Econ F241 Comprehensive

ECON F241 COMPREHENSIVE EXAMINATION
SEMESTER II 2020-2021
Writing Time: 3.00 pm to 5.00 pm TOTAL MARKS: 40
Attempt all questions
Note: For excel based questions write the answers in the answer script. Working is required in excel
and to be sent along with the scanned script.
Question 1
a) Serial independence implies that the disturbance occurring at one point of time does not carry over into
another period. True/ False. [1]
b) Exclusion of a relevant explanatory variable from the model, whose successive values are correlated,
will make the disturbance term associated with the model autocorrelated. True/ False. [1]
c) The autocorrelation problem becomes more prevalent when time series data with shorter time intervals
are considered. True/ False. [1]
d) Consider the following autocorrelation relationship:  t   t 1  t . If the correlation coefficient

between  t and  t -1 is 0.91, what is the value of  . [1]
Question 2
a) Consider the model [3]

Yi  1   2 X 2i  3 X 3i  i
Suppose heteroscedasticity takes the following form:  i2   2 X1i
How would you transform the model to remove heteroscedasticity. Also verify that the transformed
model is free from the problem of heteroscedasticity.
b) Show that the zero order autocorrelation coefficient is unity. [3]

c) Show that the random walk model, with or without drift, is a non-stationary process. [3]
d) In the presence of autocorrelation the OLS estimators are still biased. True/False. Prove your
claim. [3]
e) In a system of equations if there are more than one endogenous regressors it is possible to have at
least one equation identified. True/False. Explain with the help of an example. [3]
f) Consider the datasheet named Q2 (f) data. For the regression model of Y on X, it is suspected that
the explanatory variable suffers from heteroscedasticity. Use the Goldfeld Quandt test to verify it.
Clearly write the null and alternate hypothesis, computed test statistic and the decision (cut out 8
central observations). [3]
Question 3
Consider the data of log Y and log X provided in the data sheet named Q3 data. Assume that both the
series are I(1) [you are not required to test this]. [2+2+2=6]
a) Estimate the long run relationship between the two variables (log X is the regressor). Write the
estimated equation.
b) Use a suitable test equation to determine if the long run relationship holds. Write down the
parameter value of the test equation. Clearly write the null hypothesis and alternate hypothesis
and the conclusion at 5% significance level. What are the computed and the critical value of the
test statistic?
1
c) Is there disequilibrium in the relationship between the two series in the short run? Explain by
running a suitable model and write the estimated equation of the model. Can the equilibrium be
restored in the long run? Explain.
Question 4
Consider the following system of equations and solve the following questions. [2x4=8]
Model 1:
Demand function : QD  1   2 P   D
Supply function : QS  1   2 P  S
Equilibriu m condition : Q  QD  QS
a) In the above system neither of the equations is identified. True/False. Verify your response using
the reduced form equations.
b) Suppose you add income (denoted by I) as an exogenous variable in the demand function of
model 1. Supply function remaining the same. Call this new system as model 2. Can you identify
the slope and intercept of the supply equation in model 2. Show algebraically.
c) The assumption in Model 2 is that income does not influence supply. If this assumption is not
correct, your estimates of the supply curve in model 2 would be meaningless. True/ False.
Explain your answer algebraically.
d) Suppose you add income (denoted by I) as an exogenous variable in the demand function and
exogenous technology variable (T) in the supply function of model 1. Call this new system as
model 3. How many structural coefficients and reduced form coefficients are present in this
system? Is the identification of all the structural coefficients from the reduced coefficients
possible? Explain.
Question 5
Consider the following model:

Yi   0  1Di   2 X i  i
Where Y=annual salary of professor
X=years of experience
D=dummy for gender
Consider three ways of defining the dummy variable:
i) D=1 for female, 0 for male
ii) D=1 for female, 2 for male
iii) D=1 for female, -1 for male
a) Interpret the regression model for each dummy assignment. [3]
b) Is one method preferable to another? Justify your answer. [1]

Econ F241 Comprehensive

Uploaded by

Copyright:

Available Formats

Econ F241 Comprehensive

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Econ F241 Comprehensive

Uploaded by

Copyright:

Available Formats

ECON F241 COMPREHENSIVE EXAMINATION

d) Consider the following autocorrelation relationship:  t   t 1  t . If the correlation coefficient

a) Consider the model [3]

b) Show that the zero order autocorrelation coefficient is unity. [3]

Consider the following model:

You might also like