What are the modeling steps of GARCH model?

Time series modeling should start with stationarity test, and go through stationarity test (if multiple series are considered, co-integration test is needed), mean model (arima, etc. ) Start to test the residuals of the mean model, and if arch effect is found, establish Garch model for the residuals.

The full name of ARCH (Autoregressive Conditional Heteroscedasticity Model) is "Autoregressive Conditional Heteroscedasticity Model", which solves the problems brought by the second hypothesis (variance unchanged) of time series variables in traditional econometrics. GARCH model, called generalized ARCH model, is an extension of ARCH model and developed from Bollerslev( 1986).

Time series is a set of random variables sorted by time, which is usually the result of observing a potential process at a given sampling rate at equal intervals.

Time series data essentially reflects the changing trend of one or some random variables with time, and the core of time series prediction method is to mine this law from data and use it to estimate future data.

Composition: long-term trend, seasonal change, periodic change and irregular change.

1) Long-term trend (t) The general trend of changes caused by some fundamental factors in a long period of time.

Seasonal variation this phenomenon changes regularly and periodically with the change of seasons in a year.

3) Periodic change (c) The regular change of the fluctuation pattern of the phenomenon with a period of several years.

4) Irregular change (i) is an irregular change, including strictly random change and irregular sudden change with great influence.