Cass Undergraduate Programme

BSc (Hons) Degree in Finance

BSc (Hons) Degree in Accounting and Finance
BSc (Hons) Degree in Banking and International Finance
BSc (Hons) Degree in Investment & Financial Risk Management

FR2202 Financial Econometrics

Stage 2 Examination

January 2021 2 hours 15 mins

Instructions to students:

Candidates should answer all 5 questions from part A and question 6 from part B

Only your first answers within the rubric will be marked. Do not attempt additional questions
outside of the rubric. For example, if you are asked to answer 3 from a selection of 5 questions
- either in an assessment or a section within an assessment - and you answer 4, only the first
3 as they appear in your script will be marked, the 4th will be disregarded. Any work you do
not wish to be taken into consideration should be deleted or clearly crossed through.

If you believe a question to be ambiguous you are advised to make reasonable assumptions
provided they are consistent with the facts given in the question. State any such assumptions
clearly in your answer and proceed. If you believe there to be an error in the question state
this and proceed.

Section A carries 50 marks . Section B carries 50 marks

This examination paper consists of X printed pages including the title page. Students need
access to the statistical Tables and to the excel spreadsheet of data .

Materials:

This paper should be typed. Eviews output must be pasted and copied on the paper

All assessment will need to be submitted to the Moodle dropbox as one single pdf file

All submissions must be labelled in the correct format of your Student ID followed by a full stop
followed by the module code:
o 9digitnumber.ModuleCode (shown on the cover of all assessment as two letters followed by four
numbers)

Any queries should be addressed to: cass-ug-examinations@city.ac.uk

Internal Examiner: Dr Malvina Marchese

External Examiner:

Page 1 of X
 PART A ( 50 Marks)
Answer all questions. Each question carries 10 marks

Question 1 (10 marks)

Answer all the following points :

a) What is an estimator ? Provide one exmple of an estimation principle you have been using
and dicuss its limitations.
b) Explain what is meant by consistency of an estimator Why is consistency such an important
property of an estimator ?
c) Explain what is meant by unbiasdness of an estimator. Disucss the difference between un
biasdness and consistency and their relation.
d) Explain the difference between the standard error of the regression and the standard error
of the estimator
e) Explain the difference between time series , cross section and panel data type of sample.

Question 2 ( 10 marks)
A researcher is interested in assessing the determinants of log wage. She has collected a
sample of cross-sectional data on log wage (LWAGE), education (EDUC), experience
(EXPR) for a total of 526 observations and she has obtained the following output

Dependent Variable: LWAGE

Method: Least Squares

Sample: 1 526
Included observations: 526

Variable Coefficient Std. Error t-Statistic Prob.

C 0.220231 0.106767 2.062720 0.0396

EDUC 0.091908 0.007620 12.06171 NA
EXPER 0.007813 NA 4.777777 0.0000
MARRIED NA 0.043443 4.368781 0.0000

R-squared 0.275822 Mean dependent var 1.623268

Adjusted R-squared NA S.D. dependent var 0.531538
S.E. of regression NA Akaike info criterion 1.264505
Sum squared resid 107.4171 Schwarz criterion 1.296940
Log likelihood -328.5647 Hannan-Quinn criter. 1.277205
F-statistic 66.27242 Durbin-Watson stat 1.803025
Prob(F-statistic) NA

i) Compute and interpret the coefficient of MARRIED

ii) Is the model significant in the overall? Justify your answer
iii) How much of the variability of the sample of log wage is explained by the model?
iv) Compute the standard error of the regression and the standard error of the estimator of
the beta experience
v) Is there evidence of first order serail correlation? Justify your answer

Page 2 of X
Question 3 ( 10 marks)
The researcher decides to run post estimation diagnostic tests on the model provided in
question 2. For each of the outputs reported below , name the test , its distribution under the
null, the null hypothesis ,find the critical value and draw conclusions

a) Test one

Value df Probability
F-statistic 2.829737
Likelihood ratio 17.02882

F-test summary:
Mean
Sum of Sq. df Squares
Test SSR 3.421853 6 0.570309
Restricted SSR 107.4171 522 0.205780
Unrestricted SSR 103.9953 516 0.201541

LR test summary:
Value
Restricted LogL -328.5647
Unrestricted LogL -320.0503

Unrestricted Test Equation:

Dependent Variable: LWAGE
Method: Least Squares

Sample: 1 526
Included observations: 526

Variable Coefficient Std. Error t-Statistic Prob.

