72883 - Econometrics for Corporate Decisions

Learning outcomes

At the end of the course the student is able to understand the basic linear and non-linear regression methods (logit and probit models), useful for the analysis of cross-sectional data that can be support the decisions of economic agents operating in firms. In particular, the student is able to: i) understand and critically read the applications of the different methods in recent empirical literature; (ii) apply the acquired tools for the analysis of data using the econometric package STATA (in case campus license is available) or equivalent econometric software available free of charge (such as GRETL).

Course contents

1. Simple and Multiple Linear Regression Model on sectional data: theory and applications with STATA (in case CAMPUS license is available) or equivalent econometric software available for free (such as GRETL). Example of applications: a) trend of sales of a product in stores of a franchise chain and adherence to the format; (b) evaluation of the effectiveness of marketing actions.

2. Association or causal relationship? Critical discussion of the conditions that allow a random interpretation of the estimates of a regression model.

2. Introduction to estimation with the maximum likelihood method also through some applications, such as: verification of the presence of scams in the 'Game of Fortune'; evaluation of the difficulty of the exams in different appeals

3. Logit and probit models on sectional data: theory and applications with STATA (in case CAMPUS license is available) or equivalent econometric software available for free (such as GRETL). Example of applications: model the probability of choosing to buy a product between two brands.

Teaching material related to laboratory exercises will be made available to students through the University's e-learning platform.

The course considers as pre-requisites by students basic knowledge on the following topics: a) Typical formats of economic data (sectional data, time series, longitudinal data);b) random variable, distribution of a random variable, moments of marginal and conditioned distribution; population, parameters and random sampling; verification of hypotheses on the sample mean (first and second type error, significance level, p-value). Useful references for reviewing these topics can be found in any textbook introducing econometrics (e.g. Stock and Watson (2012; third Italian edition) chapters 1,2,3,4; Hill et al. (2008) chapters 1,2,3 and Appendix B.

For students without basic knowledge, the CLAMDA course will offer a pre-alignment course of Econometrics.

Readings/Bibliography

Wooldridge, J. (2016) Introductory Econometrics: A Modern Approach, 6e

R. C. Hill, W. E. Griffiths, G. C. Lim, (2011) Principles of Econometrics (4th edition, International Student Version), Wiley

Joshua Angrist and Jörn-Steffen Pischke (2009) Mostly Harmless Econometrics: an empiricist's companion

Joshua Angrist and Jörn-Steffen Pischke (2015) Mastering 'Metrics: The Path from cause to effect

Franses, P.H. and Paap, R. (2007) Quantitative Methods for Marketing Research

For students who attended the course in the Academic Year 2020-2021: please contact the instructor before buying a reference book

Teaching methods

Lectures involve the presentation of theoretical and applied issues of the various econometric methods. Applications are discussed in class and replicated during the practice sessions package STATA (in case campus license is available) or equivalent econometric software available free of charge (such as GRETL).

 

Tutor : Giovanni Righetto

Assessment methods

Final draft and discussion.

A written paper of up to 5 pages (including the list of bibliographic references and information on the output) must be presented. The paper must be delivered at least 10 working days before the discussion. During the oral discussion, methodological and applied in-depth questions may be asked (i.e. both demonstrations and definitions of the theory, and requests for additional analysis)

The characteristics and structure of the final project is defined by the teacher as below.

1. Motivation

The working group must define the objective of the empirical analysis (maximum 200 words)

2. Description of data and congruence with the objective of empirical analysis

The working group shall write a paragraph which: (a) clarifies the source of the data used; b) clarifies the characteristics of the data used (sectional vs longitudinal, survey vs census, etc.) and the adequacy of the same to the objective of the analysis

3. Dataexploration: descriptive analysis and documentation related to the data cleaning carried out (elimination of outliers, harmonization of data from different sources)

4. Set up, estimate and evaluate models suitable to answer the question. All groups are invited to use multiple linear regression models and, where appropriate and possible, also logit and probit models. If one of the two classes of models is not usable for empirical analysis, justification must be given in the paper in the section on the choice of tools for empirical analysis. In this section students must demonstrate that they know how to use appropriately adequate diagnostic tools (tests for the verification of hypotheses, properly defined; indices of goodness of adaptation and / or indicators of forecasting capacity)

5. Conclusion section reserved for the critical discussion of the results.

Other useful information

A. Data

Workgroups must find data for their project. The data must allow for exploration and empirical analysis and must be adapted to the objective of the project. They don't necessarily have to be very large data bases. The data must be shared at the time of delivery of the work.

