Economics 312

Spring 2010
Econometric Project #1: Capital-Asset Pricing Model

Due 6am, Tuesday, February 9: Berndt Chapter 2, Exercises 2, 3, and 5

To get some background on the economic issues with which you will work, you should begin by reading Berndt's Chapter 2 on the capital-asset pricing model. The data for the exercises in Chapter 2 have been combined in two Stata data files that can be downloaded from the Web site: Companies.dta and Gold.dta. (Berndt puts each company in a separate file; all of his company files are now combined in Companies.dta.)

As is the case with most of Berndt's chapters, Exercise 1 is an "exploring the data sets" activity. While I will never assign these, working through the various suggested activities is a great idea  in order to become more familiar with the data.

Exercises 2, 3, and 5 are required. (Note: You should not use the same period and industries in Exercises 2 and 5.) As discussed here, your report should include responses to all of the parts of each exercise and should include a copy of your Stata output from all statistical procedures.

The Berndt datasets are quite dated: many of the firms whose stock returns are used in the data no longer exist! It is possible to download stock-price data from Yahoo! Finance (and probably many other places). If you want to pick a more recent sample period of comparable length and more contemporary firms, you may download the data and use them for Exercises 2 and 5. All I ask is that you clear your choices with me before you proceed.

A few pointers if you choose to download data from Yahoo! Finance:

• You'll need to figure out the stock ticker symbol (usually 3 or 4 characters but sometimes less: Ford is just F) for the companies of interest. Type the symbol into the box at the top, then when you get the page for that stock, click "Historical Prices" on the left to get a historical time series of prices.
• The Berndt data are monthly returns to the individual stocks. To get monthly returns from Yahoo!, you should set the frequency to Monthly and set the time period in which you are interested. Once the data are shown on the screen, scroll to the bottom and click Download to Spreadsheet.
• Once you have the data in Excel, you probably want to clean them up there before transferring to Stata. To calculate the monthly return, you should use the "adjusted close" column of the data, which adjusts for dividend payments and stock splits. The monthly return is calculated as [P(t) - P(t-1)] / P(t-1), where P(t) is the (adjusted closing) price of the stock for month t. To get something corresponding to the MARKET variable in Berndt, you can do the same thing with one of the broad stock indexes. You can't get the same index that Berndt has, but there are several very broad indexes available by clicking the "View more indices" link underneath the Market Summary section. You might choose the Wilshire 5000 index (under the Other US tab) as comparable to Berndt's index.
• For your risk-free bond series, you can choose between two alternatives, both of I have prepared for you in the spreadsheet rkfree73-09.xls. The 3-month Treasury bill interest rate is r3mo (Column C) and is available in the spreadsheet for 1973-2009. The 4-week Treasury bill rate is r4wk (Column D) and is only available since July 2001. The two series do not differ that much. Use r4wk if you are using data only since July 2001 since this series is most closely symmetric to Berndt's series. Use r3mo if you need data from before July 2001. Don't mix them.
• Once you have the data organized in Excel, transfer to Stata is easy. Put the (Stata) variable name(s) that you want to use in the top row of the Excel spreadsheet with the data aligned in the column(s) below, then copy the cells containing the data (and the names). Then go to Stata, open the data editor, put the cursor in the top data cell of the first empty column, and hit paste. The data should be in your Stata dataset and usable. When doing the data entry this way, Stata will not make sure that the dates of your observations are correctly aligned. Thus, if you copy multiple series from Excel, make sure that the month corresponding to the top row is the same for all variables. (I like to have month and year variables on the left side of the speadsheet to make sure that I align things properly, even if I don't copy those columns into Stata.) For the rkfree series, you can either copy and paste these separately (delete all the rows between the header row and the start of your desired sample, and delete all the rows below the end date of your sample) or you can copy only the appropriate time interval into an empty column of a spreadsheet with your other series, aligning the dates correctly.

Project Teams

Project teams for the first assignment are below, with partners for this assignment shown in the rows of the table. The list is based on the 19 students who have attended class and are currently enrolled. With an odd number, one person will work alone on each project. If a student should drop the class this week, we'll pair up his or her former partner with the currently unmatched student if that is feasible. You should make contact with your partner as soon as possible to arrange a work schedule.