30 October 2006
G601, Indiana University
Professor Rasmusen, erasmuse@indiana.edu
Http://www.rasmusen.org/g601/tasks/ps-regression.txt
Due: December 7
REGRESSION ASSIGNMENT
For this assignment, you will run a regression, present it as a
table, and explain it. Include a titlepage.
1. Using Compustat, pick an industry and sample at least 30 firms from
it. Tell me how you chose your sample.
2. Choose a measure of profitability and 4 variables that might explain
it. One (and only one) of the variables should be a dummy variable. I will not
be
commenting on how sensible your variables are, so you can be fairly arbitrary.
3. You will run several specifications of the regression equation.
A. Regress profit on the 4 variables and a constant.
B. Regress profit on 3 of those variables and a constant.
C. Regress Log(Profit) on the log of one of the continuous variables, on the
dummy variable, and on a constant.
4. Show your results in a table with one column for each specification.
5. Explain what the size of each coefficient means in
specifications A and C, ignoring statistical significance. Your explanations
should be something like, "Regression A says that if the ratio of
inventories to assets increases by .1 then the return on equity rises by
.02," or "Regression A says that if the ratio of inventories to assets
increases by 10% then the return on equity falls by 3%."
Email me your latex and pdf files as attachments.