Introduction
This is
a course in applied time series econometrics for Ph.D. students that have
completed Econometrics I and II. There is no single text, but I will refer to
material in “Time Series Analysis”' by James Hamilton, and Fumio
Hayashi’s “Econometrics”.
Course
requirements
I will assign homework (20%), and there will be a take-home final
exam (80%).
Reaching Me
My office is 565 in the Economics Department; phone, 7-1570.
Office hours are by appointment.
Outline
1.
Linear Time Series (
a. ARMA Models
b. Lag operators
c. Auto Correlation and Autocovariance Functions
d. Prediction and Impulse Response Functions
e. State Space Representations
f. Stationarity and Wold’s Theorem
2. Estimation (Reading
a. Conditional/Unconditional Maximum Likelihood
b. The Kalman Filter
3. VARs (Reading
a. Identification
b. Estimation and Inference
c. Granger Causality
d.
Present Value Restrictions
4.
Spectral Representations (Reading
a. Fourier Analysis
b. Spectral Densities
c. Filtering
d. Band Spectrum Analysis
e. Estimation and Inference
f. The H-P filter
5. Unit Roots (Reading
a. Spectral Implications
b. Spurious Regression
c. The Beverage-Nelson Decomposition
d. Distribution Theory for I(1) processes
e.
Unit Root Tests
6.
Cointegration (Reading
a. Definition and Spectral Implications
b. Common Trends
c. Impulse Responses
d.
Estimation and Inference
7. Nonlinear Time Series
a. Markov Switching Models (Reading
b. Conditional Heteroskedasticity (Reading
c. Threshold Autoregressions