Optimal Pricing and Hedging of SOFR Derivatives

Battle Count: 73

Highly relevant for quantitative trading desks dealing with SOFR derivatives. The paper provides a practical, computationally efficient framework (sub-1-minute pricing on a laptop) for pricing and hedging OTC SOFR derivatives in realistic incomplete markets with bid-ask spreads and liquidity constraints. The sparse hedging portfolios and explicit hedging error quantification are directly actionable for trading strategies. The sensitivity analysis with respect to risk aversion and transaction costs informs practical parameter calibration.

The mathematical framework (convex optimization, indifference pricing theory) is sophisticated but well-structured. Implementation requires: (1) a calibrated stochastic term structure model with jumps, (2) construction of payout matrices for hundreds of derivatives across thousands of scenarios, (3) solving a convex optimization problem via interior-point methods (MOSEK), and (4) a one-dimensional search for indifference prices. The use of standard tools (Python, NumPy, MOSEK) and the sub-minute computation time suggest moderate implementation complexity for a quantitative developer familiar with convex optimization.

The paper specifies hardware (Intel Core i7 @ 3.00 GHz x 8, 16 GB RAM), software stack (Python, NumPy, Pandas, Numba, MOSEK Optimizer), and computation time (<1 minute). The term structure model is referenced from [48]. However, no code repository is provided, and the specific Bloomberg quote data (28 August 2024 at 15:30:00) is not publicly shared. The methodology is well-described mathematically, enabling reimplementation.