C 1.805472 17.58084 0.102695 0.9182

EDUC 1.914806 22.84279 0.083825 0.9332
EXPER 0.162156 1.941735 0.083511 0.9335
MARRIED 3.952696 47.17435 0.083789 0.9333
FITTED^2 -27.21484 651.0620 -0.041801 0.9667
FITTED^3 -8.880505 890.3776 -0.009974 0.9920
FITTED^4 44.89036 696.0700 0.064491 0.9486
FITTED^5 -36.72008 313.8324 -0.117005 0.9069
FITTED^6 12.55764 76.02440 0.165179 0.8689
FITTED^7 -1.592404 7.667936 -0.207671 0.8356

R-squared 0.298891 Mean dependent var 1.623268

Adjusted R-squared 0.286663 S.D. dependent var 0.531538
S.E. of regression 0.448933 Akaike info criterion 1.254944
Sum squared resid 103.9953 Schwarz criterion 1.336034
Log likelihood -320.0503 Hannan-Quinn criter. 1.286694
F-statistic 24.44190 Durbin-Watson stat 1.792650
Prob(F-statistic) 0.000000

Page 3 of X
 a) Test 2

F-statistic 1.730344
Obs*R-squared 10.37453

Test Equation:
Dependent Variable: RESID
Method: Least Squares

Sample: 1 526
Included observations: 526

Variable Coefficient Std. Error t-Statistic Prob.

C 0.001991 0.106983 0.018614 0.9852

EDUC 0.000106 0.007626 0.013859 0.9889
EXPER 0.000128 0.001640 0.078350 0.9376
MARRIED -0.009150 0.043572 -0.210007 0.8337
RESID(-1) 0.093824 0.044234 2.121058 0.0344
RESID(-2) -0.022316 0.044335 -0.503356 0.6149
RESID(-3) 0.029611 0.044429 0.666487 0.5054
RESID(-4) 0.053044 0.044211 1.199793 0.2308
RESID(-5) 0.068879 0.044491 1.548134 0.1222
RESID(-6) 0.014044 0.044319 0.316881 0.7515

R-squared 0.019723 Mean dependent var -6.46E-17

Adjusted R-squared 0.002626 S.D. dependent var 0.452332
S.E. of regression 0.451738 Akaike info criterion 1.267398
Sum squared resid 105.2985 Schwarz criterion 1.348487
Log likelihood -323.3256 Hannan-Quinn criter. 1.299148
F-statistic 1.153563 Durbin-Watson stat 1.999073
Prob(F-statistic) 0.323017

Question 4 ( 10 marks)
The rearcher mentioned in question 2 and 3 wishes to test for heteroskedasticity
i) Disucss the test she can run to detect more effectively heteroskedasticity . make sure
to state the null and the alterntivae hypothesis , the test equation and the distribution
of the test statistics under the null.
ii) In performing the test abve she finds a pvalue of 0.0001 what can she conclude?
iii) Explain how the reseacher can solve the problem of heteroskedaticty to re estimate the
model in question 3 and what will change in the regression output.
iv) Will step iii) impact on the significance of the regression? Justify your answer.

Page 4 of X
Question 5 (10 marks)

Open the data for question 5 . The dependent variable of interest is the returns on a uk
based portfolio ( POrRTFOLIO). The other variables are : the MARKET(MARKET) is the
excess return on the FTSE100; the size factor (SIZE) is the return on a portfolio of small
stocks less the return on a portfolio of large stocks, the value factor (VALUE) defined as
the return on a portfolio of high value stocks less the return on a portfolio of low value stocks
(where value is measured using the book-to-market ratio), and the momentum
(MOMENTUM) factor which is the return on stocks that have performed strongly over the
last year less the return on a portfolio of stocks that have performed poorly. All variables are
log returns. Adress all the folliwng points:
1) Estimate a dynamic multiple linear regression model of portfolio on market size and value.
Find if there are significant breaks and identify the break dates.
2) Using your results of part 1) generate dummy for the break dates
3) Test if the break/breaks impact the elesticity of the portfolio to market

Page 5 of X
PART B ( 50 Marks) EVIEWS PRACTICE QUESTION
Answer all parts of question 6 .

Question 6
Import the Excel spreadsheet of data provided into Eviews called data for question 6.
You are working as the main econometric advisor of the US Governamnet . The
dependent variable of interest is the M1 – money supply, billions of USD. The
explanatory variables are :. IP –industrial production, index levels (source: Board of
Governors of the Federal Reserve); Tbill – 3-month US Treasury rate (source: Board
of Governors of the Federal Reserve) and CPI – Consumer Price Index, level (source:
Bureau of Labor Statistics)

1) Estimate a preliminary multiple linear regression model for M1 using all the explanatory
variables and comment on the validity of the model (copy the eviews output in your
answer)
2) Conduct all the post estimation diagnostic tests that you deem appropriate to establish
the validity of the results above . For each test , copy the eviews output in your answer (
including the test equations) and draw conclusions.
3) Using the results of point 2 , estimate the multiple linear regression model that you have
found to be the more appropriate for M1, dropping any non significant explanatory
variable you may have found and adding any interaction term you find appropriate
.Comment on the appropriatness of the model you are proposing and interpret your
findings.
4) Generate a dummy variable for the month of January . Is there a January effect in the
dyanmics of M!? Does the impact of IP on M1 change in January ? Justify your answers and
comment.
5) Is there a structral break in Junary 2000? Justify your answer and explain how you can
include this information in your model if your aim is to forecast furture values M1..
6) Can you improve on point 5 ? Discuss and provide the best model you could use to advise
the US Governament.

Page 6 of X