Scribd
Upload
5 views6 pages

Exam2 Solution

Problem solving method

Uploaded by

yeabsiraabreham37
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
5 views6 pages

Exam2 Solution

Problem solving method

Uploaded by

yeabsiraabreham37
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/ 6

Econometrics - Exam 1

Exam
Problem 1: (15 points)
Suppose that the classical regression model applies but that the true value of
the constant is zero. In order to answer the following questions assume just
one independent variable.
1. Give the formulae for the two least squares slope estimators (the one
with and the one without the constant).
2. Calculate their variances.
3. Compare the variance of the least squares slope estimator computed
without a constant term with that of the estimator computed with an
unnecessary constant term.

Solution:

1.
P
(xi − x̄)(yi − ȳ)
y = β1 + β2 x + ε → β2 = i P 2
P i (xi − x̄)
xi yi
y = β˜2 x + ε → β˜2 = Pi 2
i xi

2.
σ2
V ar(β2 ) = P 2
i (xi − x̄)
σ2
V ar(β˜2 ) = P 2
i xi
Econometrics - Exam 2

3. The ratio of these two variances is


2
Pσ 2
V ar(β˜2 ) i xi
= σ2
V ar(β2 ) P 2
i (xi −x̄)

− x̄)2
P
(xi
iP
= 2
i xi
X X X X
(xi − x̄)2 = (x2i − 2xi x̄ + x̄) = x2i − 2nx̄x̄ + nx̄2 = x2i − nx̄2
x2i
− nx̄2
P
iP
= 2
i xi
nx̄2
= 1− P 2 ≤1
i xi

It follows that fitting the constant term when it is unnecessary inflates


the variance of the least squares estimator if the mean of the regressor
is not zero.
Econometrics - Exam 3

Problem 2: (15 points)


Suppose that y has the pdf f (y|x) = β10 x e−y/(β x) , y > 0. Then E[y|x] =
0

β 0 x and V ar[y|x] = (β 0 x)2 . For this model, prove that the GLS and MLE
estimators are the same, even though this distribution involves the same
parameters in the conditional mean function and the disturbance variance.

Solution:
First the GLS estimator:
!−1 !
X xi xi 0 X xi yi
β̂ GLS = (X 0 Ω−1 X)−1 X 0 Ω−1 y =
i
(β 0 xi )2 i
(β 0 xi )2

Next the MLE estimator:


Y 1 0
L = 0 e−yi /(β xi )
i
β xi
X X
ln L = − ln(β 0 xi ) − yi/(β 0 xi )
i i
∂ ln L X xi X
= − 0 + yi/(β 0 xi )2 xi = 0
∂β i
β xi i
X yixi X xi X xi X xi xi 0 β
= Now write =
i
(β 0 xi )2 i
β 0 xi i
β 0 xi i
(β 0 xi )2
X yixi X xi xi 0 β
=
i
(β 0 xi )2 i
(β 0 xi )2
!−1
X xi xi 0 X yi xi
β̂ M LE =
i
(β 0 xi )2 i
(β 0 xi )2
Econometrics - Exam 4

Problem 3: (15 points)


The following model is estimated using a balanced panel of five firms over
20 years: Iit = β1 Fit + β2 Cit + εit , where the regressors are market value (F )
and capital (C ) and the dependent variable is investment (I ). Suppose that
the true error structure of the model is εit = αi + ηit , where α is uncorrelated
with the regressors.
1. If the model is estimated as a fixed effects model, what will be the
statistical properties, in terms of efficiency and consistency, of the esti-
mates?
2. The estimates for pooled OLS, fixed effects (using dummies) and ran-
dom effects models are given in the table below. Use the statistics
shown to decide whether the data support a fixed effects or random
effects specification. Carefully explain your reasoning.
Dependent Variable is Investment
Estimation Constant Market Value Capital
(a) OLS -48.030 (-2.236) 0.10509 (9.236) 0.30537 (7.019)
(b) Fixed Effects - 0.10598 (6.669) 0.34666 (14.348)
(c) Random Effects -61.575 (-0.775) 0.10549 (6.859) 0.34641 (14.350)
(t-ratios are shown in brackets)
Breush-Pagan LM test for random effects (1 df): 453.82
Hausman test of fixed vs random effects (2 df): 1.27

Solution:
1. If the individual effects are strictly uncorrelated with the regressors
then a random effects model is the appropriate model. However, if a
fixed effect model is estimated the estimates will be consistent but not
efficient.
2. Breush-Pagan LM test: Test statistic is 453.82, the critical value from
the chi-squared table is 3.84, so the null hypothesis that random effects
are not needed can be rejected.
Hausman Test: Test statistic is 1.27, the critical value from the chi-
squared table is 5.99, so the null hypothesis of the random effects model
cannot be rejected.
Econometrics - Exam 5

Problem 4: (15 points)


Consider the stochastic processes given below. For each process determine
what the effects of first differencing the process, i.e. computing yt − yt−1 , on
autocorrelation are, e.g. reduction of the autocorrelation.
1. yt = yt−1 + εt , where εt is normally distributed white noise.
2. yt = β0 + β1 t + εt , where εt is normally distributed white noise.
3. yt = β 0 xt + εt , where εt = ρεt−1 + ut and ut is normally distributed
white noise.
[Hint: Compare the autocorrelation of εt and the autocorrelation of
(εt − εt−1 ).]

Solution:
1. ∆yt = yt − yt−1 = εt , white noise, no more autocorrelation
2. ∆yt = yt − yt−1 = β1 + εt − εt−1 . This is an MA(1) process with
θ 1 1
autocorrelation 1+θ 2 = 1+1 = 2 .

3. ∆yt = yt − yt−1 = β 0 (xt − xt−1 ) + vt , where vt = εt − εt−1 .

σu2
V ar(εt ) =
1 − ρ2

V ar(vt ) = V ar(εt − εt−1 ) = V ar(ρεt−1 − εt + ut )


σ2 2σu2
= V ar[(ρ − 1)εt−1 + ut ] = (ρ − 1)2 u 2 + σu2 =
1−ρ 1+ρ

Cov[vt , vt−1 ] = Cov[εt − εt−1 , εt−1 − εt−2 ]


= E[εt εt−1 − ε2t−1 − εt εt−2 + εt−1 εt−2 ]
σ2 σu2 2 σu
2
σu2 σu2 (2ρ − 1 − ρ2 )
= ρ u 2− − ρ + ρ =
1−ρ 1 − ρ2 1 − ρ2 1 − ρ2 1 − ρ2
σu2 (ρ − 1)2 σ 2 (ρ − 1)
= = u
(ρ − 1)(ρ + 1) ρ+1
Cov[vt , vt−1 ] ρ−1
=
V ar[vt ] 2
Econometrics - Exam 6

Compare the two autocorrelations:

ρ−1
|ρ| > Assume ρ > 0 and |ρ| < 1
2
ρ−1 1
ρ > − →ρ>
2 3
If the original autocorrelation is greater than 1/3 (For economic data,
this is likely to be fairly common.) the differenced process has a smaller
autocorrelation.

You might also like