Time series analysis

[Gia Hao]

Hi all,
I do a summary about time-series models and hope you all could give me suggest or any advices. Thanks all !
1. Stationary data --> Arma (p,q)
2. Non-stationary --> Arima (p,d,q)
3. Heteroscedasticity (Arch effect) but effect of residual is symmetric (the negative or positive shock produce the same volatility)
--> Arch or Garch model, and the Conventional Mixture ARMA-GARCH model
4. Volatility clustering, asymmetric volatility (non-normal), leverage effect/ inverse leverage effect
---> Asymmetric Garch-family model ( SGacrh, GJRGarch, QGarch, NGarch, EGarch or the combined versions: Arma-SGarch, Arma-GJRGarch ... and Arma-Garch-M, Arma-NGarch-M... )
Hope you all could give me suggestion !

[Roberto Liebscher]

Dear Gia,

If you would add a precise question to your post I am sure one of the members here will be able to help you. If your are concerned with applying these models using Stata, Sean Becketti's "Introduction to Time Series Using Stata" might be a great resource. But you could also type help time in Stata which opens a comprehensive overview of Stata's time series capabilities.