Course descriptions for the winter term 2011/2012

040131 UK Einführung in die empirische Wirtschaftsforschung

UK, 2 Wochenstunden

Zeit und Ort:

Montag 14:30–16:00 Hörsaal 23, Hauptgebäude

Beginn: Montag, 3. Oktober 2011

Beschreibung des Kurses: Die Studierenden sollen in leicht fasslicher Form mit den Methoden und Begriffen empirischen ökonomischen Arbeitens bekannt gemacht werden. Als Grundlage dient das Lehrbuch von Ramu Ramanathan "Introductory Econometrics with Applications" (5th edition, South-Western). Das Buch enthält eine umfangreiche Sammlung von Daten und Anwendungsbeispielen. Leider ist es derzeit vergriffen, daher stelle ich diverse Kursunterlagen (Skriptum, Slides) zur Verfügung. Der Kurs bereitet auch auf das eigenständige empirische Arbeiten in Praktika vor.

Aufbau des Kurses:

Ökonometrisches Arbeiten, Streudiagramme, Modell, Parameter, Schätzen und Testen

Einfaches lineares Regressionsmodell (Kleinstquadrateschätzung OLS, R², t-Statistiken)

Multiples lineares Regressionsmodell (R², Modellauswahl, F-Statistiken, Multikollinearität)

Gebräuchliche Spezifikationstests (Durbin-Watson u.a.)

Leistungsfeststellung durch schriftliche Teilprüfung während des Semesters (midterm, 45%) und schriftlichen Abschlusstest (45%) am Ende des Semesters sowie durch Mitarbeit und Hausübungen (10%).

---

040787 UK Applied Time Series Analysis

UK, 2 hours per week

Time and location: Monday, 17:00–18:30 Seminarraum 2 Hohenstaufengasse 9 1.Stock

First meeting: October 3, 2011

Course description: This course focuses on time-series analytic methods that are empiricall relevant in current economics. Two main issues are:

Linear models for stationary variables (definitions of stationarity, correlogram, ARMA model, information criteria)

Models for difference-stationary variables (integration and cointegration, Dickey-Fuller test, Johansen procedure)

All methods will be illustrated through empirical examples and printouts obtained using econometric software.

The definition of the UK course requires the course grade to be based on two or more components. A suggestion would be to have a written test before Christmas (45% weight) and additionally ask participants to work out a small empirical project using time-series methods (55% weight). The empirical project may be elaborated in groups of up to three persons and can be presented in class in January. A written project report will be graded and should be in by January 31, 2012. The definitive organization will be convened in the first units and may also depend on the number of participants.

Participants are assumed to have some basic knowledge of econometric methods.

A time-series textbook popular with economists is: Hamilton: Time Series Analysis. A clear presentation of the basic issues is also contained in Brockwell and Davis: Introduction to Time Series and Forecasting (Springer).

---

390045 UK PhD-E: Vector autoregressive models

UK, 2 hours per week

Time and location:  Tuesday, 18:00-19:30, Seminarraum 2 (Hohenstaufengasse)

First meeting: October 4, 2011

Course description: The subject of this course are vector autoregressive models (VAR) that are an indispensable tool for current empirical macroeconomics. The course is mainly based on the important textbook by Helmut Lütkepohl: New Introduction to Multiple Time Series Analysis (Springer). In particular, the following aspects should be covered:

    (1) Stable vector autoregressions
          (a) Basics: stationarity, MA representation, autocorrelation function
          (b) Forecasting
          (c) Granger causality, impulse response
    (2) Estimation of models
    (3) Model selection, specification tests, information criteria
    (4) Cointegrated VAR models

These topics correspond roughly to the chapters 2-4 and 6 of the textbook, with chapter 3 (estimation) to be surveyed only.

The definition of the UK course requires the course grade to be based on two or more components. A suggestion would be to assign a weight of 45% to a written midterm test and 55% to the presentation of a book section that was not covered in the first part of the course or, alternatively, of a project using VAR models. Participants' presentations should be in the second part of the course after the test. The definitive form of this course will be convened in the first meeting and may depend on the number of participants.

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