Regression

GrAssignment

GB462R

Assignment#4 (G2)

Regression

This assignment is a team assignment and has two parts. You only need to upload one copy of the write- up per team. The uploaded file should be a pdf file (no excel file needed) and should have your team identifier. Make sure you list each of your team members on the cover page.

Please read each question carefully and note that you will have to make assumptions (and compromises) along the way. Note: All relevant information pertaining to this assignment is in this document and you will not need any further details from either the course or lab instructor.

1 Estimating Firm β’s

The Capital Asset Pricing Model (CAPM) relates the expected excess return on an asset to the expected excess return on the “market portfolio” (see discussion on p. 122 of Stock and Watson posted on blackboard):

−#=  +()−#)

where α (the intercept) is assumed to be zero in theory, but you should include it in your estimation. You will estimate  (the slope) for a few individual stocks.

Collect data from Yahoo Finance (finance.yahoo.com) and compute estimates of β’s for the following firms for the calendar years 1990 through 1999:

1. Home Depot (HD)
2. Apple (AAPL)
3. Verizon (VZ)
4. Cisco (CSCO)

For each stock you will have to download data, choose an appropriate market index, construct excess returns, and decide on a relevant frequency of data (daily, weekly, monthly …). In collecting your dataset, you might also find the following website useful:

http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html

As a write-up construct and fill in the table on the next page and use a maximum of one page of text to justify your choice of market index and data frequency, discuss the estimates, and tests related to α and β.

2 Computing Price Elasticities (and optimal prices)

Using the data posted on the class website (rfj_data.xls) run three simple regressions of the following type

3=  +3+3

for each of the three brands in the data (Tropicana, Minute Maid and Private label). (a) Using the regression results compute own price elasticities for each brand. Remember:

=

(b) Assume that the per unit cost of producing an additional unit is the same across brands and is 1 cent per ounce. Based on the data provided and the estimated demand equation from above, compute the optimal price for each brand.

Using a maximum of one page describe how you went about answering questions (a) and (b) and comment on your findings.