Which of the following is true about ARCH and GARCH models?
(I) They are used for modelling and forecasting volatility
(II) They are non-linear models
(III) They can both be estimated using OLS
(IV) Series estimated using these models must have a unit root process
A) I only
B) I and II only
C) I, II and III only
D) I, II, III and IV

Question

Quizplus · Accepted Answer

ARCH (Autoregressive Conditional Heteroskedasticity) and GARCH (Generalized Autoregressive Conditional Heteroskedasticity) models are indeed used for modeling and forecasting volatility, making them particularly useful in financial time series analysis where volatility clustering is common. They are non-linear in the sense that they model the variance of the current error term as a function of the past squared errors (ARCH) or both past squared errors and past variances (GARCH), which introduces non-linearity. However, they cannot be estimated using Ordinary Least Squares (OLS) due to this non-linearity; instead, methods like Maximum Likelihood Estimation (MLE) are used. Additionally, the requirement for a unit root process is not a prerequisite for ARCH or GARCH models; they are designed to model volatility regardless of the presence of a unit root in the series.

Which of the Following Is True About ARCH and GARCH