Second Thoughts: How 1-second subslots transform CEX-DEX Arbitrage on Ethereum

Battle Count: 94

Directly addresses CEX-DEX arbitrage strategy, execution risk modeling, and the impact of confirmation latency on trading decisions. The risk-averse agent model with entry thresholds and fallback logic is highly relevant to quantitative trading strategy design. The finding that 535% more transactions and 203% more volume emerge under faster execution has direct implications for HFT firms operating cross-venue arbitrage. The paper quantifies how execution speed changes the viable opportunity set, particularly for marginal trades in low-fee pools.

The simulation framework requires: (1) OLS regression for CEX-DEX price reversion with interval-based recalibration, (2) Monte Carlo simulation with backward induction over nested loops (failed attempts × subslots), (3) noise trading distribution estimation from empirical data, (4) risk-averse utility computation at each decision node, (5) competitive agent interaction modeling. The mathematical formulation is clear but implementing the full backward induction with MC estimation across multiple configurations and fee tiers requires significant engineering effort.

The paper provides detailed parameter specifications (α=0.35, λ=0.01, θ=0, 16 MC paths, 300s reversion window, 3s wait horizon), explicit mathematical formulations for the agent model, price interpolation mechanism, and simulation design. However, no code repository is provided, and the specific data processing pipeline for Binance/Uniswap v3 data is not fully detailed. Robustness tests with parameter grids are described.