ESTIMATING THE PARAMETERS OF STOCHASTIC DIFFERENTIAL-EQUATIONS BY MONTE-CARLO METHODS

We propose a method for the simultaneous estimation of the drift and d iffusion coefficients of stochastic differential equations (SDE) from panel data. The method involves matching the distribution of the exper imental/field data with a panel of simulated data generated by a Monte Carlo experiment. The fit between the two distributions is assessed b y means of the chisquare goodness-of-fit statistic leading to a confid ence function computed from an incomplete gamma function. A numerical optimisation algorithm then optimises the choice of parameters to maxi mise this function. Preliminary evidence is presented which suggests t hat it is possible to estimate the coefficients of the generating SDE very accurately.