Quality-Adjusted Hit-Ratio Targeting in Corporate Bond Market Making: Credit Alpha, Client Flow Quality, and Style-Aware Warehousing

Battle Count: 77

Directly addresses a core operational problem in electronic corporate bond market making: how to set RFQ quotes under client-service KPI constraints while managing adverse selection. The quote decomposition formula (riskless spread + inventory skew - credit-alpha skew + residual-toxicity charge - quality-hit-ratio subsidy) is immediately actionable for fixed-income trading desks. The dualization technique preserves tractability while changing the economic objective, making it implementable in production systems. The style-aware warehousing extension connects market making to alpha generation through inventory recycling.

The core algorithm requires: (1) factor model estimation for credit covariance, (2) markout decomposition to identify residual toxicity, (3) solving a one-dimensional nonlinear fixed point per tier per quote state (Newton/bisection), (4) Riccati matrix computation for multi-bond inventory curvature, (5) logistic fill curve inversion via Lambert W function. The separability of the dualized HJB keeps the problem tractable, but the multi-bond factor model, quality weight estimation, and integration with existing RFQ infrastructure add practical complexity. The style-aware warehousing extension adds a second participation-sizing layer.

The paper provides a detailed numerical algorithm (Section 8), exact dual fixed-point equation, logistic fill curve closed-form quote map, and quadratic/Riccati approximation. Synthetic simulation design is described (20-bond EUR IG universe, 300 MC paths, Newton iterations for dual solve). However, no code repository is linked, and the synthetic universe generation details (sector/rating/maturity attributes, factor covariance) are not fully specified. Appendix B provides a replication roadmap with recommended path counts and reporting standards.