Can volatility in the stock market be modelled?

Michael Lo, in the Department of Statistics and Actuarial Science successfully complete his M.Sc. project on "Generalized autoregressive conditional heteroscedastic (GARCH) time series models."

While the title is a mouthful, his project uses a class of models that try to account for several features in the financial markets: high variability of daily returns, and periods of high variability followed by periods of low variability. He found that the Standard and Poor's 500 was fit reasonably well by this model. Alas, the model also shows that the past provides no new information on predicting the future value of the S & P Index.

This type of interdisciplinary work is a hallmark of our program in Applied Statistics at Simon Fraser University. For more information, please contact Michael Lo (shiu (at) stat.sfu.ca) or his supervisor Richard Lockhart (lockhart (at) stat.sfu.ca).