Department of Electrical Engineering Systems, Tel Aviv University,

Tel Aviv, Israel

School of Mathematical Sciences, Monash University,

Victoria, Australia

We propose an adaptive algorithm for tracking historical volatility. The algorithm borrows ideas from nonparametric statistic. In particular, we assume that the volatility is a few times diferentiable function with a bounded highest derivative. We construct the Kalman filter-type estimator, which guarantees the well known from statistical inference asymptotics relative to the sample size n. A tuning procedure of this flter is simpler than for GARCH flter.

Back to the Workshop's Program