Topics in Macroeconometrics

Contents and Goals

The course provides an introduction to structural vector autoregressive (SVAR) and limited dependent time series models and related econometric methods. After the course, the student should be able to critically evaluate empirical research based on these models and to implement them in practice. In particular, they should be familiar with (1) a wide range of different identification schemes for SVAR models and (2) different econometric models designed for limited dependent time series, be able to assess their adequacy and apply them in practice.

In addition to lectures, there are a number of small homework assignments, whose solutions are discussed during the lectures. The homework consists of both theoretical problems and empirical hands-on applications involving actual economic data. In class, we use the R software to demonstrate the practical implementation of some of the methods discussed, but any suitable software package can be used for the empirical homework assignments.

Lecture slides are posted on the Moodle website before each lecture. In some lectures an empirical application is discussed, and the related text is posted on the Moodle website before the lecture. The course is not based directly on any textbooks. However, the following may useful as references:

  • Hamilton, J. (1994). Time Series Analysis. Princeton University Press.
  • Lütkepohl, H. (2005). New Introduction to Multiple Time Series Analysis. Springer.
  • Martin, V., Hurn, S., Harris, D. (2013). Econometric Modelling with Time Series: Specification, Estimation and Testing. Cambridge University Press.
  • Verbeek, M. (2012). A Guide to Modern Econometrics. 4th Edition. Wiley.

The course consists of two separate parts, whose contents and learning goals are described in detail below.

Part 1

The first six lectures by Prof. Markku Lanne concentrate on the structural vector autoregressive model widely employed in empirical macroeconomics and finance. The following topics are dealt with:

  • Vector autoregression and impulse response analysis
  • Structural vector autoregression
  • Identification of the SVAR model by short-run and long-run restrictions
  • Identification by sign restrictions
  • Statistical identification

The theoretical results related to the identification and estimation of SVAR models and impulse response analysis are illustrated through a number of empirical applications drawn from the previous literature.

After the first part of the course, the student should

  • be familiar with the stationary and nonstationary VAR model,
  • know the common identification procedures in SVAR analysis, and
  • be able to apply at least some of the methods covered in empirical research.

Time permitting, finally the following issues are discussed:

  • Relationship between dynamic stochastic general equilibrium (DSGE) and VAR models
  • Non-fundamentelness in structural VAR analysis

After this provisional part, the student should also

  • have a basic understanding of the relationship between the SVAR and DSGE models and be aware of their pros and cons, and
  • understand the concept of non-fundamentalness and its significance for SVAR analysis.

Part 2

The remaining six lectures by Postdoctoral Researcher Henri Nyberg consider limited dependent time series models and their macroeconomic and financial applications. The plan is to consider the following models:

  • Binary response models
  • Unordered and ordered multinomial models
  • Count data models
  • Models for duration data
  • Models for non-negative variables

We concentrate on, in particular, the properties of the above-mentioned models and their empirical applications typically examined in the previous literature.

After the second part of the course, the student should

  • be familiar with various limited dependent models,
  • understand their main differences compared with the conventional (AR) models for continuous dependent variables,
  • have a basic understanding of their statistical properties (including statistical inference and goodness-of-fit measurement), and
  • be able to apply the models and related methods in empirical research.



The course builds upon the courses Advanced Econometrics I and Advanced Econometrics II, whose contents the student is expected to have good command of. In particular, familiarity with the basics of VAR models, unit root processes, cointegration and the vector error correction model as well as linear algebra is required.

This course can be taken on both the Master’s and doctoral levels.


The final grade of the course is the weighted average of the grades of

  • the final examination, and
  • the term paper

with weights 70% and 30%, respectively. Both are graded on a scale from 0 (fail) to 5, and in order to pass to course, an acceptable grade in both is required. The final examination is common to all students, but the requirements of the term paper are different at the Master’s and doctoral levels. The term paper instructions are posted on the Moodle area in due course.


See the study guide for the dates of the lectures and examinations. The deadline of the term paper will be posted along with the instructions on the Moodle area in due course.

Link to the Moodle page of the course