CORPFIN 2503 - Business Data Analytics
2020 S2, Group Assignment
Sigitas Karpavičius
Instructions
1. The assignment can be done in groups of one to three students. All team
members are expected to contribute approximately equally to a group assignment. A group can eliminate an underperforming member who then will need
to do the assignment individually or join another group. All group members
will get the same mark for the assignment.
2. The maximum score is 30 points.
3. The presentation of your write-up is important. Up to 5 bonus points will
be given for a properly formatted report.
4. All numerical analysis, all tables and figures need to be done using SAS (however, you may use Excel or Word etc. to make tables for regressions as the
standard SAS output for regressions is not very nice).
5. Please retain your SAS code and make sure that it is user-friendly (use comments where necessary). Using your submitted code and data set, one should
be able to produce all your results, tables, and figures.
6. Please retain a copy of the problem set that is submitted.
7. Only one member of a group submits:
• a SAS code,
• a data file (e.g., in txt, csv, xls, xslx formats), and
1
• the report (in doc, docx, or pdf format) with ‘Group Assignment Cover
Sheet’, which must be signed (electronic signature is okay) and dated
by all group members before submission; the report should be properly
formatted and be similar to a business report; font: 12 pt Times New
Roman; maximum number of pages: 15.
8. Lecturer can refuse to accept assignments, which do not have a signed acknowledgment of the University’s policy on plagiarism.
9. Any suspected plagiarism will be severely punished. This includes any student
that submits copied work or any student that allows their work to be copied.
10. You must acknowledge any external material you use in your answers, e.g.,
material from websites, textbooks, academic journals and newspaper articles.
11. All queries for this project should be directed to Lecturer.
12. The submission deadline for the problem set is 6pm, Sunday the 18th of October, 2020.
13. The submission must be done through MyUni.
14. Late submission will be penalized 3 points per day.
Agenda
The assignment is based to some extent on Workshop 4. Assume that you are a
bond analyst and you have been asked to look at the U.S. corporate bond market.
1 Sample and description statistics (5 points)
From Eikon download bond issues with the following characteristics:
1. Sector: Consumer Goods, Manufacturing, Telephone, Transportations
2. Domicile: United States
3. Amount Outstanding: > 100,000,000
4. Coupon: > 0%
2
5. Maturity: > 1-October-2024
6. S&P Long-term Issue Credit Rating: > D.
7. make sure that the following filters are visible on the left hand side:
(a) Currency
(b) Convertible
(c) Callable
(d) Putable
(e) Seniority
(f) Market of Issue
You should end up with more than 2,000 bonds; thus, you will need to download
the data to Excel in batches (e.g., first, download small bond issues, then medium,
and lastly with large bond issues).
Discuss briefly your sample, including the number of observations, outliers. Provide the descriptive statistics of the sample. How you choose to do this is entirely at
your discretion. However, it is recommended that you consider using both summary
statistic and graphical methods (this task should include at least one pie chart, one
histogram, and one scatter plot) while also noting any peculiarities within the data
set. You should put more emphasis on variables that are the dependent variables in
the regressions estimated in other tasks.
2 Which bonds are more likely to include “a call”
feature? (7 points)
A bond issuer can repurchase its bonds before their maturity if they include a call
feature. Firstly, identify statistically significant characteristics of callable bonds.
You may consider, issue size, maturity, industry, credit rating, and other variables
available in Eikon. Secondly, compute the average of the individual marginal effects
of the amount outstanding on the probability that a bond includes a call feature.
Thirdly, estimate the probability that a bond with the following characteristics
includes a call feature:
3
• maturity: 10 years
• coupon: 2.5%
• amount: $750,000,000
• currency: US dollars
• seniority: senior unsecured
• S&P credit rating: A
• sector: Electronics
• domicile: USA
• convertible: no
• market of issue: Global
• puttable: no.
Do the results suggest that this bond is callable or not?
In this task, you are expected to use a logit regression analysis. To ensure that
the results are robust, estimate at least two regression models (e.g., in the first
regression model, one includes amount in $ and in the second model, one uses the
natural logarithm of amount in $).
3 Which bonds are more likely to be issued in the
domestic and foreign markets? (6 points)
“Market of Issue” can take the following values:
• Domestic: a bond is issued in the US
• Global: a bond is issued in the US and foreign markets
• Eurobonds: a bond denominated in USD is issued in the foreign market.
You need to estimate the probabilities that a bond with the following characteristics is issued in each market:
4
• maturity: 10 years
• coupon: 3%
• amount: $750,000,000
• currency: US dollars
• seniority: senior unsecured
• S&P credit rating: A
• sector: Electronics
• domicile: USA
• convertible: no
• callable: yes
• puttable: no.
According to the analysis, what is the most likely market of issue of this bond?
In the analysis, estimate multinomial logit regression model and briefly discuss
the determinants of “Market of Issue.”
4 Estimating yield for a hypothetical bond (7 points)
Lastly, you need to estimate the yield for a bond with the following characteristics:
• maturity: 10 years
• coupon: 3%
• amount: $750,000,000
• currency: US dollars
• seniority: senior unsecured
• S&P credit rating: A
• sector: Electronics
5
• domicile: USA
• convertible: no
• callable: yes
• puttable: no
• market of issue: Global.
To ensure that the results are robust, estimate at least 3 regression models (e.g.,
in the first regression model, one includes amount in $, in the second model, one
uses the natural logarithm of amount in $, and the third model features something
else). Briefly discuss the determinants of yield.
Using one of the regression models, compute two additional yields:
1. the amount is $1,000,000,000, other bond characteristics the same as above
2. S&P credit rating is AA, other bond characteristics the same as above (i.e.,
amount: $750,000,000 etc.).
Are the results the same as the main estimate? Why?
Additional information
1. Before implementing statistical and regression analysis, check whether your
sample includes any outliers and duplicate observations. If needed, take necessary actions to deal with them.
2. To ensure that regression residuals “behave well,” you may need to scale or
transform one or more variables. For example, to use a natural logarithm
value of the variable instead of its raw value.
3. In the analysis, you should only use the data that can be downloaded from
Eikon.
4. You may move technical calculations to Appendix if you think it helps your
report to look more professional.