Discretely observed diffusions: Classes of estimating functions and small \(\Delta\)-optimality (Q2722307)

The main purpose of this paper is to discuss criteria for choosing good estimating functions when estimating the parameters of an ergodic diffusion that is observed only at discrete points in time. Ergodic diffusions in several dimensions depending on an unknown multivariate parameter are considered. For estimation, when the diffusion is observed only at finitely many equidistant time points, unbiased estimating functions leading to consistent and asymptotically Gaussian estimators are used. Different types of estimating functions are discussed and the concept of small \(\Delta\)-optimality is introduced to help select good estimating functions. Explicit criteria for small \(\Delta\)-optimality are given. Some exact optimality conditions are presented as well as, for one-dimensional diffusions, methods for improving estimators using time reversibility. The conditions are presented in generality for multidimensional diffusions depending on multidimensional parameters, allowing for any combination of parameter dependence in the drift and diffusion coefficients, and apply to any flow that is known to yield consistent and asymptotically Gaussian estimators for arbitrary \(\Delta\).