1907.03082 · v2 · pith:C22YIKNRnew · submitted 2019-07-06 · 💱 q-fin.MF

Systemic Risk and Heterogeneous Mean Field Type Interbank Network

Pith reviewed 2026-05-25 01:52 UTC · model grok-4.3

classification 💱 q-fin.MF
keywords interbank lendingmean field gameNash equilibriumRiccati equationsheterogeneous groupssystemic risklog-capitalizationliquidity rate
0
0 comments X

The pith

Heterogeneous interbank groups reach Nash equilibria through solvable coupled Riccati equations in the large-bank limit.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper models interbank lending and borrowing among heterogeneous groups of banks using a mean-field game where each bank minimizes costs to avoid default based on log-capitalization averages. It establishes that both closed-loop and open-loop Nash equilibria are determined by coupled Riccati equations. In the two-group case with many banks, existence follows from the solvability of these equations as group sizes approach infinity. The model extends to an arbitrary number of groups, confirming the existence of epsilon-Nash equilibria. Numerical studies show how parameters affect liquidity rates under these equilibria.

Core claim

In the heterogeneous mean-field interbank model, the Nash equilibria for lending strategies are driven by coupled Riccati equations, whose solvability in the infinite-bank limit per group guarantees the existence of equilibria in the two-group case and epsilon-Nash equilibria for d groups. The equilibria incorporate a mean-reverting term and contributions from group averages due to heterogeneity between groups.

What carries the argument

Coupled Riccati equations that determine the closed-loop and open-loop Nash equilibria in the heterogeneous mean-field game.

Load-bearing premise

The coupled Riccati equations remain solvable as the number of banks in each group tends to infinity.

What would settle it

Finding a set of parameters where the coupled Riccati equations have no solution for large finite groups, or where the derived strategies fail to form an approximate equilibrium in simulations with thousands of banks per group.

read the original abstract

We study the system of heterogeneous interbank lending and borrowing based on the relative average of log-capitalization given by the linear combination of the average within groups and the ensemble average and describe the evolution of log-capitalization by a system of coupled diffusions. The model incorporates a game feature with homogeneity within groups and heterogeneity between groups where banks search for the optimal lending or borrowing strategies through minimizing the heterogeneous linear quadratic costs in order to avoid to approach the default barrier. Due to the complicity of the lending and borrowing system, the closed-loop Nash equilibria and the open-loop Nash equilibria are both driven by the coupled Riccati equations. The existence of the equilibria in the two-group case where the number of banks are sufficiently large is guaranteed by the solvability for the coupled Riccati equations as the number of banks goes to infinity in each group. The equilibria are consisted of the mean-reverting term identical to the one group game and the group average owing to heterogeneity. In addition, the corresponding heterogeneous mean filed game with the arbitrary number of groups is also discussed. The existence of the $\epsilon$-Nash equilibrium in the general $d$ heterogeneous groups is also verified. Finally, in the financial implication, we observe the Nash equilibria governed by the mean-reverting term and the linear combination of the ensemble averages of individual groups and study the influence of the relative parameters on the liquidity rate through the numerical analysis.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper models heterogeneous interbank lending/borrowing as a mean-field game with d groups of banks, where log-capitalization evolves via coupled diffusions driven by relative averages (within-group and ensemble). Banks minimize heterogeneous linear-quadratic costs to stay away from a default barrier. Both closed-loop and open-loop Nash equilibria are characterized via coupled Riccati equations; existence for the two-group case with large group sizes is asserted via solvability of the limiting Riccati system, ε-Nash equilibria are verified for general d groups, and numerical experiments examine liquidity-rate dependence on relative parameters.

