The module will consider a range of techniques that can be used to forecast operating income, earnings, share prices, interest rate, foreign exchange rate, etc. Some of the techniques that will be considered are linear and no linear regressions with lagged variables, including ARIMA, ARMAX, ARCH (q) and GARCH (pq). Techniques of determining the existence of unit roots, random walks, trends and cointegration will be considered together with those that can be used to correct errors, forecast seasonal and cyclical series, measure accuracy of a forecast, test the stability and non-linearity of market returns on shares, bonds and foreign exchange rates, as well as stimulation. Market microstructure in developed and emerging markets as well as recent studies where these techniques were used, will also be considered.

By the end of the module the student should be able to:

• apply procedures for model-building in finance;
• test financial theories and forecast financial variables using real-world data;
• apply specifically the classical linear regression, limited dependent variable, time series analysis and simultaneous equations models to financial data;
• identify when the model’s assumptions are satisfied in practice and how to deal with violations;
• demonstrate comprehensive knowledge and understanding of the theory and application of univariate time series modelling and forecasting using ARMA and ARCH-GARCH techniques.