Market-Informed Valuation of GMMB Riders with Surrender Options under a Heston Stochastic-Local Volatility Model

Battle Count: 61

This paper is primarily focused on actuarial/insurance pricing rather than quantitative trading. However, the Heston SLV framework, local volatility calibration, and stochastic volatility modeling techniques are relevant to derivatives pricing desks. The Markovian projection and leverage calibration methods have applications in exotic option pricing. The paper's core contribution (model risk for path-dependent/early-exercise claims) is relevant to structured product desks but not directly to trading strategies.

The implementation requires: (1) a recombining trinomial tree with moment-matching and admissibility search for CIR dynamics, (2) a decorrelating transformation for correlated Brownian motions, (3) a conservative fully implicit finite-volume scheme for forward density propagation, (4) iterative leverage calibration via Markovian projection with relaxation, (5) backward finite-difference PDE solves with obstacle projection for American-style surrender, (6) boundary condition handling in transformed coordinates, (7) substepping logic based on drift indicators, and (8) low-density fallback mechanisms. The paper spans 41 pages with extensive appendix detailing every numerical step.

The paper provides extremely detailed numerical specifications including grid construction, tree algorithms, finite-volume schemes, boundary conditions, and parameter values. However, no code repository is provided. The market data (EURO STOXX 50 options, VSTOXX futures) are from Bloomberg and Eurex, which are proprietary. The synthetic LV surface formula is fully specified. Reproduction would require significant implementation effort but is feasible given the detailed appendix.