This paper deals with parametric inference for continuous-time stochastic volatility models observed at discrete points in time. We consider approximate maximum likelihood estimation: for the "k"th-order approximation, we pretend that the observations form a "k"th-order Markov chain, find the...