Modelling the Probability Distribution of Stock Price Changes: Financial Time Series

The book is a survey of major probability models of interest to academics and practitioners working in finance modelling. The book borrows models that were uniquely applied to forecast averages in time series analysis and applied them to the volatility of stock price changes. Intervention analysis of major world crises such as the oil crises of 1973-1974 and the market crash of the 1987 indicate that volatility responds differently to each crisis in terms of the impact and the lasting effect. The major argument of the thesis is that the assumption of stationary financial relationships is only valid in the short term. Models that use long time series of financial data should be used carefully so that periods of crises do not affect the inference about the financial variables relationship. The thesis also considers other important aspects of modelling financial time series such as volume-volatility relationship, financial leverage effect, seasonality, and non normal probability distributions.