Asymptotic equivalence of nonparametric diffusion and Euler scheme experiments (Q2510830)

The authors consider two statistical experiments: a continuously observed (on a time interval \([0,T]\)) diffusion with nonparametric unknown drift and its Euler scheme. Diffusions are supposed to have nonconstant known distribution coefficients. A global asymptotic equivalence of these two experiments is proved in the sense of Le Cam's theory by constructing equivalence mappings based on random time changes. As a consequence, the equivalence of the discretized observation of the diffusion and the corresponding Euler scheme experiment is derived. The results justify the use of the Euler scheme instead of the discretized diffusion process for inference purposes.