Forecasting methods in economics and finance

August 22-24, 2012

Title: Forecasting methods in economics and finance

Location: Department of Economics and Business, Bartholins Allé
Organized by CREATES

  • Lecturer: Allan Timmermann, University of California and CREATES
  • Course description:
    This course will introduce forecasting methods in the context of the econometric analysis of financial and economic time-series. How can we build prediction models for interest rates, stock returns, GDP growth and other variables? Answering this question entails addressing issues such as how to select the functional form and predictor variables of the prediction model, and how to estimate any unknown model parameters. If a variety of forecasts are available, are there gains from combining such forecasts, rather than trying to identify a single best model? This will be addressed in the context of forecast combination analysis. Moreover, can we even tell if some models produce significantly better forecasting performance than others? This leads to the forecast evaluation techniques. Illustrations of the techniques will be provided through a variety of empirical examples.
    Knowledge of regression methods, statistical analysis and matrix algebra is a prerequisite for this course.

  • Course program
  • Download lecture notes (zip)
  • Registration (closed for registration)
    Price: 0 DKK for participant from AU, KU, CBS or SDU, and Nordic universities outside Denmark; 300 Euro for other participants.