ESTIMATION OF STOCHASTIC VOLATILITY MODELS VIA MONTE-CARLO MAXIMUM-LIKELIHOOD

This paper discusses the Monte Carlo maximum likelihood method of esti mating stochastic volatility (SV) models. The basic SV model can be ex pressed as a linear state space model with log chi-square disturbances . The likelihood function can be approximated arbitrarily accurately b y decomposing it into a Gaussian part, constructed by the Kalman filte r, and a remainder function, whose expectation is evaluated by simulat ion. No modifications of this estimation procedure are required when t he basic SV model is extended in a number of directions likely to aris e in applied empirical research, This compares favorably with alternat ive approaches. The finite sample performance of the new estimator is shown to be comparable to the Monte Carlo Markov chain (MCMC) method, (C) 1998 Elsevier Science S.A. All rights reserved.