Dear Murat Tuna

If Arch-Garch test appear significance where p-value =. 000 that is mean presence of heteroskedasticity problem you should examine residuals series if there is a period of high volatality followed by period of low volatality and so on. In this case you should estimate GARCH famlily model.

Dear Murat Tuna

If Arch-Garch test appear significance where p-value =. 000 that is mean presence of heteroskedasticity problem you should examine residuals series if there is a period of high volatality followed by period of low volatality and so on. In this case you should estimate GARCH famlily model.

It is better in this case to use ARIMA-GARCH models.

Hi,

You can think about several options.

1. Check whether the seasonality is present, if so you may need to use seasonal AR terms and seasonal MA terms in addition to non seasonal AR and MA terms (if the seasonality is stochastic). Or you can use seasonal dummies with AR and MA terms (if the seasonality is deterministic)

2. Try using more AR or MA terms

3. If the issue is not yet solved, use ARIMA(p,q)-GARCH(m,n), ARIMA(p,q)-GJR(m,n), ARIMA(p,q)-EGARCH(m,n) or other ARIMA(p,q)+other GARCH family models depending on the nature of data you have, one of these will fit your your data. Use AIC, or SIC to select the best model.

As you're using Eviews for your analysis, use AUTOMATIC ARIMA Forecasting feature available to identify AR and MA terms and SAR and SMA terms.

Kind regards

Thushara

the presence of GARCH effects means that you have to use GARCH specifications which takes into account heteroscedasticity (presence of volatility) in the residuals. Thushara is wright, Eviews is very simple and helpful in forecasting.

The data is daily so I cant check seasonality at Eviews.This is my thesis subject and ı notices at last time . So ıcant change model only I can fing a reference to continue .To use citation.