California Institute of Technology

3/14/1996

3:45pm-5:00pm

In modeling a financial time series, one can validate the prediction of the (time-varying) mean by directly comparing the predition to the actual value of the series. To validate the prediction of the (time-varying) variance, however, one uses less direct methods such as the likelihood function. The value of the variance or volatility does not only provide an error bar on the mean prediction, but also is a tradable quantity in its own right, e.g., as in options and other derivatives.

In this talk, we evaluate the of the likelihood function as a means of training (in sample) and validating (out of sample) a volatility model. The fidelity, or lack thereof, is then used to modify and combine different models to come up with a better model for the actual volatility.

*Refreshments will be served immediately before the talk at 3:30pm.Hosted by Andreas Weigend.*

Department of Computer Science

University of Colorado Boulder

Boulder, CO 80309-0430 USA

webmaster@cs.colorado.edu

University of Colorado Boulder

Boulder, CO 80309-0430 USA

webmaster@cs.colorado.edu