90349 - Structural Macroeconometrics

Learning outcomes

At the end of the course the student has acquired a comprehensive knowledge of the main identification and estimation methods which can be featured by Structural Vector Autoregressions (SVARs) in order to quantify the dynamic causal effects of macroeconomic structural shocks of interest including, among others, the monetary policy shock and uncertainty shocks. In particular, he/she is able to - analyze critically the implications of macroeconomic theories in terms of estimated impulse response functions, and to make inference on the identified dynamic causal effects; - apply SVAR analysis to Euro area and/or U.S. monthly/quarterly data by available econometric packages with the idea of replicating existing results or producing new ones.

Course contents

0. Pre-course bootcamp

   VAR, VMA, companion form, Cholesky, baseline IRFs

   The treatment of unit roots

1. What is a macro shock?

     Structural shocks, exogeneity, unanticipatedness,

      IRFs, causal interpretation

2. Why Cholesky is not enough ?

     DSGE-to-SVAR intuition, non-recursive structures,

     full impact matrix

3. Identification as a mapping problem

    Rothenberg approach, reduced form vs structural form,

    order/rank conditions

4. Point-identified SVARs

   B-model, A-model, AB-model, Maximum Likelihood estimation

5. Inference on IRFs

   Delta method (intuition only), bootstrap, bootstrap consistency,

   FEVD (intuition only)

6. Modern identification

   Sign restrictions, heteroskedasticity, non-Gaussianity

7. proxy-SVARs (SVAR-IVs) and LP-IVs

    LP vs SVAR, weak instruments 

       

 

Readings/Bibliography

- Slides provided by the teacher available on Virtuale

- Kilian, L. and H. Lutkepohl (2017), Structural Vector Autoregressive Analysis, Cambridge University Press.

- Lutkepohl. H. (2015), New Introduction to Multivariate Time Series Analysis, Springer

- Amisano, G. and C. Giannini (1997), Topics in Structural VAR Econometrics, 2nd edn, Springer, Berlin.

Teaching methods

Traditional classes and "virtual labs" (i.e. the students bring their laptops in the classroom with freely or Unibo licenzed econometric softwares installed).

In case the course will be held online the exam will consist in the development of exercises that will be sent online to the professor according to the rules that will be negotiated

Attending classes is crucial to fully understand the spirit of this course

Assessment methods

The exam aims to verfy that the student has achieved the basic ingredients necessary to quantify the impact of macroeconomic shocks on the macroeconomy by SVAR methods.

More in detail, the students is supposed to have acquired:

- the knowledge of VAR models as key tools to capture dynamic properties of macroeconomic variables;

- methods to address the identification problem implied by the SVAR methodology;

The student is also supposed to carry out independent empirical work.

For 2026/2027 the exam consists in one of the following options:

(i) oral presentations of papers or topics of interest by groups of 2 student, no longer than 30 minutes;

(ii) addressing an econometric problem of interest of the teacher. The topic will be identified by the end of the course;

(iii) writing a short paper on a project (related to the topics covered during classes) assigned by the teacher.

Artificial Intelligence (AI) may be used as a valuable tool to support independent study by providing additional explanations, summaries, and self-assessment activities. However, during examinations (option (i) above), the use of AI is strictly prohibited. Any use of AI during the examination constitutes a violation of academic integrity.

For options (ii) and (iii) AI can be used for coding and editing. 

Grades of the form XX/30 are given. Overall, the meaning of grades is as follows

<18 failed
18-23 sufficient
24-27 good
28-30 very good
30 e lode excellent.

Should the course be held online (because of Pandemic issues, floods, etc.) the exam will still consist in writing a short paper on a project (related to the topics covered during classes) assigned by the teacher.

Teaching tools

Software used are:

Gretl wich open source and is freely downloadle from the web

Matlab for which Unibo has a licence which means that students can download and freely install it on their laptops, etc.

Links to further information

https://sites.google.com/site/lucafanelliunibo/home

Office hours

See the website of Luca Fanelli