ECON452 (Queen's, Winter 2012)

Applied Econometrics

Important Announcements/News

  • Apr. 19: You can check some examples of good final reports in the section with the same name below.
  • Apr. 19: I am done marking final reports. You should receive your final grade as well as some comments by email soon. Take some time to look at the general comments about the final paper in the section below.
  • Apr. 11: I have received all the final papers from students registered in class. I have begun grading them.
  • Mar. 27: Both the homework and the intermediate report can be collected at the distribution center (Dunning, 3rd floor).
  • Mar. 20: I am done marking the homework. You will receive your grade by email tonight. Some general comments can be found below.
  • Mar. 19: The solution to the homework can be found below.
  • Mar. 14: You can find below the marking scheme for the final paper.
  • Mar. 14: Vincent's office hours are now on wednesdays, from 10h00 to 11h30.
  • Mar. 9: I am done marking the mid-reports. You should have received an email with your grade and some comments about your paper.
  • Feb. 29: I have a added a recurrent question about the first and second report. Here is the link .
  • Feb. 29: Vincent's office hours are cancelled this week. They will resume the following thursdays from 13h to 14h30.
  • Feb. 29: Some of you have requested information about LaTeX. Here is an introduction to it. In no way this is mandatory for this class. However, if you intend to pursue in academia, it is a good thing to learn it.
  • Feb. 25: I have included a FAQ about the first report. You can find it here.
  • Feb. 15: I added a simple example of Maximum Likelihood Estimation problem in the readings below.
  • Feb. 9: I have put online a paper that we will cover in class once we get back from the reading week. Please read it.
  • Feb. 8: Revised slides for week five are now availlable below. I have made some changes regarding the ADF test. We will cover this during next week's tutorial.
  • Feb. 7: My office hours changed from thursdays after class to tuesdays after class (11h30 to 13h00). Feel free to send me an email anytime, though.
  • Jan. 17: Revised slides for week 2 are availlable below. Some minor errors were corrected. I have also added stars (*) to persons without partners in the page about groups. This might help if you are still looking for a partner.
  • Jan. 16: The department's dropboxes are actually open until 15:00 for homeworks. All homeworks deposited after this time will be placed in a recycle bin after this hour.
  • Jan. 12: Groups for laboratories have been made. You can look at the list here (with some likely misspell of your name).
  • Jan. 11: I have put online some revised notes for week one. Tomorrow morning, I ask for your preferences regarding laboratories groups.
  • Jan. 10: I have heard of only one conflict for holding tutorials on wednesdays 10:00. Since the group will be split in two, we will keep this period. (Note that there is no tutorial session this week.)
  • Jan. 10: The homework and reports should be submitted in the drop-boxes on the second floor of Dunning. Keep in mind that the building/drop-boxes is/are open from 8:30 to 15:00, so you should plan accordingly.
  • Jan. 9: Both the homework and the first set of lecture notes are availlable online.
  • Jan. 3: Students can choose to work in pairs for their reports. The group is likely to be split in two for teaching laboratory sessions. Students willing to do so should thus rapidly find a partner so they can be in the same session.
  • Jan. 3: The Syllabus is now availlable below.

Syllabus

This course is intended for persons willing to perform meaningful time series analysis after they graduate. The semester will be divided in two periods. In the first one, there will be classes about the theory of time series as well as applied examples. Tutorials will cover an introduction to a statistical software. In the second period, Students will have to apply theory to a project of their own (and there will thus be no formal classes).

There will be no midterm or final exam. Most of the grades will be given according to a final essay based on students projects as well as an intermediary report. Some additional grades for class attendance and an homework will be given.

Course Slides

  1. Week one (revised).
  2. Week two (revised).
  3. Week three.
  4. Week four (and perhaps five, depending on how it goes).
  5. Week five (revised).
  6. Week six.
  7. (reading week).
  8. A paper that we will discuss in week 8 (click here for an annotated version) and some lecture notes.

Laboratories Files

  1. First laboratory: an introduction to Stata. (Note that ".do" files can be opened with any text editor.)
  2. Second laboratory: (first do-file), (second do-file).
  3. Third laboratory: use the dataset for tutorial 3 below. My do-file and my log-file (pabsta).
  4. Fourth laboratory: some discussions about what should be in a report.
  5. Fifth laboratory: we will deal with stationarity issues on the GDP and the CPI/TSX datasets (below): do-file for the GDP (pabsta), do-file for the TSX.
  6. Sixth Laboratory: Forecasting, building nice graphs and nice tables.: do-file. Vincent also use the command estout which is similar to outreg in principle.

Homework (due March 16th)

  1. The solution to the homework can be found here.
  2. Some general comments:
    1. It was well done overall. The average is 7.3. The Standard Deviation/Min/Max are respectively 0.8/6/8.7.
    2. Everybody needs to review how to perform a one sided test.
    3. Even though, the 1.96*s.d. works in an ARIMA setup (we assume errors are white noises), it is not always a good approximation in an OLS framework. When the sample size is small the difference is too big to use this approximation. But most importantly, this is not how Stata computes the confidence region when using OLS.
    4. When applying OLS to the model estimated in question 2, it is meant to regress the process on its past values. Since the model is an AR(1,2,3,5), this means performing the following command:
      reg process L.process L2.process L3.process L5.process, nocons.
    5. A lot of you wrote question 2 as if no information was given about the process. This leads to useless analysis as we know, for instance, that the process as no moving average component.

Intermediate Report

  1. Some general comments can be found here.
  2. An example of a good intermediate report (submitted by Macklem & Lyman).
  3. Another example of a good intermediate report (submitted by Gasser & Power).

Final Report

  1. The grading scheme for the final report is the following:
    Item Marks
    Use of methodology seen in class (including forecasts) 35
    Dataset description 5
    Formulation of an hypothesis based on the litterature review and the dataset 5
    Overall Look & Readability (in particular: nice graphs/tables and a meaningful abstract) 5
    Total: 50

    Please note that a literature review is still expected. You will be marked on your ability to form an hypothesis based on it and your dataset. Keep the best of the papers you found in your first report.

  2. Some general comments about the final paper can be found here
  3. An example of a good final report by Richards and Yu.
  4. An example of a good final report by Karlov and Xi.

Some Datasets Used

  1. Dataset for question 2 of the homework.
  2. An AR(1) process: someAR1.dta.
  3. Canadian Business Cycles (.csv file)
  4. Dataset for the second part of tutorial 2.
  5. Dataset for tutorial 3.
  6. Consumer Price Index Dataset, monthly.
  7. Seasonally Adjusted Gross Domestic Product, quarterly.
  8. Toronto Stock Exchange, Closing Value, Monthly Average

Some Useful Readings

Copyright of Course Material

The material on this webpage is copyrighted and is for the sole use of students registered in ECON452 at Queen's University. The material on this webpage may be downloaded for a registered student’s personal use, but shall not be distributed or disseminated to anyone other than students registered in ECON452 at Queen's University. Failure to abide by these conditions is a breach of copyright, and may also constitute a breach of academic integrity under the University Senate’s Academic Integrity Policy Statement.