Time Series Cointegration: Econometric Analysis Course
1Summary
Most economists eventually run into the same wall: applying standard regression to economic time series that aren't stationary produces results that look convincing but are actually spurious. This course from the Arab British Fellowship Training Academy is built to get participants past that wall.
Participants build an initial understanding of the technical characteristics shared by time series data, work through the concept of stationarity and what it means for analysis, and gain the modelling techniques needed when a unit root is present in the data.
2Objectives and target group
Who Should Attend?
- Researchers and analysts in development economics, general financial and tax policy, social and health economics, financial markets, and international trade and finance.
Course Objectives:
By the end of the course, participants will be able to:
- Understand and apply non-stationary time series analysis.
- Explain the concept of stationarity and apply unit root testing.
- Apply advanced econometric techniques to cointegration processes and error correction models, particularly in a multivariate context.
Participants sit an open-book test on the last day of the course and receive a certificate upon successful completion.
3Course Content
Module 1: Stationarity and Data Generation
- Fixed (stationary) versus non-fixed (non-stationary) time series.
- Data generation processes behind time series data.
Module 2: Cointegration Foundations
- Overview of non-linear common (cointegration) analysis.
- Common integration in uniform (simultaneous) equations.
- Multivariate integration.
Module 3: Correction and Response Modelling
- Error correction models.
- Response function and contrast (variance) analysis.
Module 4: Applied Cointegration Techniques
- The Johansen joint integration methodology.
- The self-regression (vector autoregressive / VAR) model.