Relevance
9/10
Highly relevant to quantitative trading, particularly for FX market makers and algorithmic execution desks. The paper provides actionable insights on optimal pricing strategies in the presence of internal liquidity, execution thresholds, and the trade-off between P&L optimization and client service. The identification of passive market impact mechanisms and the demonstration that optimal strategies significantly outperform naive benchmarks (by ~10-15% in P&L) are directly applicable to real-world trading systems. The framework bridges theoretical market making with practical internal exchange management.
Implementation Complexity
8/10
The HJBQVI formulation is mathematically sophisticated, combining stochastic control with optimal stopping. Numerical solution requires careful handling of the quasi-variational inequality structure, multiple state variables (time, price, inventory, cash, order size), and the execution region boundary. The backward Euler scheme with multiple order sizes and Poisson jump processes adds computational complexity. However, the dimensionality reduction via the ansatz v(t,s,q,x,l) = x + qs + h(t,q,l) simplifies the problem. Real-time implementation would require efficient PDE solvers and careful calibration.
Reproducibility
3/5
The paper provides detailed model specifications, parameter calibrations (kappa, lambda, phi, alpha, psi values), and describes the numerical scheme (backward Euler). However, the anonymised HSBC GBPUSD trade data used for calibration is not publicly available. The HJBQVI numerical solver implementation details are limited. The model structure is fully specified mathematically, enabling independent replication of the theoretical framework.