Homework Assignment 2
Calculating growth rates and betas for financial data
series
You should complete this homework assignment before the date
indicated on the homework assignments page. You are allowed to ask for help
from other students or to compare answers with other students, although you
should try to do as much as possible of this on your own.
This assignment should take between one and two hours to
complete.
This is a detailed assignment that requires you to use an
Excel Workbook labeled Beta Calculation HW and to retrieve and use raw
data from Yahoo’s stock price historical data source at http://finance.yahoo.com.
The purpose of this homework is to show you how to, using a
data series, (a) calculate real continuous growth rates, and (b) calculate and
compare a relative measures of risk,
the Beta.
Note: Keep this sheet and data in good
order. We will eventually use it for more advanced risk and other calculations.
You will want to review the formula for the Beta
that I have provided to you in Lecture 2 Compounding
Discounting and Risk. For much of
this I am not going to tell you what you need to do. I want you to figure it out. You will have to use many of Excel’s
built in functions to do this.
Here are the steps to completing your homework.
- Open
the Excel workbook labeled Beta Calculation HW and look sheet
labeled Beta Calculation.
This will be your working sheet. Although there is no data this
sheet is already formatted so be careful not to change the formatting
of cells in this sheet. On this sheet, you are going to download two years
of weekly data for the S&P 500 ETF SPY and for three stocks, Dendreon
DNDY, Johnson and Johnson JNJ, and a stock that you choose. Then you are
going to calculate continuous growth rates and Betas relative to
the S&P500 for your three stocks.
- Now go
to the Yahoo finance at http://finance.yahoo.com.
[Note: As of the date this homework was assigned, you cannot complete this
assignment using Google Chrome because it does not allow you to download
to an Excel spreadsheet. So use IE or compatible]. On this site you will
need to find historical prices for the stocks. The format of yahoo’s
finance site changes often but at the time this was prepared you would do
the following: On the Investing tab at the top of the home page
select Stocks. On the page that appears, select Historical
Quotes under Research Tools. A default page should appear
showing historical prices for a stock like General Electric GE. You are
going to use this page to download your data.
- We are
going to normalize our Beta calculations to the S&P500. But
rather than use actual S&P500 data, we are instead going to use data
for the ETF tracking stock SPY, which is priced at 10% of the S&P500
quotation (arbitrage keeps it very close to this value). So your first step
will be to override the GE default historical prices with historical
prices for SPY (don’t worry about the date range yet). Do that and make
sure that you are looking for data for SPY.
- For
all of the data, use a starting date of
September 6, 2006 and an ending date of September 2, 2008. Set
those dates and check the Weekly box for the data series, then Get
Prices. Weekly data beginning on September 2, 2008 working backwards
will appear (the first closing value should be around 124).
- On the
bottom of the page you will see a link that says Download to
Spreadsheet. Use that link to download this data to a blank Excel
spreadsheet (do not load the data into Beta Calculation HW spreadsheet
– it is too easy to mess up that spreadsheet). Check your data in the new
spreadsheet. There should be 105 observations with the observation dates
clearly labeled.
- You
will want to copy some of this data to the appropriate labeled columns in
the Beta Calculation HW spreadsheet. You are going to copy only the
date and Adj Close columns (Adjusted Close adjusts historical data
for splits and dividend payments). There is a problem though. The data are
listed from most current to oldest as you go down the column and because
we are going to calculate historical growth rates you need to invert this
column. Since the only data that you are going to copy over are the dates
and Adj Closes you might minimize mistakes by deleting the middle
five columns. Don’t eliminate the date column. Then using the Sort function,
sort all columns by sorting the date column as Ascending. This
should flip your data upside down. Check and make sure that your most
current datum is at the bottom of its column and that the SPY value of
around 124 is down there with it. Obviously this homework is gibberish if
you don’t do this right. Copy the date and SPY Adj Close columns to
their appropriate labeled columns in the Beta Calculation HW spreadsheet.
You should still have 105 dates and observations. Although eventually you
are going to repeat this procedure for DNDN, JNJ, and the stock that you
choose, you will next do some operations on the data that you have
downloaded for SPY.
- First,
in the adjacent appropriate column, use Excel’s LN function and the COPY
and PASTE feature to calculate the natural logs of the historical data for
SPY.
- As
discussed in Lecture 2 Compounding Discounting and
Risk take the difference in logs for each observation to calculate the
104 observations for the continuous weekly growth rates in the
column labeled CGR (this column should be set to default at four
decimal places). Your first CGR number should have a value like 0.0172,
which you would interpret to mean that the continuous weekly growth rate
is 1.7%. If we were using discrete rather than continuous growth rates we
would take the geometric mean, but we don’t have to do that here. Take a
look at the data. Remember, these data are not annualized. (Also remember
that the market, although volatile, was pretty flat during this period).
- At the
bottom of the spreadsheet you will see a row labeled AVERAGE. Use
Excel’s AVERAGE function to calculate the average weekly continuous
compounding growth rate (i.e. the geometric mean, given how we have set
this up) of SPY in the yellow cell in the CGR column. You
should get an answer that is very close to one (because the market was
static on net during this period).
- Below
this you will find a row labeled ANN AVG. To calculate the annual
average take value in the previous cell and multiply it times 52. If you have
done everything right, you should get a negative number near zero for an
answer. Although trivial here, this won’t be the answer for your three
stocks. Note also that this “return” on an investment in SPY does not
include dividends, which exist for SPY and you may have noticed them when
you accessed the original data before download.
- Now
using the Excel VAR and STDEV commands, calculate the variance and
standard deviation of the continuous growth rate. The standard deviation,
which will equal the square root of the variance, should have a value of
around 0.02. Note that we have calculated the variance of the continuous
growth rate and not the unconverted price data.
- Now
find and download the data for C, DNDN, and the stock of your choice, then
calculate the same values for all of them and compare them to SPY. What
does this tell you?
- Using
SPY as the base, and using the formula provided in Lecture
2 Compounding Discounting and Risk calculate the covariances (either
using the Excel COV formula or putting in the formula yourself) and Betas
of the three stocks. My two
examples should clearly provide an example of a high Beta and a low
Beta and that is the result you should get.
- How do
you interpret these Betas? You will note that the Beta calculation
for C looks about right, but the same for DNDN does not. Look at the price
history for DNDN. It’s hugely volatile (look at April of 2007). There must
be some deficiency with the Beta and there is, which will be
discussed in class. (Hint: the correlation coefficient between the two is
0.09). We will need to find better measures of volatility.
This homework is complete. You may have noticed that we
didn’t use the time columns in this homework. Keep the spreadsheet
because we will return to it later in the class to make some more sophisticated
risk and volatility estimates.