Optimal strong convergence rate of a backward Euler type scheme for the Cox-Ingersoll-Ross model driven by fractional Brownian motion (Q2309581)

The goal of this paper is to give the first results on strong convergence rate for the numerical approximation of the Cox-Ingersoll-Ross (CIR) model driven by fractional Brownian motion with Hurst parameter \(H\in (1/2,1).\) To deal with unbounded diffusion coefficients, the Lamperti transformation is used to obtain a more convenient auxiliary equation. Then for the considered fractional Brownian case, the authors provide the Malliavin derivative of the exact solution and the boundedness of the inverse moments of the exact solution. Using these a priori estimates, it is proved that the backward Euler scheme applied to the auxiliary CIR model (as well as the corresponding numerical approximation to the initial CIR model) converges with order one in the uniformly strong sense and ensures the positivity of the numerical solution. Numerical experiments are presented to enlighten theoretical results.