The QRM volatility lessons specify how conditional variance evolves. This short bridge asks the different econometric question: when do ARMA residuals justify a conditional-variance model, and what does inference remain valid for when the fitted innovation density is not actually Gaussian?
By the end of this note you should be able to:
- run and interpret Engle’s ARCH LM test;
- state the quasi-likelihood object and the conditions behind QMLE consistency;
- distinguish consistent parameter estimation from naive Gaussian standard errors;
- select robust covariance inference for a fitted GARCH specification.
In this lesson
- Contents
- The big picture
The full lesson is for members.
Read the full lessonNew to SelfMadeQuant.com? See membership options or browse the Econometrics and Financial Time Series syllabus.