An Introduction to Advanced Techniques for Macroeconomic Forecasting

Topics:

  • ARIMA and dynamic regression forecasts with an excursus to nonlinear forecasts using the example of the IMK Konjunkturindikator
  • VAR forecasts
  • Bayesian VAR forecasts

Description:

These lectures aim to introduce the audience to advanced techniques for macroeconomic forecasting, and the underlying regression models - whether a single equation or multi-dimensional approach - to generate economically meaningful predictions. The research workshop builds on two chapters of the textbook Forecasting: Principles and Practice by Rob J. Hyndman and George Athanasopoulos. The authors provide a comprehensive introduction to forecasting methods and sufficiently in-depth information about each method to enable readers to run their own applications with the help of the statistical freeware R.

The research workshop expands the textbook content in two ways. On the one hand, non-linear real-time short-term forecasts are discussed using the example of the IMK business cycle indicator. On the other hand, lectures also cover the Bayesian VAR approach, as introduced by Litterman (1986), using the example of the most recent high inflation phase to review and extend the rather parsimonious VAR specification of Stock and Watson (2001).

The workshop uses R for most of the models presented. Participants learn how to perform a forecast analysis coded in R. R is free and available on almost every operating system.

About the instructor:

Thomas Theobald is senior economist for financial markets and business cycle analysis at the Macroeconomic Policy Institute (IMK) in the Hans-Böckler Foundation (HBS). He obtained a PhD from the Freie Universität Berlin in 2014 and studied mathematics (Dipl.-Math.) and economics (B.Sc.) at Goethe University Frankfurt before. His broad research interests include macro-financial modelling, forecasting and time series econometrics. Among others, his work has been published in the Journal of Economic Dynamics and Control, Cambridge Journal of Economics, and International Journal of Forecasting.  

Schedule:

Wednesday, 07. May 2025 (F21/03.48)

13:15h - 14:45h: ARIMA and dynamic regression with an excursus to nonlinear forecasts

Thursday, 08. May 2025 (F21/03.79)

13:15h - 14:45h: Vector Autoregressive forecasts with a special focus on a 3-variable system (inflation, interest rate and unemployment rate)

Friday, 09. May 2025 (F21/03.80)

13:15h - 14:45h: Bayesian Vector Autoregressive forecasts with a special focus on a high dimensional system