Autoregressive conditional heteroskedasticity
Error variance in a time-series model that depends on the size of past errors, so volatile periods cluster. Where present, standard errors are unreliable, but the variance itself can be modelled and forecast.
Where you will meet it
Autoregressive conditional heteroskedasticity appears in 3 learning outcomes on the CFA syllabus. A few of them:
- Explain how residual autocorrelations let you check whether an autoregressive model suits a time series
“If none is significant, the model passes this check, though stationarity and ARCH still need testing.”
- Say what autoregressive conditional heteroskedasticity (ARCH) is and show how ARCH models help forecast a time series' variance
“Autoregressive conditional heteroskedasticity means the variance of a model's error in one period depends on the size of past errors, so calm and turbulent periods cluster.”
- Choose a suitable time-series model for a given investment problem and give reasons for the choice
“Then test for ARCH and correct if needed.”
See also
Practise it
Practise autoregressive conditional heteroskedasticity until it is automatic: questions that explain every wrong option.
Take the free diagnostic →