• Identify a research question and discuss its relevance.
• Use a theoretical model from finance or economics to form hypothesis about the research question.
• Specify a suitable econometric model including the assumptions of the model.
• Collect the data necessary for conducting the analysis.
• Estimate the model using an appropriate method using SAS or R
• Discuss and solve any important problems encountered in relation to the analysis (e.g. heteroscedasticity for a cross-sectional data set).
• Perform hypotheses testing of both simple and more composite hypotheses.
• Report and interpret the results of the analysis clearly and effectively to a reader who does not have a technical background of econometrics, for example, the CEO of a firm.
• Conclude upon the analysis.

This course provides an opportunity for students to gain direct experience estimating a simple model from economics or finance. The course will introduce three widely used approaches in econometrics that extend skills learned in the basic course in Applied Econometrics at the Applied Economics and Finance cand.merc. line. Students will be required to perform an analysis that uses one or more of the techniques of the course. This course will be very useful to students who plan to undertake empirical analysis as part of their Masters Thesis.
 

The course will begin with a brief summary of basic econometric methods with a focus on practical issues such as data quality, recoding raw data and treatment of missing data. We will also review the important classical assumptions, emphasizing when empirical results can be interpreted as causal effects or correlations between variables.
 

The course will cover the following topics:
 

1) Binary outcomes– Outcomes that are of particular interest within corporate governance, finance and many other fields are often binary in nature. For example, one might want to study the factors that affect whether a firm will make a public offering. In this section, students will learn the different approaches to estimating models with binary outcomes: linear probability models, probit and logit models.

2) Program/policy evaluation using difference-in-difference estimation– Difference-in-difference (DD) estimators are a relatively easy way to estimate the causal impact of a program or policy on any outcome of interest. For example, one might estimate the effect of minimum wages on unemployment, whether free trade agreements affect stock market prices, or how anti-trust laws impact cartel behaviour?

3) Trending time series and cointegration –These methods are very relevant for students who plan to write their thesis within finance or more generally those who plan to use time series data. In this section of the course, students will learn how to work with data that has a time trend and how to test whether two time series are cointegrated. The methods learned in this section can be used to test various important hypotheses from economic and finance models.
 

Students will:
•become proficient in the use of SAS or R;
•develop the ability to communicate clearly and effectively in writing.

Main text book:
Either

Stock, J.H. and Mark W. Watson. Introduction to Econometrics,  3rd edition, ISBN 978-0201715958
or
Gujarati and Porter, Basic Econometrics, 5th edition. ISBN 978-0073375779
 

Relevant articles will be selected from each of the three main topics.