Dynamic Collateral Control for Permissionless Spot Perpetual Basis Trading

Battle Count: 56

Highly relevant to quantitative trading, particularly in the DeFi/crypto space. The paper provides a complete framework from static collateral allocation through dynamic band control to execution-aware implementation. It addresses practical questions of liquidation risk management, execution cost asymmetry, minimum trade sizing, and strategy capacity. The live deployment validation (ERC-4626 vault on Arbitrum with 1inch routing and Hyperliquid hedging) demonstrates real-world applicability. The asymmetric control architecture (solvency-driven lower boundary, economic upper boundary) is directly implementable. The finding that realized performance is predominantly explained by funding environment has important implications for strategy selection and venue choice.

Moderate-to-high complexity. The static control problem requires numerical solution of first-passage time equations (no closed-form solution). The dynamic layer involves state-dependent band control with asymmetric triggers. Execution layer requires integration with DEX routing (1inch RFQ), perpetual order books (Hyperliquid), and on-chain settlement (Arbitrum). However, the paper provides a clear operational algorithm (Table VI) with well-defined inputs and decision logic. The bootstrap robustness exercise and Monte Carlo simulation add computational requirements. Smart contract implementation (ERC-4626 vault) adds infrastructure complexity. The control rule itself is interpretable and rule-based rather than ML-driven, which aids implementation.

The paper provides extensive reproducibility: Jupyter notebooks, processed datasets, and scripts are publicly available on GitHub (Krestenko, 2026). Raw market data sourced from public Binance REST endpoints and Hyperliquid Info API (free, unauthenticated). Live execution sample reconstructed from on-chain transactions of public vault addresses. No proprietary datasets or restricted credentials required. The Fractal research framework is also publicly available. All calibration parameters, venue rules, and control algorithms are explicitly documented.