no. 17 · Wednesday 18 November 2026
Uniform spatial Schauder estimates for backward stochastic partial differential equations with random coefficients
Abstract
We prove spatial Schauder estimates for super-parabolic backward stochastic partial differential equations. All coefficients can be random and depend on time and space, and the estimates are uniform in probability and time. The key ingredient is a weighted conditional energy norm adapted to the parabolic scale of each frequency block. Together with heat-kernel bounds and coefficient freezing, it controls the gradient-martingale coupling without assuming a bounded martingale integrand. As an application, we establish Schauder solvability for semilinear stochastic Hamilton–Jacobi–Bellman equations.
- fields
- Probability, Analysis of PDEs, Optimization and Control