QUASI-INDIRECT INFERENCE FOR DIFFUSION-PROCESSES

We discuss an estimation procedure for continuous-time models based on discrete sampled data with a fixed unit of time between two consecuti ve observations. Because in general the conditional likelihood of the model cannot be derived, an indirect inference procedure following Gou rieroux, Monfort, and Renault (1993, Journal of Applied Econometrics 8 , 85-118) is developed, It is based on simulations of a discretized mo del. We study the asymptotic properties of this ''quasi''-indirect est imator and examine some particular cases, Because this method critical ly depends on simulations, we pay particular attention to the appropri ate choice of the simulation step. Finally, finite-sample properties a re studied through Monte Carlo experiments.