||This is a PhD level course that covers time series analysis. The goal of the course is to present modeling techniques for time series data. We consider models for time varying means, time varying variances, and time varying correlations. Specifically, we begin with ARMA models including interpretation, estimation, and forecasting. These models are extended to models for time varying means of vector processes. Next we move to models for time varying variances and variance covariance matricies. Two popular approaches are considered here; DCC models and factor models. Tests for non-stationary time series are presented and implications are explored including cointegration and error correction models. Time permitting; we will discuss models for high frequency data including models for (intertemporally) irregularly spaced data.
- ARMA models
- Maximum Likelihood Estimation and Inference
- Forecasting and Forecast Evaluation
- Vector Autoregressions (VARs)
- GARCH Models for Time Varying Volatility
- DCC and Factor models for Time Varying Variance Covariance Matricies
- Unit Roots and Time Trends
- Modeling techniques for High Frequency Data
We will use the full version of the Eviews software. I have arranged for special prices for the course. After you have registered for the course I will send you a form that needs to be filled out and signed by me to give you the special prices. Here is the webpage: http://www.eviews.com/
Grades will be determined by homework (15%), a midterm (35%) and a final exam (50%).