Time Series Econometrics
Economics 513
Fall 2009

Taught by: Chris Sims

Contact information

New information is in red.

Course Syllabus and Reading List

Link to materials from last previous version of the course.

The course meets 10:40-12:10 Monday and Wednesday in (room not yet determined).

Exercises, links to additional notes, announcements about the course, etc. will appear here.

Takehome Final Exam
The exam is due by Tuesday, 1/19 at 9AM. You can submit it either on paper or electronically. You can submit it on paper to Kathe Woodside in the department office, or in my box in the department office. If you submit it at a time much different from 9AM on 1/19, email me when you submit it.
Supplemental Notes
This directory contains notes that were used in class but not previously posted. Three of the four are very similar to the versions from previous years on the same topics. The SVAR notes are quite different from previous years' versions.
Notes on hidden Markov chains
Exercise on hidden chain regression due Wednesday 12/16
matlab code for the exercise
Further explanations about the exercise
Exercise on VAR forecasting, due 11/25
Useful code for the exercise: R; Matlab.
Notes on Wiener process and linear differential equations and on the frequency domain.
Exercise on the frequency domain. Due Monday, 11/16.
Data for the exercise
Notes on multivariate ARMA
Exercise on Wold decompositions, ARMA models, due Wednesday, 10/21
Notes on stochastic processes, conditional expectations, ARMA processes
Exercise on MCMC for a small AR2 model, due Monday, 10/11
Data and code examples for R
Data for Matlab (csv format)
The R code is only for evaluating the multivariate normal density and for the MCMC iterations. It does not include the code for evaluating the unconditional likelihood. Though it is only in R, it should be fairly easy to translate to Matlab.
Lecture notes on the Bayesian approach to inference.
We followed the first part of these notes in the first part of the first lecture. The later notes on the SNLM and on MCMC methods are for topics we will take up later.
Notes on the Kalman filter and smoother.
These notes include examples we did not discuss in class. They also contain "practice exercises", which are meant as study aids, not for you to turn in.
Exercise on the Kalman filter for seasonal adjustment, due 9/28.
For this exercise, you are asked to prepare graphs showing your results. Have your graphs on a USB stick for class 9/28. We will display and discuss some in class. You may work with others, but will be on your own if asked to discuss the results in class.
Code and data for the Kalman filter exercise:
R, Matlab
Reading and pencil-exercise assignment for 9/28.
Read sections 2.1-3 and chapter 3 in MacKay. Work through exercises 2.4, 2.5, 3.5, 3.6, 3.8-10, and 3.12. These are supposed (by MacKay) to take about 15 minutes each. You do not need to write up and hand in answers (most are available in the solutions sections of the book). You do need to be able to discuss the answer if called on to do so in class, so take some notes.
). You do need to be able to discuss the answer if called on to do so in class, so take some notes. tes.