Time Series Econometrics TSE48M1
Assignment: Due 25 June 2021
100 MARKS
QUESTION 1 [30 marks]
(a) Explain why the linear probability model is inadequate as a specification for limited
dependent variable estimation. [5]
(b) Assuming that all seven classical assumptions, the OLS estimators and can have
several desirable properties. List those properties and explain concisely what each
means. [8]
(c) List and briefly explain the steps used in the Box-Jenkins approach to constructing an
ARMA model. [ 9]
(d) Briefly note the advantages of constructing a parsimonious ARMA model [8]
QUESTION 2
[20 Marks]
In a bid to establish the impact of income on consumption, the following regression was
estimated by a researcher.
Dependent Variable: CONS
Method: Least Squares
Date: 02/28/14 Time: 18:48
Sample: 1982 1996
Included observations: 15
Variable Coefficient Std. Error t-Statistic Prob.
C -184.0780 46.26198 -3.979034 0.0016
INCOME 0.706408 0.007827 90.24707 0.0000
R-squared 0.998406 Mean dependent var 3964.087
Adjusted R-squared 0.998284 S.D. dependent var 489.6614
S.E. of regression 20.28525 Akaike info criterion 8.981231
Sum squared resid 5349.390 Schwarz criterion 9.075638
Log likelihood -65.35923 F-statistic 8144.534
Durbin-Watson stat 2.081830 Prob(F-statistic) 0.000000
a. In general, what is the economic interpretation of the intercept in a consumption
function? [2]
b. Do your results with respect to the intercept make economic sense? Elaborate.
[2]
c. What is the economic interpretation of the slope coefficient? [3]
^β =3
d. How would you react to a colleague’s estimation results showing that 2 ? [1]
e. Give some examples of the kind of factors that the error term would capture in this
model? [2]
f. How well does your model fit the data? What is the measure of goodness of fit and
how do you interpret its value? [3]
g. A test for normality of the residuals was conducted and the results are illustrated in
figure 1.
Figure 1: Normality Test
6
Series: Residuals
Sample 1982 1996
5
Observations 15
4 Mean -3.03e-13
Median 1.760866
Maximum 31.30563
3
Minimum -39.33004
Std. Dev. 19.54736
2 Skewness -0.571746
Kurtosis 2.636503
1
Jarque-Bera 0.899815
Probability 0.637687
0
-40 -30 -20 -10 0 10 20 30 40
i. Interpret the results. [2]
ii. What five assumptions are usually made about the unobservable error terms in the classical
linear regression model (CLRM)? [5]
QUESTION 3 [10 marks]
(a) Why have VARs become popular for application in economics and finance, relative to
structural models derived from some underlying theory? [4]
(b) Discuss any weaknesses you perceive in the VAR approach to econometric
modelling. [6]
Question 4 [30 MARKS]
(a) Calculate the t-ratios for the coefficients in the model below.
yˆ t =0.638+0.402 x 2 t −0.891 x 3 t R2=0.96 , ^
R 2=0.89
s.e (0.436) (0.291) (0.763) [6]
(b) Are regression parameters significantly different from zero? Explain [4]
(c) Do you suspect multicollinearity from results in question 1(a) above? Explain fully
what is meant by multicollinearity. [4]
2
(d) Explain what the R mean from the regression model [3]
2
(e) Explain the adjusted R and what it means [3]
(f) What is the solution to multicollinearity? Explain more than four solutions [10]
QUESTION 5 [10 MARKS]
Consider a regression model of relating Y (the dependent variable) to X (the independent
variable) Yi = 0 + 1Xi+ i where i is the stochastic or error term. Suppose that the
estimated regression equation is stated as Yi=❑0 +❑1 Xi and ei is the residual error term.
(a) What is ei and define it precisely. Explain how it is related to i. [4]
(b) What is i and define it precisely. What are the four reasons for the inclusion of
this error term in the population regression function (model)? [4]
(c) Distinguish or make contrast between an estimator and an estimate. [2]
