·

Task: write a (roughly) 15 page report on some statistical analysis that you have

performed. The report should use the Word template provided on Moodle and should

(for all but the last part of the analysis) use the data also provided on Moodle.

The statistical analysis to be performed by each group is an examination of how

a particular stock return varies with stock market returns and industry returns. It

entails some basic statistical analysis plus some regression estimation. All regressions

should contain an intercept and have appropriate checks on residuals performed before

they are interpreted and before any t-tests are performed. All t-tests should be

conducted at the 5% level (and perhaps also at the 1% level).

Basic instructions:

1. Find your assigned company in the spreadsheet FR2202 stocks to groups.xlsx.

2. Then look in the spreadsheet FR2202 coursework data.xlsx where you will

nd two years of daily prices for your company. The spreadsheet also contains

data on daily levels of the S&P-500 (column CY) and (in column CZ onwards)

data on the levels of 10 sector indices of the S&P-500. Select the single sector

index that best matches the area in which your company operates.

3. Create daily stock, market and sector returns. Plot them and check for outliers

and errors.

4. Present a table containing basic summary statistics for the returns of your

stock, the returns on the market and the returns on the sector. Discuss the

features of the returns using the summary statistics and the plots and compare

across the three series.

5. Run a regression of the stock return on the market return. After diagnostic

checking, interpret the regression. Test the null that the intercept is zero against

a two-sided alternative and the null that the slope is zero against the alternative

that it is greater than zero. Also test the null that the slope is one against a

two-sided alternative. Be sure to give economic interpretations of the results of

these tests.

6. Run a multiple regression of the stock return on the market and the sector

returns. Again, after diagnostic checks, interpret the regression and run the

appropriate t-tests.

7. Run an F-test of the null that both slope coecients are zero against the alter-

native that at least one of them is not. Interpret the test results.

8. Use your own research and/or intuition to modify the multiple regression model

in a sensible way and then estimate the modied model. For example, you

might add another determinant of the stock return to the right-hand side of

the regression or you might check for a structural break in the model and re-

estimate if a structural break is found. This is a chance for you to add your

own twist to the model and to see if it works. You will be judged on the quality

of the idea and the execution, not on whether the modication works or not.

Report structure: use the Word template on Moodle.

1. The rst page should contain the title of the report, the names of group mem-

bers and a 100 word (no more) abstract.

2. Then add a 1 page introduction describing what you are going to do and brie

what you nd. On this page also briey describe the nature of the company

you have been assigned.

3. The next section should contain return plots, a table of summary statistics and

surrounding discussion. It should be no more than 2 pages long.

4. The following section should contain the basic regression, its diagnostic tests,

interpretation and t-tests. No more than 3 pages.

5. Then there should be a section on the multiple regression with all the accom-

panying tests and tables. No more than 3 pages.

6. The nal empirical section should motivate and present your extension to the

model. Again, no more than three pages.

7. Provide a 1 page conclusion

8. There should be a closing reference list if you have referred to any papers or

books.

2

Note that all tables and plots should be in the text rather than at the end of the

document. Do not change the font, spacing or margins in theWord template. Convert

the document to PDF before submission.

Group participation: this is a piece of group work. Any group member who has

not participated in the analysis or the writing should be identied to me and action,

including the possibility of failing non-participants on this assessment component,

will then be taken.

Submission: the deadline for submission is 1pm onWednesday the 5th of December.

Reports should be submitted electronically via the FR2202 moodle page.

所有文章

×