This research will develop and apply a new approach to modeling the variability of economic and financial time series. The suggestion is to combine the models of autoregressive conditional heteroskedasticity developed in the early 1980s with time series models involving (Markov) switching processes. Combining these approaches offers promise of capturing more realistically the time series properties of dramatic economic events such as the stock market crash, financial panics or major changes in monetary or fiscal policy. Being able to characterize these events in a coherent time-series framework may prove to be critical in understanding the behavior of financial markets.
