Who Restores the Peg? A Mean-Field Game Approach to Model Stablecoin Market Dynamics

Battle Count: 80

Highly relevant for stablecoin market-making strategies, arbitrage execution across primary/secondary venues, and understanding de-peg dynamics for risk management. The model provides actionable insights on when arbitrage capacity breaks down (non-linear threshold in primary friction κ_P), which is directly applicable to stablecoin trading strategies. However, it is more of a structural/equilibrium model than a direct trading signal generator. Useful for understanding market microstructure frictions, congestion effects, and optimal routing across venues during stress events.

High complexity due to: (1) multi-population LQ-MFG formulation requiring policy iteration with convergence checks, (2) GARCH volatility coupling with state-dependent parameters, (3) multi-venue routing with softmax allocation, (4) Differential Evolution calibration over high-dimensional parameter space, (5) exploitability computation for ε-Nash validation, (6) three-phase event splitting and regime-specific parameter estimation. Requires expertise in game theory, stochastic control, numerical optimization, and market microstructure.

Code is publicly available on GitHub (https://github.com/ANRGUSC/stablecoin-peg-mfg). Model parameters are fully specified in Table I. Calibration methodology is described in detail. Data source (Binance API) is publicly accessible. However, the Differential Evolution calibration process and specific event-phase splitting criteria could benefit from more detailed documentation. The LQ-MFG framework is well-defined mathematically.