| Model | Conditional variance equation | Proposed by |
| IGARCH | σ2t=ω+αε2t−1+(1−α)σ2t−1 | Engle and Bollerslev [30] |
| EGARCH | ln(σ2t)=ω+αε2t−1+γ(|εt−1|−E(|εt−1|))+βln(σ2t−1) | Nelson [31] |
| GJR | σ2t=ω+αε2t−1+γI(εt−1<0)ε2t−1+βσ2t−1 | Glosten et al. [32] |
| APARCH | σδt=ω+α(|εt−1|−γεt−1)δ+βσδt−1 | Ding et al. [33] |
| CSGARCH | σ2t=qt+α(r2t−1+qt−1)+β(σ2t−1+qt−1) |
|
|
| qt=ω+ρqt−1+ϕε2t−1−σ2t−1 | Engle and Lee [34] |
| TGARCH | σt=ω+ασt−1(|εt−1|−η1εt−1)+βσt−1 | Zakoian [35] |
| AVGARCH | σt=ω+ασt−1(|εt−1−η2|−η1(εt−1−η2))+βσt−1 | Schwert and Seguin [36] |
| NGARCH | σδt=ω+ασδt−1(|εt−1|)δ+βσδt−1 | Higgins and Bera [37] |
| NAGARCH | σ2t=ω+ασ2t−1(|εt−1−η2|)2+βσ2t−1 | Engle and Ng [38] |
| FGARCH | σδt=ω+ασδt(|εt−1−η2|−η1(εt−1−η2))δ+βσδt−1 | Hentschel et al. [39] |
| FIGARCH | ϕ(L)(1−L)dε2t=α0+[1−β(L)]νt | Baillie et al. [40] |