Determining Insolvency Regions in Banks: A Stochastic Dynamic Approach Integrating Liquidity and Credit Risk

Battle Count: 75

The paper is primarily focused on bank regulation, supervisory stress testing, and financial stability rather than trading strategies or market microstructure. However, the liquidity-credit spiral dynamics and insolvency boundary concepts could inform credit risk models used in fixed-income trading, bank equity valuation, and systemic risk monitoring relevant to macro trading desks. The surrogate function could potentially be adapted for real-time risk signal generation.

The core model requires solving a non-linear PDE (HJB equation) in three state dimensions (W, lambda, d) with inequality constraints (Basel III ratios) using Finite Difference Methods. The surrogate fitting involves non-linear least squares optimization over continuous exponents. Calibration requires granular balance-sheet data. The mathematical sophistication (stochastic control, KKT conditions, convexity proofs) and numerical PDE solving represent significant implementation challenges, though the surrogate function simplifies real-time deployment.

The authors provide an interactive web-based dashboard (https://insolvency-model.onrender.com) and a video demonstration in supplementary material. However, the underlying CODAL balance-sheet data from the Iranian banking sector is not publicly available in a downloadable format. The HJB numerical solution methodology is described but no code repository is explicitly provided. Parameter calibration details are partially specified.