The source of the data must always be clearly indicated: projects that do not comply with this feature will not be admitted.

Some potentially interesting data repositories and/or search engines are

Google's data set search engine: https://datasetsearch.research.google.com/

Search engine: Data Portals [http://dataportals.org/]

Search engine: http://opendatamonitor.eu/

File: Kaggle [https://www.kaggle.com/]

Archive: GitHub - awesomedata/awesome-public-datasets: A topic-centric list of HQ open datasets. [https://github.com/awesomedata/awesome-public-datasets]

Data that some journals make available to replicate the results of published articles and that you can possibly consider as a starting point for analysis (for example, updating the data and replicating the analysis on more recent data and / or from other countries and / or collecting similar data through surveys). These repositorys also provide code examples.

JOURNAL OF APPLIED ECONOMETRICS DATA ARCHIVE http://qed.econ.queensu.ca/jae/

AMERICAN ECONOMIC JOURNALS; examples

AMERICAN ECONOMIC JOURNAL:APPLIED ECONOMICS [https://www.aeaweb.org/articles.php?doi=10.1257/app.6.4]

AMERICAN ECONOMIC JOURNAL:ECONOMIC POLICY [https://www.aeaweb.org/articles.php?doi=10.1257/pol.6.3]

B. Code to replicate the analysis

The working group will have to share the code used for the analysis at the same time as the delivery of the paper and data.

C. Instruments

c1) The working groups must be provided by a minimum of 2 and a maximum of 4 students. The composition of the working groups (self-managed by the students) must be communicated to the teacher according to the deadlines indicated in class.

c2) The working groups must fill out a form (provided by the teacher, according to the deadlines that will be agreed in class) where they indicate the responsibilities and activities in which each member of the group is involved and the deadlines for the organization of the activities (including any live or online meetings)

c3) All groups are invited to use multiple linear regression models and, where appropriate and possible, also logit and probit models. If one of the two classes of models is not usable for empirical analysis, justification must be given in the paper in the section on the choice of tools for empirical analysis. In this section students must demonstrate that they know how to use appropriately adequate diagnostic tools (tests for the verification of hypotheses, properly defined; indices of goodness of adaptation and / or indicators of forecasting capacity)

D. Assessment

The final evaluation of the project will take into account the difficulty posed by the cleaning and preparation of the data, the originality of the project, the congruence between motivation and data and the level of critical and analytical skills documented by the empirical analysis performed by the students.

The final evaluation of the project will weigh the evaluation of the project and the oral discussion.

According to the indications of the council of the School of Economics and Management, the indications on the graduation of the grade are reported.

• insufficient <18

• 18-23 sufficient

• 24-27 good

• 28-30 excellent

• 30 and excellent praise

The course is part of an integrated course for which a single grade will be recorded. The final grade of the integrated course is represented by the arithmetic average of the marks of the two modules of the integrated course. The grade in the "Econometric Models for Business Decisions" module will contribute to the aforementioned average only if it is equal or it exceeds 18.

Teaching tools

Slides, teaching material, practice sessions using package STATA (in case campus license is available) or equivalent econometric software available free of charge (such as GRETL)..

Lectures involve the presentation of theoretical and applied issues of the various econometric methods. Applications are discussed in class and replicated during the practice sessions using the package STATA (in case campus license is available) or equivalent econometric software available free of charge (such as GRETL).

Software GRETL (available for free from the web): http://gretl.sourceforge.net/

Software STATA: available for students of the Department of Economics (CAMPUS license) and at the Computer Lab of the School of Economics and Management.

Office hours

See the website of Margherita Fort