Scribd
Upload
123 views2 pages

Yule-Walker Equations and Moving Average Models

The document discusses the Yule-Walker equations and moving average models. It introduces the Yule-Walker equations as a set of equations that can be used to estimate the parameters of an autoregressive (AR) model from the autocorrelation function. It presents the equations in matrix form and describes how they can be used for parameter estimation under the method of moments approach. It then defines the moving average (MA) model and the moving average operator, presenting the MA model compactly in terms of this operator.

Uploaded by

sellary
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
123 views2 pages

Yule-Walker Equations and Moving Average Models

The document discusses the Yule-Walker equations and moving average models. It introduces the Yule-Walker equations as a set of equations that can be used to estimate the parameters of an autoregressive (AR) model from the autocorrelation function. It presents the equations in matrix form and describes how they can be used for parameter estimation under the method of moments approach. It then defines the moving average (MA) model and the moving average operator, presenting the MA model compactly in terms of this operator.

Uploaded by

sellary
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Yule-Walker Equations and Moving Average Models

7.1: Yule-Walker Equations 3.2 Moving Average Processes Homework 2c


Yule-Walker Equations
Start with the mean zero AR(p) model:
Z
t
=
1
Z
t 1
+
2
Z
t 2
+ +
p
Z
t p
+a
t
()
Multiply both sides of () by Z
t h
for h = 1, . . . , p:
Z
t
Z
t h
=
1
Z
t 1
Z
t h
+
2
Z
t 2
Z
t h
+ +
p
Z
t p
Z
t h
+a
t
Z
t h
()
Take expectations throughout:
(h) =
1
(h 1) +
2
(h 2) + +
p
(h p) (1)
Now take the expectation of () with h = 0:
(0) =
1
(1) +
2
(2) + +
p
(p) +E(Z
t
a
t
)
. .

2
( )
Rearranging ( ) gives

2
= (0)
1
(1)
p
(p) (2)
Equations (1) and (2) are the Yule-Walker Equations.
Arthur Berg Yule-Walker Equations and Moving Average Models 3/ 9
7.1: Yule-Walker Equations 3.2 Moving Average Processes Homework 2c
Y-W Equations in Matrix Form
From the recurrence
(h) =
1
(h 1) +
2
(h 2) + +
p
(h p) (1),
extract the p equations
_

_
(1) =
1
(0) +
2
(1) + +
p
(p 1)
(2) =
1
(1) +
2
(0) + +
p
(p 2)
.
.
.
(p) =
1
(p 1) +
2
(p 2) + +
p
(0)
_
_
_
_
_
(1)
(2)
.
.
.
(p)
_
_
_
_
_
. .
p
=
_
_
_
(0) (1) (p 1)
(1) (0) (p 2)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
(p 1) (p 2) (0)
_
_
_
. .
p
_
_
_
_
_

2
.
.
.

p
_
_
_
_
_
. .

Hence
p
=
p
where
p
= {(k j )}
p
j ,k=1
.
Arthur Berg Yule-Walker Equations and Moving Average Models 4/ 9
7.1: Yule-Walker Equations 3.2 Moving Average Processes Homework 2c
Estimation Using the Y-W Equations
Under the method of moments approach, we estimate
= (
1
,
2
, . . . ,
p
) with

p
1

p
and also

2
= (0)

1
(1)

2
(2)

p
(p)
= (0)


p
= (0)
_

p
1

p
_


p
= (0)
p

p
1

p
where

p
= { (k j )}
p
j ,k=1
and
p
= ( (1), . . . , (p))

.
Arthur Berg Yule-Walker Equations and Moving Average Models 5/ 9
7.1: Yule-Walker Equations 3.2 Moving Average Processes Homework 2c
Moving Average Model MA(q)
Denition (Moving average model MA(q))
The moving average model of order q is dened to be
Z
t
= +a
t
+
1
a
t 1
+
2
a
t 2
+ +
q
a
t q
where
1
,
2
, . . .
q
are parameters in R.
The above model can be compactly written as
Z
t
= + (B)a
t
where (B) is the moving average operator.
Denition (Moving Average Operator)
The moving average operator is
(B) = 1 +
1
B +
2
B
2
+ +
q
B
q
Arthur Berg Yule-Walker Equations and Moving Average Models 7/ 9
7.1: Yule-Walker Equations 3.2 Moving Average Processes Homework 2c
Homework 2c
Read 3.3 and 3.4.
Do exercise #3.1(a,b) and #3.5.
Arthur Berg Yule-Walker Equations and Moving Average Models 9/ 9

You might also like