| SØREN
JOHANSEN UNIVERSITY OF COPENHAGEN EMAIL:SØREN JOHANSEN |
ECONOMETRICS,
SPRING 2006This page will be edited during term |
 |
| SØREN
JOHANSEN UNIVERSITY OF COPENHAGEN EMAIL:SØREN JOHANSEN |
ECONOMETRICS,
SPRING 2006This page will be edited during term |
|
Date: |
StatØ3 |
|
Week 1 |
The lectures will be based upon the monograph S. Johansen (2006) Likelihood-Based Inference in Cointegrated Vector Autoregressive Models. Oxford University Press, Oxford (1996). The first week I shall give an overview of the theory from S. Johansen (2006) Cointegration: a survey, Chapter 14 in T.C. Mills and K. Patterson (eds.) Palgrave Handbook of Econometrics: Volume 1, Econometric Theory, Basingstoke, Palgrave Macmillan. The slides are in Slides The lectures held by S. Johansen Tuesday 10.15-12.00 Aud 7 (HCØ) Friday 10.15-12.00 Aud 5 (HCØ) The exercise classes are held by Kristoffer Laursen Tuesday 13.15-16.00, A101 (HCØ) OBS! The first time, Tuesday April 25, the exercises start at 14.15. The exercises will start with a discussion of Example 3.1 and 3.2 on page 37 and the definitions of I(0), I(1), and cointegration in Chapter 3. After that you can try to do the Exercises 3.1, 3.2, 3.5, 3.6, 3.7, 3.8 on pages 42 to 44 Computer exercises Friday 8.15 -10.00, Aud 05, but not on April 28. The program CATS2 is installed and available as an application in RATS.
The lecture for Friday 28 is based upon the Slides April 28 This week I started on a survey lecture and continued with a discussion of the Granger Representation Theorem |
Week 2 May 1 |
The topics for this week are 1. Chapter 5 on the interpretation of the I(1) model. and 2. Chapter 6 on the statistical analysis of the I(1) model. Exercises for Tuesday May 9 are 5.2, 5.3, 5.4, 5.6 This week we covered Chapter 5 and the role of deterministic terms. Friday we had a demonstration of RATS and CATS for RATS. |
Week 3 May 8 |
The topics for this week are 1. Chapter 6 on the statistical analysis of the I(1) model. Note: No lecture on Friday May 12. (Store Bededag.) A substitute session will be given on Wednesday June 14 by Kristoffer because by that time you will need to discuss exercises and I shall be at the NordStat2006 meeting in Rebild. Exercises for Tuesday May 15 are 6.1 (not question 6.1.2) 6.2 6.5 (not question 6.5.3) 6.4 in the form: Discuss the estimation problem for the models in exercises 5.2 and 5.6 |
Week 4 May 15 |
The
topics for this week are 1. Weak exogeneity, partial models and test on alpha, see Chapter 8. 2. Asymptotic properties of the process and product moments, see Chapter 10 and Appendix B I have been through Appendix B including section B.5.' Exercises for Tuesday May 23 are here Exercise May 23 Please note that you can hand them in to have them corrected. |
Week 5 May 22 |
The
topics for this week are 1. The rest of Appendix B 2. Asymptotic properties of the process and product moments, see Chapter 10 Exercises for Tuesday May 30 are here Exercise May 30 This week I discussed Appendix B.6 and B.7, and applied the results to the test for cointegrating rank in Chapter 11 for the model without lagged differences and deterministics |
Week 6 May 29 |
The
topics for this week are 1. The modifications of the rank test for lagged differences and deterministics (Chapter 11 and 12) 2. Asymptotic properties of the estimator for the cointegrating relations (Chapter 13) Exercise June 6 This week we discussed the modifications of the rank test and how to determine the cointegrating rank from the tables. We also showed consistency of the estimators of \alpha and \beta and found the asymptotic distirbution of the estimator of \alpha. |
Week 7 June5 |
The
topics for this week are 1. Niels Framroze Møller will present a paper on "Linking Economic theory models and the cointegrated VAR- Some simple illustrative examples" I then continue with 2. Asymptotic properties of the estimator for \beta and the mixed Gaussian distribution. An extra exercise is found in Extra Exercise I have essentially finished the topics in the book that I wanted to discuss. This friday I discussed the consequences of "running a regression" against doing maximum likelihood in two simple examples, where difficiult inference is a consequence of unmodelled correlation between error terms or due to ignoring lack of weak exogeneity. |
| Week 8 June 12 |
There
is an extra lecture on Friday June 16, 10-12 in Aud 5 This last lecture I shall discuss various exercises, including the Extra Exercises posed in week 7 Solutions of Extra Exercise |