Master seminar "Forecasting in a Data-Rich Environment" (Summer 2024)

Instructor:        Prof. Dr. Ralf Brüggemann

Important Dates

April 12, 2024:                              Kick-off meeting 8.15 in room E405

June 28 and 29, 2024:             Seminar presentations by students 08:15-18:30 in H303

September 30, 2024:                Students hand seminar paper (15 pages)

Time and Location

The seminar will be held as a block seminar. There is a kick-off meeting on April 12, 2024 (8.15, room E405). The seminar presentations take place on June 28 and 29, 2024 in room H303

Description

The seminar focuses on economic forecasting applications using recent econometric techniques for data-rich environments. You will work on a topic and technique tailored to (high-dimensional) macroeconomic and/or financial time series data with many potential predictors. Depending on your choice this will involve dimension reduction methods (factor models), variable selection and shrinkage approaches or other machine learning methods.

The learning objective of this seminar is to provide you with an overview of recent advances in econometric techniques in the area of time series forecasting. Students should demonstrate their ability to understand and apply econometric techniques.

The work on the seminar topics typically includes reading, understanding and summarizing the main techniques and results from a paper assigned to you.

In addition, we ask you to empirically implement a small subset of the methods from the paper in your own illustrative empirical application. This includes some data preparation and programming using e.g., Python, R or Matlab. The specific nature of the empirical part has to be discussed on an individual basis.

Finally, you present and discuss your project in class and write a summary in form of a seminar paper.

Seminar papers are due on September 30, 2024. Seminar presentations and papers are in English. More information will be given during the kick-off meeting

Prerequisites

This is a seminar for Master’s students. Participants must have knowledge in econometrics (similar to “Econometrics” and “Advanced Econometrics” at the UKon). Ideally, students already have some background in time series econometrics. If not, it is highly recommended to also attend the lecture on “Applied Time Series Analysis”. Some knowledge in using and programming econometric software (e.g. Python, R or Matlab) is recommended

Assessment

Grading is based on the presentation (30%) and the seminar paper (70%).

Organizational details

You have to register for the seminar via the obligatory online seminar registration form. For details see webpage of the Master’s program.

If you need any additional information, please contact me directly at ralf.brueggemann@uni-konstanz.de

Seminar topics:  will be offered to the seminar participants.