Course descriptions for the summer semester 2013



040027 Advanced Econometrics


UK, 4 hours per week (8 ECTS)


Language of instruction: English


Time and location:

Tuesday 11:00-12:30 Seminarraum 1 Hohenstaufengasse 9

Thursday 11:30-13:00 Hörsaal 26 Hauptgebäude


Starts: March 5, 2013


Course description: The course focuses on econometric methods whose knowledge is expected at the Master level, but which are not covered in the econometric core instruction. In particular, time-series methods such as unit-root testing and cointegration analysis are in focus, as well as some extension to panel data. Thus, the course material corresponds roughly to the sections 8 to 10 of the textbook by Verbeek: A Guide to Modern Econometrics (Wiley, 4th edition 2012). The methods are highlighted in empirical applications in Stata.


Plan of the course: The definitive plan will be convened in the first unit and may depend on the number of participants. A suggestion is that the course grade comprises two tests (35% each) and an empirical project (30%) that can be done in groups of up to three persons. The second test should be in early June, such that participants can focus on their project afterwards.



390022 PhD-E: Working with panel data


UK, 2 hours per week (4 ECTS)


Language of instruction: English


Time and location:

Friday 11:30-13:00 Prominentenzimmer Hauptgebäude, Tiefparterre


Starts: March 1, 2013


Course description: The lecture course surveys econometric techniques that are used on two-dimensional data sets, one dimension of which is time. Such data sets, generally named panels or sometimes longitudinal data, occur in macroeconomics, for example in inter-country comparisons, where the time dimension dominates, as well as in labor economics, where a short time dimension and a large cross-section dimension is the rule.


The following topics could be covered:


Models for panel data (fixed and random effects, one-way and two-way)

Estimation procedures for panels (LSDV or within-groups estimator, GLS)

Tests (poolability, Hausman test)

Heteroskedasticity and autocorrelation

Dynamic panels (Nickell bias, instrumental variable estimation)

Panel tests for unit roots


The methods are explained using empirical examples that use the by computer software Stata. As important literature related to this course, I would recommend the textbooks by Hsiao: Analysis of Panel Data (Cambridge University Press, 2003) and by Baltagi: Econometric Analysis of Panel Data (Wiley, 2005).


In line with the course form UK, the allocation of grades to participants is to be composed of several parts. The definitive form will be convened at the first meeting. It may also depend on the number of participants. In recent comparable courses, the course grade was a weighted average of a written test (40%) in May, and of a presentation of an advanced topic from the above list or of an independent empirical project on panel data (60%). For the latter point, joint work in groups of two or three participants is encouraged. Grades will be based on the written report of the presented project to be submitted by the end of the course. For literature presentations, the report may consist of elaborate presentation slides. For empirical projects, the report will indicate clearly the aim of the project and its potential conclusions. Printouts of software results or screen shots are not acceptable as reports.



390022 PhD-E: Non-linear Time Series Analysis


UK, 2 hours per week (4 ECTS)


Language of instruction: English


Time and location:

Tuesday, 14:00-15:30 Seminarraum 2 Hohenstaufengasse 9


Starts: March 5, 2013


Course description: While the supply of nonlinear time-series models is almost limitless, this course will focus on three parametric model classes that are used in empirical economics and are treated in the monograph by Fan & Yao: Nonlinear time series (Springer 2005, Chapter 4):


Threshold models (SETAR, threshold autoregressions)

ARCH and GARCH models (conditional heteroskedasticity)

Bilinear models


The first course units are devoted to a short introduction and repetition of linear time-series models, comparable to Chapter 2 of the Fan & Yao textbook, then the basic features of the three classes above are addressed. A written test should close this part of the course and carries 40% of the grade. The remaining 60% of the course grade are based on presentations by participants, with topics related to nonlinear time-series analysis. The material for these presentations can be taken from the book by Fan & Yao or from comparable monographs, in particular:


Teräsvirta, Tjostheim, and Granger: Modelling Nonlinear Economic Time Series (Oxford University Press, 2010)


Written versions of all presentations should be handed in by June 30, 2013.


Participants are assumed to have some basic knowledge of econometrics and time-series methods.