Check this page for more details -> investopedia.com/articles/07/ewma.asp
- It only works with (I made sure that the script plots the only when "D" time frame is selected)
- It doesn't plot the estimated variance, but the estimated std. dev. Compute the square of the std.dev to obtain the variance, if you need it.
Risk-metrics model is a very simple method to estimate the variance/std.dev. of daily log-returns. It doesn't require any parametric estimation technique (like OLS or Maximum likelihood) because it uses a fixed weighting parameter set to 0.94 (however you can change the parameter value from the settings window). The result can be used to implement more complex models for forecasting purposes. The recursive nature of this model also lets you extrapolate future values of the itself.
Please show me your support if you like my scripts. Hope you enjoy them.