Hands-On Machine Learning for Algorithmic Trading

How to build a volatility-forecasting model

The development of a volatility model for an asset-return series consists of four steps:

1. Build an ARMA time series model for the financial time series based on the serial dependence revealed by the ACF and PACF.
2. Test the residuals of the model for ARCH/GARCH effects, again relying on the ACF and PACF for the series of the squared residual.
3. Specify a volatility model if serial correlation effects are significant, and jointly estimate the mean and volatility equations.
4. Check the fitted model carefully and refine it if necessary.

When applying volatility forecasting to return series, the serial dependence may be limited so that a constant mean may be used instead of an ARMA model.

The arch library ...