Course Description

Instructor: Dimitris Korobilis, Professor of Econometrics, Adam Smith Business School, University of Glasgow, UK

Course aims and overview

The main aim of this course is to help develop an understanding of Bayesian methods relevant for the analysis of modern financial and macroeconomic time series. The emphasis throughout this course is on Bayesian estimation and computation, with emphasis on flexible modelling and machine learning inference for high-dimensional cases.

This short course will introduce a very large spectrum of time series models used in macroeconomics and finance. Instead of focusing on the theoretical time-series properties of these popular models, we will delve deeply into estimation issues which are of practical importance for applied researchers and PhD students.

By the end of this course the student should be able to:

1. Specify flexible regression models that account for nonlinearities, stochastic volatility, or models that allow flexible modelling of the whole density of the data (quantile regression; density regression)
2. Estimate models with more parameters that observations, be this a simple linear regression or a more complex multivariate model
3. Compute parameters using a variety of traditional (e.g. MCMC) as well as machine learning algorithms (e.g. variational Bayes)
4. Devise new models and algorithms in order to tackle novel empirical problems

Background

This course is appropriate for students who already possess experience in time-series methods, and they want to take their skills to the next level. The ideal profile is PhD students in applied macro/finance who want to enhance their research potential by adding new tools in their toolbox, or interns and researchers in central banks (or large organizations) who want to monitor large panels of data in a time series context.

Undergraduate-level knowledge of econometrics is essential, and any further knowledge of econometrics/statistics/probabilistic data science, would be beneficial. We will need to rely heavily on distributions such as the Normal, Bernoulli, Gamma, and Wishart so students should be familiar with the concept of a p.d.f., a c.d.f, and understand (but not remember by heart) their basic functional forms.

Computations are in MATLAB. I will provide all the code in a very accessible form, so that even colleagues with no knowledge of programming can attend this class. Nevertheless, people who are serious about using Bayesian econometrics in their research, are expected to have at least some basic MATLAB skills (e.g. know how to estimate a regression with OLS using basic commands, i.e. ” >> beta_OLS = X/Y “), although more experienced users will be able to keep up more easily with the fast pace of the course (Note: less experienced programmers will inevitably need to self-study after the course is over with the material that I will provide).

Readings and resources

• Korobilis, D. and Shimizu, K. (2022). “Bayesian approaches to shrinkage and sparse estimation”, Foundations and Trends® in Econometrics, 11 (4), pp. 230-354.
• Koop, G. and Korobilis, D. (2010). “Bayesian Multivariate Time Series Methods for Empirical Macroeconomics”, Foundations and Trends® in Econometrics, 3, pp. 267-358.
• Koop, Gary (2003) Bayesian Econometrics, Wiley.
• Koop, G., Poirier, D. and Tobias, J. (2007) Bayesian Econometric Exercises, Cambridge University Press
• Bauwens, L. and Korobilis, D. (2013). “Bayesian Methods”, in Handbook of Research Methods and Applications on Empirical Macroeconomics.