Double exposure businessman and stock market or forex graph suitable for financial investment concept. Economy trends background for business idea and all art work design. Abstract finance background.

Bayesian Analysis for Central Bankers

With Professor Dimitris Korobilis

Key information

Adam Smith Business School Summer School on Empirical Macroeconomics

Programme dates: 6 - 10 September 2021

Location: Online


Early bird (before 10 August 2021): £600
Standard price (from 11 August 2021): £800
Alumni: £600

Course overview

You will develop an understanding of Bayesian methods relevant for the analysis of financial and macroeconomic time series. The emphasis throughout this course is on Bayesian estimation and computation, and specification of flexible models. You will cover several topics and examples from macroeconomics and finance.

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, you will delve deeply into estimation issues which are of practical importance for applied researchers and PhDs/interns. You will focus on the use of Bayesian methods to jointly model many time series with short numbers of observations (as is the case with quarterly macro data for most countries). At the same time, you will also focus on flexible econometric modelling of the Great Recession, the current pandemic, and beyond.

Who should attend?

This programme is designed for those with a basic knowledge of regression and time series models. In particular, you should be familiar with time series concepts at the level of: Brooks, C. (2019) Introductory Econometrics for Finance, 4th Edition, Cambridge University Press, or: Tsay, R. (2010) Analysis of Financial Time Series, Wiley Series in Probability and Statistics.

This course will be useful for central bank researchers and advanced PhD students engaged in macroeconomic modelling work and anyone who wants to develop their professional skills in macroeconomic modelling. It will be appealing to a wide range of analysts who want to improve their analytical and practical skills in Bayesian econometrics.

The course will focus on the Bayesian inference aspect of time-series models. As it won’t review basic time series concepts such as autocorrelation, or how to convert an AR into an MA and vice-versa, it will 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 their basic functional forms.

Computations are in MATLAB. You will be provided with all the code in a very accessible form, so that even those with no knowledge of programming can attend this class. There is an expectation you will have at least some basic MATLAB skills (for example, know how to estimate a regression with OLS using basic commands, such as " >> beta_OLS = X/Y ").


Learning outcomes

By the end of this course participants will be able to:

  1. estimate and programme high dimensional univariate regressions with more predictors than observations
  2. build and estimate VARs, possibly of very high dimensions, for a single country
  3. devise panel, global and/or factor VARs for joint modelling of time series of many countries
  4. specify extensions of all the above models using time-varying parameters, structural breaks, or stochastic volatility.

Readings and resources

Bauwens, L. and Korobilis, D. (2013). “Bayesian Methods”, in Handbook of Research Methods and Applications on Empirical Macroeconomics.

  • Korobilis, D. (2018). Machine Learning Macroeconometrics: A primer
  • 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.
  • Chan, J., Koop, G., Poirier, D. and Tobias, J. (2020) Bayesian Econometric Exercises, Cambridge University Press

Further references will be provided during the course.

What you will learn


Lectures: 9.00am - 10.45am BST

Labs: 11.15am - 1.00pm BST

Day 1

Lecture 1a: An overview of Bayesian Inference; basics of Bayesian computation; Inference in the Normal linear regression model

Lab 1: Inference in the linear regression; Bayesian Lasso

Day 2

Lecture 2: Bayesian VARs using analytical results; the Minnesota prior; Hierarchical VARs and Bayesian shrinkage

Lab 2: Measuring monetary policy using various priors for VARs

Day 3

Lecture 3: Panel VARs, Factor-Augmented VARs

Lab 3: Monetary policy evaluation using panel VARs and FAVARs

Day 4

Lecture 4: State-space models; Homoskedastic time varying parameter VAR

Lab 4: Unobserved Components Stochastic Volatility (Stock and Watson, 2007)

Day 5

Lecture 5: Time varying parameter (TVP) VARs; High-dimensional TVP-VAR; TVP FAVAR

Lab 5: TVP-VAR for measuring the transmission of monetary policy (Primiceri, 2005); a large TVP-FAVAR for extracting a Financial Conditions Index (Koop and Korobilis, 2014)


  • The course will be conducted using zoom, therefore participants will need a free zoom account with zoom app installed
  • Participants will need a stable and good broadband connection, a fairly new pc or laptop with a camera and microphone - one which is capable of running MATLAB code - and speakers/headphones
  • Matlab code and course material will be distributed to participants ahead of the course
  • Computer illustrations will be based exclusively on matrix programming language MATLAB. All participants need to make sure they download and install in their own computers a free, trial copy of MATLAB from Mathworks website

Programme leader

Professor Dimitris Korobilis

Dimitris Korobilis  is Professor of Econometrics at the Adam Smith Business School. Before joining Glasgow he was Professor of Finance at Essex Business School, University of Essex. He works in advanced statistical inference using economic and financial data. Recent research involves the development of machine learning algorithms for high-dimensional inference in macroeconomic models with large datasets; research that has been published in Journal of Econometrics and Journal of Business & Economic Statistics, among other journals. His models have been used extensively by policy-making institutions to monitor financial conditions (International Monetary Fund) and to forecast inflation (European Central Bank). He has been a consultant for major international institutions (e.g. US Department of Energy) and government (e.g. Scottish Government), and he delivers regularly specialised training in central banks on topics related to statistical inference that supports policy decision-making.

He is in the all-time top 4% of authors internationally in Economics on REPEC (among 55,000 registered economists), in the top 1% (position 290 out of 55,000) measuring output only for the past 10 years, and he was No 1 young economist in the UK (PhD 5 years or less).