0% found this document useful (0 votes)
3 views2 pages

null

The document is a take-home assignment for a course on Econometrics: Time Series Econometrics at Lilongwe University of Agriculture and Natural Resources. It includes several questions related to time series concepts such as stationarity, ergodicity, VAR models, cointegration, and ARCH processes, with specific tasks and calculations required for each question. The assignment emphasizes the use of a reference book and has a deadline of July 4, 2022.
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
0% found this document useful (0 votes)
3 views2 pages

null

The document is a take-home assignment for a course on Econometrics: Time Series Econometrics at Lilongwe University of Agriculture and Natural Resources. It includes several questions related to time series concepts such as stationarity, ergodicity, VAR models, cointegration, and ARCH processes, with specific tasks and calculations required for each question. The assignment emphasizes the use of a reference book and has a deadline of July 4, 2022.
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

LILONGWE UNIVERSITY OF AGRICULTURE AND NATURAL RESOURCES

BUNDA COLLEGE OF AGRICULTURE

DEPARTMENT OF AGRICULTURAL AND APPLIED ECONOMICS

AAE 522: ECONOMETRICS: TIME SERIES ECONOMETRICS

TAKE HOME ASSIGNMENT

DEADLINE: JULY 4, 2022

INSTRUCTIONS: ALL ANSWER QUESTIONS AND USE THE WATER ENDERS BOOK,
SHARED ON THE WHATSAPP AS A REFERENCE MATERIAL.

QUESTION ONE: Answer the following questions:


(a) Explain what is meant when it is said that a time series is
1. Stationary (2 Marks)
2. Invertible (3 Marks)
3. Weakly stationary (2 Marks)
4. ergodic (3 Marks)
(b) Prove that the following univariate processes are weakly stationary and ergodic
1. 𝑦𝑡 = 𝑐 + 𝜀𝑡 + 𝜃1 𝜀𝑡−1 + 𝜃2 𝜀𝑡−2 (5 Marks)
2. 𝑦𝑡 = 𝑐 + ∅1 𝑦𝑡−1 + 𝜀𝑡 |∅| < 1 (5 Marks)
QUESTION TWO: Consider the following simple bivariate system and assume that 𝜀𝑦𝑡~𝑁𝐼𝐷(0,𝜎𝑦2 )
and 𝜀𝑧𝑡~𝑁𝐼𝐷(0,𝜎𝑧2 ) :

𝑦𝑡 = 𝑏10 − 𝑏12 𝑧𝑡 + 𝛾11 𝑦𝑡−1 + 𝛾12 𝑧𝑡−1 + 𝜀𝑦𝑡

𝑧𝑡 = 𝑏20 − 𝑏21 𝑦𝑡 + 𝛾21 𝑦𝑡−1 + 𝛾22 𝑧𝑡−1 + 𝜀𝑧𝑡


a. If one estimates such a structural model, explain the main problem associated with such
estimation (2 Marks)
b. Using matrix algebra, obtain the VAR model (5 Marks)
c. Show that the error terms of the VAR model are composites of the two structural shocks
𝜀𝑦𝑡 and 𝜀𝑧𝑡 (5 Marks)
d. Find the following:
1. The expected value, (2 Marks)
2. The variance, and (3 Marks)
3. The covariance of the error terms of the VAR model (3 Marks)
e. Attempt the following: One way to identify the model is the type of restriction proposed
by Sims (1980). Setting b21 = 0 is a method to decompose the residuals, known as the
Choleski triangular decomposition. Implement such restriction in the bivariate system and
recover the structural shocks. (10 Marks)

QUESTION THREE: Briefly explain how you might test for cointegration, paying particular
attention on the characteristics of the test statistics involved and the problems of their
implementation. (10 Marks)

QUESTION FOUR:
a. Explain briefly what is meant by
1. A Trend Stationary Model (TSM) (2 Marks)
2. Difference Stationary Model (DSM). (2 Marks)
3. Why is the distinction important? (1 Marks)
b. Consider the following process:
𝑦𝑡 = 𝛼 + 𝛽𝑇 + 𝜋𝑡
Where 𝑇 is a trend and error term is 𝜋𝑡 has a serial correlation of order two.
1. Re-parameterize the equation in such a way that you can test whether the
variable 𝑦𝑡 is a trend stationary or it contains a unit root. (3 Marks)
2. Explain the relevant hypothesis test you would have to make. (2 Marks)

c. Consider the Dickey Fuller and Augmented Dickey Fuller Approach to test for unit roots
in economic time series. Clearly explain how you would go about to test for
1. A random walk (2 Marks)
2. Random walk with a drift (3 Marks)

QUESTION FIVE

a. Suppose that the {𝜀𝑡 } sequence is the ARCH (1) process: 𝜀𝑡 = 𝑣𝑡 √ℎ𝑡 , where ℎ𝑡 = 𝛼𝑡 +
2 2
𝛼𝑡 𝜀𝑡−1 , 𝑣𝑡 is white noise and 𝜎𝑣2 = 1 and 𝑣𝑡 as well as 𝜀𝑡−1 are independent.
1. Write down the unconditional expectation and variance. (4 Marks)
2. Write down the conditional mean and the conditional variance of the process
[hint: conditioning on information at time 𝑡 − 1]. (4 Marks)
3. What do you notice about the nature of 𝜀𝑡 (2 Marks)
b. Consider the ARCH-M model represented by the following process:
𝑦𝑡 = 𝜇𝑡 + 𝜀𝑡 where
2
𝜇𝑡 = 𝛽 + 𝛿ℎ𝑡 and ℎ𝑡 = 𝛼0 + 𝛼1 𝜀𝑡−1 , let {𝜀𝑡 } be the white noise and for simplicity,
2 2
let 𝐸 (𝜀𝑡 ) = 𝐸(𝜀𝑡−1 ) = 1,
1. Find the unconditional variance of 𝑦𝑡 (5 Marks)

You might also like