hi bill, one of the things I'm spending time on this year is developing some good volatility models,predictors, and correlators (for asset allocation), which is what brought me to this site in the first place(VIX). I'm not sure what level you're diving in to volatility measures and methods at, but I've found some interesting info on GARCH (generalized autoregressive conditional heteroskedasticity) methods as part of a package offered by Mathsoft. Have you looked into this at all? Some info is below:
"A single bad trade can wipe out a year's profits," said Richard Bohdanowicz, vice president and general manager of the MathSoft Data Analysis Products Division,"so measuring risk has become an exceedingly important issue in the financial world. Until recently, quants have not had good statistical models that predict future
volatilities of financial returns. S+GARCH, on the other hand, is rapidly adaptive to changing volatility, providing more up-to-date risk calculations for investment firms. Enormous amounts of money are at stake, and S+GARCH can give a financial organization a crucial edge."
Prior to S+GARCH, statistical models available to financial analysts didn't adequately take into consideration that volatility varies over time, and that it comes in clusters. But S+GARCH, with its ability to handle both univariate and multivariate models, brings real-world variance modeling to financial analysts doing options pricing, term structure modeling, asset-allocation modeling, risk management and asset pricing. For example, if a quant is trying to measure the volatility of a stock portfolio with several different stocks, each with its own volatility, S+GARCH can accurately model the time-varying correlations among the stocks as well as their individual volatility and provide the analyst with a single risk measurement for the portfolio.
dh |
