Student Seminar Series – September 29, 2009
University of Minnesota
School of Statistics
College
of Liberal Arts

 

Generalized Autoregressive Conditional Heteroscedastic Time Series Models with Applications

Jia Liu

 Tuesday, September 29, 2009
3:45 PM
300 Ford Hall
Minneapolis, East Bank Campus

Refreshments at 3:15 PM
300 Ford Hall



Abstract

This paper gives a survery of the univariate GARCH models for financial time series with applications. ARMA models are reviewed at first. T hen, ARCH and GARCH models are introduced with specifications, important properties, estimation, and diagnostic evaluations. A more detailed analysis is implemented to the GARCH (1,1) models with forecasting. Moreover, the properties of the estimated parameters of the GARCH (1,1) models are explored by monte-carlo simulations. ARMA-GARCH models, GARCH in Mean models and EGARCH models are also covered briefly as an extension. As examples, Standard & Poor's 500 stock price index and IBM stock price are used to fitting different models.