Child pages
  • Information for foreign students
Skip to end of metadata
Go to start of metadata

Back to Pekka Pere's home page
Back to Econometrics pages

Teaching and exams of time series analysis and econometrics at the department

The core of econometric teaching at the department is composed of three courses:

  • Stationary time series: Chapters 1-5 (intermediate level, 8 credits)
  • Multivariate time series: Chapters 7-11 (advanced level, 10 credits)
  • Nonstationary time series: Chapters 15-20 (advanced level, 10 credits).

The chapters refer to James Hamilton's book Time series analysis (available from the bookstore at the ground floor of the Exactum building, or at the time of writing this at a discount price from Amazon). The courses are in general lectured in Finnish, but foreign students are welcome and can present questions in English. The examinations are written in English if requested.

Alternatively, one can take an exam in the general examinations of the department. One should first contact the examiner, professor Pentti Saikkonen, to make sure the date is fine (and for potential advice on the exam). To register for the exam fill the registration form in the department office in room C329 in Exactum or the 2nd floor office at Unioninkatu 37 (two weeks beforehand please!).

The Stationary time series course can be taken at an advanced level (10 credits) by having Chapter 6 examined, too, or by carrying out a small empirical exercise and writing a short report of it. The topic has to be agreed with professor Saikkonen beforehand.

University lecturer Pere examines two courses at the general examinations:

  • Second course on regression analysis and
  • Econometrics.

Literature for the exams is Russell Davidson's and James MacKinnon's Econometric Theory and Methods. Chapters 1-6 should be read for the regression analysis exam at the intermediate level (10 credits). By reading additionally chapters 7 an advanced level grade (10 credits) is given. The econometrics exam is at the advanced level (10 credits). It comprises chapters 8-12 and 15 (and Chapter 7 if not covered by the aforementioned examination). Second course on regression analysis is lectured on a regular basis. Please check the list of the lectures given each term from the study pages of statistics and mathematics.

All of the courses and exams assume knowledge of statistical inference and regression analysis at the intermediate level.

Advice about the requirements at the general examinations

Stationary time series (Chapters 1-5 of Hamilton's book):

  • The section on repeated eigenvalues in Chapter 1 need not be studied (pp. 18-19).
  • The main outcome of Section 2.5 are equations (2.5.18)-(2.5.20).
  • Ergodicity considerations will not appear in the exam.
  • Invertibility of an MA(1) process should be understood but invertibility of an MA(q) will not be asked about in the exam.
  • If you feel that section 4.4 (triangular factorization) is heavy going you are not alone with your feelings. The derivation should be studied but is not asked for in the examination. Application of the factorisation to forecasting of an MA(1) process (pp. 95-98) is for reference only.
  • This summary relates to p. 111.
  • A not so rare exam question is to ask the student to explain the presented two methods of estimating either of an AR(1) or of an MA(1) model.
  • Newton-Raphson method is the core idea of Section 5.7.
  • Section 5.9 can be skipped.

Multivariate time series (Chapters 7-11 of Hamilton's book):

  • Mixingales in Chapter 7 may be hard to digest. Your knowledge about them will not be inspected in the exam.
  • Proofs on the asymptotics of the t and F statistics (p. 212-213) capture many essential aspects of asymptotics, and the proofs have been asked for in the exams. Remember to carefully explain all the steps in the proofs by referring to appropriate theorems. Likewise, layout of a proof for the asymptotics of the OLS estimator for the parameters of an AR(p) model has been asked for in the exams.
  • Concepts of instrumental variables, 2SLS, simultaneous equations bias and identification are important topics (Chapter 9).
  • It is a good idea to prove for oneself the laws Kronecker products obey (p. 265, say). During the course this is done in the exercises.
  • Section 10.4. (spectrum) need not be studied.
  • VARs (Chapter 11) are much used nowadays and are an important topic.Section 11.7 (standard errors of impulse responses) is supplementary material only, though.
  • No labels