ARCH and GARCH

Modeling Volatility Dynamics

Modeling Unequal Variability
Equal Variability: Homoscedasticity

Unequal Variability: Heteroscedasticity
Means any variability (around the mean)
that is not homoscedasticity
Models must be developed for specific
cases
What These Acronym Mean?
ARCH
Autoregressive Conditional
Heteroscedasticity

GARCH
Generalized ARCH
Information in e
2
Let c
t
have the mean 0 and the variance o
t
.

Let e
t
be the residual of a model fitted.

Then:
e
t

estimates c
t

e
t
2
estimates the variance o
t
2
.



ARCH Modeling of o
t
2
.


ARCH(1)



ARCH as AR(1) on


2 2
( 1) t t
o = oc

= +
2 2
( 1) t t t
c = oc v

= + +
2 2
t t t
c o v = +
GARCH
GARCH(1)



GARCH (1) as ARMA(1,1) on


2 2 2
( 1) ( 1) t t t
o = oc |o

= + +
( )
( )
2 2
( 1)
1
t t t
t
c = o | c v |v

= + + +
2 2
t t t
c o v = +
Asymmetry in GARCH - TARCH
TARCH(1,1)
2 2 2 2
1 1 1 t t t s t
d o e oc c |o

= + + +
d = 1 if c
t
< 0, and = 0 if c
t
> 0
0
t
c >
Asymmetry in GARCH - EGARCH
EGARCH(1,1)
( )
2
2 2
1 1
1
1 1
0
log log
t
t t
t t
t t
o
c c
o e | o o
o o

>
= + + +
0 for asymmetriceffect =
Eviews Command


ARCH(p, q) series_name c