• Evaluation of macroeconometric models
• Forecasting of uncertainty and forecasting for policy analysis
• Properties of time series data and model design
• Dynamic specification and the use of vector auto-regression models (VARs) and error correction models (VECMs)

Module 1: EViews Basics (Optional)
Review of the main EViews commands to manage data.

Module 2: Introduction to Forecasting with EViews
Introduction to the EViews model simulator to estimate and forecast multiple equation models.

Module 3: Statistical Properties of Times Series Data
The concept of stationarity is defined as well as how to test for it. Box-Jenkins (ARMA) methodology to study time series is introduced.

Module 4: Forecast Uncertainty and Model Evaluation
How best to choose between forecasts from competing models or sources. Participants will learn the main forecast evaluation statistics and how to calculate them in EViews.

Module 5: Vector Auto-Regressions (VARs)
Understand VARs, how they used for forecasting and structural analysis, and how to estimate a well-specified VAR and generate forecasts.

Module 6: Cointegration and Vector Error Correction Models (VECMs)
Define and understand the concept of cointegration among unit-root variables and its implications for forecasting. Learn how to test for cointegration using the Johansen method and how to estimate and forecast using a VECM.

Module 7: Evaluating Regressions Models
What does it mean to have a “good model” (model evaluation and key model assumptions) and the consequences for forecasting. Introduction to model testing and dealing with error irregularities and structural breaks.

Module 8: Final Assignment: Bringing It All Together
An overview of the techniques studied is provided using a case study focused on private saving-consumption behavior in the U.S. before and after the global financial crisis.