[go: up one dir, main page]
Scribd
Upload
63 views2 pages

Hasts211 W8y23

This document is a worksheet for a course on Time Series Analysis and Econometrics at the University of Zimbabwe. It contains three main questions that explore the spectral density function of stationary time series models, including derivations, evaluations for specific models, and comparisons with white noise. The questions require students to demonstrate understanding of autocorrelation functions, sketch normalized spectral densities, and analyze the characteristics of time series plots.

Uploaded by

mashiringwaninyasha2003
Copyright
© © All Rights Reserved
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
63 views2 pages

Hasts211 W8y23

This document is a worksheet for a course on Time Series Analysis and Econometrics at the University of Zimbabwe. It contains three main questions that explore the spectral density function of stationary time series models, including derivations, evaluations for specific models, and comparisons with white noise. The questions require students to demonstrate understanding of autocorrelation functions, sketch normalized spectral densities, and analyze the characteristics of time series plots.

Uploaded by

mashiringwaninyasha2003
Copyright
© © All Rights Reserved
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

UNIVERSITY OF ZIMBABWE

TIME SERIES ANALYSIS AND ECONOMETRICS HASTS211

WORKSHEET 8: LD/2023

Attempt ALL Questions

Q1. The spectral density function f (ω) of a stationary time series model having autoco-
variance function {γk } (for k = 0, 1, 2, . . .) may be written as
∞
1 X
f (ω) = γk e−iωk , for − π < ω < π
2π k=−∞

(a) Stating any general properties of the autocorrelation function that you assume,
show that f (ω) may be written equivalently as
∞
( )
γ0 X
f (ω) = 1+ ρk cos ωk , for − π < ω < π
2π k=1

where ρk is the autocorrelation function of the time series.


(b) Deduce that f (ω) = f (−ω), for −π < ω < π.
(c) Evaluate f (ω)
(i) for a white noise at with variance σa2
(ii) for the time series Yt = at − 0.6at−1
(d) In each case (i) and (ii), sketch the normalised spectral density, g(ω) = 2πfγ0(ω) , for
o < ω < π and comment on what the shape of this function tells you about the
characteristic of the series.

Q2. The spectral density function f (ω) of a stationary time series model having autoco-
variance function {γk } (for k = 0, 1, 2, . . .) may be written as
∞
1 X
f (ω) = γk e−iωk , for − π < ω < π
2π k=−∞

2πf (ω)
(a) Show that the normalised spectral density, g(ω) = γ0
, may be written equiv-
alently as
∞  
ρk e−iωk + eiωk
X
g(ω) = 1 +
k=1

where ρk is the autocorrelation of the time series.


(b) Deduce that g(ω) = g(−ω), for −π < ω < π.

page 1 of 2
HASTS211

(c) For an AR(1) time series model

Zt = φZt−1 + at

(i) Show that


ρk = φk , k = 0, 1, 2, . . . [6]
(ii) Show that
1 − φ2
g(ω) =
1 − 2φ cos ω + φ2
(d) In the case φ = −0.7
(i) Sketch the g(ω) for 0 < ω < π and compare it with the spectral density
function of white noise.
(ii) Comment on what the shape of g(ω) tells you about how a time series plot
of a realisation of Zt would compare to a white noise.

Q3. The spectral density function f (ω) of a stationary time series model having autoco-
variance function {γk } (for k = 0, 1, 2, . . .) may be written as
∞
1 X
f (ω) = γk e−iωk , for − π < ω < π
2π k=−∞

(a) Stating any general properties of the autocorrelation function that you assume,
show that f (ω) may be written equivalently as
∞
( )
γ0 X
f (ω) = 1+ ρk cos ωk , for − π < ω < π
2π k=1

where ρk is the autocorrelation function of the time series.


(b) Deduce that f (ω) = f (−ω), for −π < ω < π.
(c) Show that for a white noise series at with variance σa2 , f (ω) is a constan.
(d) Calculate ρk for the time series Yt = at − θat−3 ,, where |θ| < 1 and show that its
spectral density function is

σa2 n o
f (ω) = 1 + θ2 − 2θ cos(3ω)
2π

page 2 of 2

You might also like