Paper: Who Restores the Peg? A Mean-Field Game Approach to Model Stablecoin Market Dynamics
Authors: Hardhik Mohanty, Bhaskar Krishnamachari
Date: arXiv:2601.18991v2, 7 May 2026
Estimated Reading Time: 15 minutes
The paper develops a dynamic mean-field game framework to model how fiat-collateralized stablecoins recover from de-peg events. The model represents interactions between two agent populations, arbitrageurs and retail traders, across primary mint/redeem channels and secondary trading venues. By solving for equilibrium behavior, the framework derives price recovery dynamics and order flows endogenously rather than imposing them through statistical assumptions. The authors calibrate the model to three historical de-peg episodes involving USDC and USDT and show that it reproduces observed recovery patterns and estimated half-lives. The analysis attributes peg restoration primarily to arbitrage activity through primary redemption channels when those channels remain operational. Sensitivity testing reveals a non-linear threshold where rising primary-market frictions substantially slow recovery regardless of secondary-market liquidity conditions. The framework is proposed as a tool for evaluating stablecoin resilience, infrastructure design, and systemic risk under stress scenarios.
Core insights
- Primary-market arbitrage dominates recovery: The model finds that peg restoration is primarily driven by arbitrageurs using mint and redeem facilities rather than by exchange trading alone. When redemption channels remain accessible, arbitrage activity creates the largest peg-reverting flows during stress events.
- Infrastructure quality determines stability: Recovery speed depends strongly on the cost and accessibility of primary redemption rails. Higher redemption frictions reduce arbitrage participation and weaken the mechanism that links market prices back to the $1 peg.
- Non-linear breakdown threshold exists: Peg recovery remains relatively fast across a broad range of market conditions until primary-market frictions reach a critical range. Beyond that threshold, recovery half-lives increase sharply and market stabilization becomes significantly slower.
- Secondary liquidity is supporting mechanism: Exchange liquidity and trading depth influence recovery outcomes but are not the primary determinant when redemption infrastructure functions normally. Secondary-market conditions become more important only after primary channels become constrained.
- Equilibrium modeling explains market flows: The mean-field framework derives aggregate trading behavior from agent optimization rather than fitting price dynamics alone. This allows decomposition of peg-restoring flows by participant type and trading channel while maintaining equilibrium consistency.
The paper examines stablecoin stability through the interaction of supply adjustment mechanisms and trading incentives. Arbitrageurs can expand or contract circulating supply through primary-market minting and redemption, while retail traders influence prices through secondary-market trading. By explicitly modeling both groups, the framework links observed price movements to underlying incentive structures. Rather than assuming recovery behavior, the model derives it from optimization problems faced by market participants. A central tokenomics implication is that supply elasticity depends on operational access to redemption infrastructure. When a stablecoin trades below peg, arbitrageurs can acquire tokens in secondary markets and redeem them at par value. This mechanism effectively reduces circulating supply and creates upward price pressure. However, how much redemption activity remains profitable when settlement delays, fees, or congestion increase? The model suggests that the answer depends on a threshold effect rather than a gradual decline. The calibrated results indicate that primary-market friction is the dominant variable governing recovery speed. Across the examined de-peg events, elevated redemption costs were associated with longer recovery periods and weaker arbitrage participation. Secondary-market execution costs also matter, but their impact remains limited while redemption channels function efficiently. This finding implies that reserve quality alone may not determine stability if operational access to reserves becomes constrained. The paper also highlights the role of market composition. The relative share of arbitrageurs and retail traders influences how quickly price deviations are corrected. A higher concentration of arbitrage capital generally improves stabilization because more participants can exploit deviations from peg. What happens when arbitrage capacity becomes limited during a period of market stress? The simulations suggest that recovery slows substantially because fewer agents are available to absorb mispriced supply and execute redemption trades. Another notable contribution is the decomposition of peg-restoring flows. In the May 2022 USDT event, recovery was attributed largely to primary-market redemptions. In the March 2023 USDC event, impaired redemption infrastructure reduced the effectiveness of primary channels until operational conditions improved. These results imply that observed price recovery may reflect different underlying mechanisms even when similar price outcomes occur. From a tokenomics perspective, the paper frames stablecoin stability as a coordination problem between supply adjustment infrastructure and trader incentives. Peg robustness emerges when arbitrage remains economically attractive and operationally feasible. The identified non-linear threshold suggests that resilience cannot be evaluated solely through reserve backing or exchange liquidity metrics. Instead, stablecoin stability depends on whether redemption systems retain sufficient capacity during periods of elevated demand for conversions, allowing arbitrage-driven supply contraction to restore equilibrium efficiently.