Significance. If the existence claims can be rigorously established, the framework would extend mean-field game techniques to systemic-risk modeling with explicit group heterogeneity, providing a tractable way to separate intra-group mean reversion from inter-group effects and to quantify how relative weighting parameters influence aggregate liquidity.

major comments (2)
  1. [Abstract (and corresponding two-group analysis)] Abstract and the two-group existence claim: the statement that 'the existence of the equilibria ... is guaranteed by the solvability for the coupled Riccati equations as the number of banks goes to infinity in each group' supplies no argument, comparison principle, a-priori bound, or explicit solution establishing global existence on the fixed horizon [0,T]. In heterogeneous LQ mean-field games the limiting matrix Riccati ODE can exhibit finite-time blow-up for admissible parameter regimes; the finite-N well-posedness does not automatically pass to the limit.
  2. [Section on general d heterogeneous groups] The ε-Nash verification for arbitrary d groups likewise rests on the same limiting Riccati solvability without additional justification that the coefficients (group-interaction weights and relative-average parameters) remain in the regime where global solutions exist.
minor comments (2)
  1. [Abstract] Abstract contains several typos and grammatical issues: 'complicity' should be 'complexity'; 'mean filed game' should be 'mean field game'; 'the equilibria are consisted of' should be 'the equilibria consist of'.
  2. [Model formulation] Notation for the relative-average process and the precise definition of the group-interaction matrix should be introduced earlier and used consistently when writing the Riccati system.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and for highlighting the need for stronger justification of the existence claims. We agree that the manuscript currently asserts solvability of the limiting Riccati system without supplying the requisite a-priori bounds or comparison arguments, and we will revise the relevant sections to address this gap.

read point-by-point responses
  1. Referee: [Abstract (and corresponding two-group analysis)] Abstract and the two-group existence claim: the statement that 'the existence of the equilibria ... is guaranteed by the solvability for the coupled Riccati equations as the number of banks goes to infinity in each group' supplies no argument, comparison principle, a-priori bound, or explicit solution establishing global existence on the fixed horizon [0,T]. In heterogeneous LQ mean-field games the limiting matrix Riccati ODE can exhibit finite-time blow-up for admissible parameter regimes; the finite-N well-posedness does not automatically pass to the limit.

    Authors: We acknowledge that the manuscript provides no explicit argument establishing global existence of the coupled Riccati ODE on the fixed interval [0,T]. The finite-N system is well-posed for each N, but passage to the mean-field limit indeed requires additional analysis to rule out blow-up. In the revision we will add a remark after the two-group existence statement that supplies sufficient conditions (e.g., sufficiently small interaction weights or sufficiently short horizon T) under which a standard comparison principle or a-priori L^∞ bound guarantees global solvability; we will also note that all numerical examples in the paper were computed with parameters for which the Riccati solution remained bounded on [0,T]. revision: yes

  2. Referee: [Section on general d heterogeneous groups] The ε-Nash verification for arbitrary d groups likewise rests on the same limiting Riccati solvability without additional justification that the coefficients (group-interaction weights and relative-average parameters) remain in the regime where global solutions exist.

    Authors: The same limitation applies to the general-d case. We will revise the ε-Nash theorem statement to make the standing assumption on global solvability of the limiting Riccati system explicit, and we will include a brief paragraph listing admissible regimes for the group-interaction weights and relative-average parameters that ensure the required global solution exists (again via comparison or smallness arguments). revision: yes

Circularity Check

0 steps flagged

No circularity: existence claim rests on external solvability assumption for limiting Riccati system

full rationale

The paper states that equilibria existence 'is guaranteed by the solvability for the coupled Riccati equations as the number of banks goes to infinity in each group' and verifies ε-Nash for d groups. This is an explicit modeling assumption on the limiting ODE system rather than a self-definitional loop, a fitted parameter renamed as prediction, or a load-bearing self-citation. No equations are shown to reduce to their own inputs by construction, no uniqueness theorem is imported from the authors' prior work, and no ansatz is smuggled via citation. The finite-N to mean-field passage and the Riccati-driven strategies remain independent of the target existence result; solvability is treated as a prerequisite condition, not derived from the equilibria themselves. The derivation chain is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The paper relies on standard assumptions from mean-field game theory and stochastic control applied to finance; no new entities or fitted parameters are mentioned in the abstract.

axioms (2)
  • domain assumption The evolution of log-capitalization is described by a system of coupled diffusions based on relative average of log-capitalization.
    This is the foundational modeling assumption for the interbank system.
  • domain assumption Banks minimize heterogeneous linear quadratic costs to avoid approaching the default barrier.
    This defines the optimization objective in the game.

pith-pipeline@v0.9.0 · 5775 in / 1341 out tokens · 32891 ms · 2026-05-25T01:52:30.895840+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.