Ross, Westerfield, and Jaffe's Spreadsheet Master

Corporate Finance, 10th edition

by Brad Jordan and Joe Smolira
Version 10.0

Chapter 11
In these spreadsheets, you will learn how to use the following Excel fun

SQRT
COVAR
CORREL
Adding a trendline
Regression estimates
SLOPE
INTERCEPT

The following conventions are used in these spreadsheets:

1) Given data in blue

2) Calculations in red

NOTE: Some functions used in these spreadsheets may require that

the "Analysis ToolPak" or "Solver Add-In" be installed in Excel.
To install these, click on the Office button
then "Excel Options," "Add-Ins" and select
"Go." Check "Analysis ToolPak" and
"Solver Add-In," then click "OK."
 the following Excel functions:

eadsheets:

equire that
 Chapter 11 - Section 2
Expected Return, Variance, and Covariance

In Chapter 10, we used the AVERAGE, VAR, and STDEV functions to calculate the average, variance, and standard dev
have built-in functions that handle unequal probabilities, so we need to create our own equations.

Supertech

(5)
(1) (2) (3) (4) Deviation from
State of Probability of Return if State Product Expected Return
Economy State Occurs (2) × (3) (3) - E(R)
Depression 0.25 -0.20 -0.05 -0.375
Recession 0.25 0.10 0.025 -0.075
Normal 0.25 0.30 0.075 0.125
Boom 0.25 0.50 0.125 0.325
Expected return = 0.175

The standard deviation is the square root of the variance, so the standard deviation is:

Standard deviation: 25.86%

RWJ Excel Tip

Excel has a built-in function, SQRT, that finds the square root of a number. SQRT is found under the Math & Trig tab.

We should also note that the square root (or any other power) can be calculated using the caret key (^). For example

And for Slowpoke:

 Slowpoke

(5)
(1) (2) (3) (4) Deviation from
State of Probability of Return if State Product Expected Return
Economy State Occurs (2) × (3) (3) - E(R)
Depression 0.25 0.05 0.0125 -0.005
Recession 0.25 0.20 0.0500 0.145
Normal 0.25 -0.12 -0.0300 -0.175
Boom 0.25 0.09 0.0225 0.035
Expected return = 0.055

Standard deviation: 11.50%

To calculate the covariance and correlation, we need to calculate the product of the return deviations, multiply this p
then sum to find the covariance. Doing so, we find:

Deviation of Deviation of
Supertech Slowpoke
Return from the Return from the
State of Probability of Expected Expected Product of the
Economy State Return Return Deviations
Depression 0.25 -0.375 -0.005 0.001875
Recession 0.25 -0.075 0.145 -0.010875
Normal 0.25 0.125 -0.175 -0.021875
Boom 0.25 0.325 0.035 0.011375
Covariance =

Since the correlation is the covariance divided by the product of the standard deviations, the correlation between Su

Correlation: -0.1639

Covariance and Correlation with Historic Data

While we just discussed the calculation of covariance and correlation using unequal probabilities, both calculations a
historic data, Excel has built-in functions that will calculate the covariance and correlation for you.

Suppose we have the following returns for the market and a stock:

Year Market return Stock return

1 18% 7%
2 27% 25%
3 5% 21%
4 13% 4%
 5 -17% -16%
6 6% 19%
7 -21% -38%
8 34% 29%
9 19% 15%
10 11% 16%

What is the covariance and correlation of the returns between this stock and the market?

Covariance: 0.0281
Correlation: 0.8648

RWJ Excel Tip

The functions for covariance (COVAR) and correlation (CORREL) are both located under More Functions, Statistical. B
the data is located.

To use COVAR and CORREL, select the first data array, tab to Array2, and select the second data array. It is irrelevant w
between A and B is equal to the correlation between B and A.

A Quick Statistics
Covariance Review are measures of how much two variables move together. If two variables tend to vary tog
and correlation
value, then the other variable tends to be above its expected value too), then the covariance and correlation betwee
when one of them is above its expected value the other variable tends to be below its expected value, then the cova
negative.

The main difference between covariance and correlation is the interpretation. Covariance is an unstandardized numb
the two variables is large, or because of a strong relationship between the two variables. Thus, the only interpretatio
positive or negative.
Correlation is standardized and will be between -1 and 1. The closer the correlation is to -1, the stronger the negativ
correlation is to 1, the stronger the positive relationship between the two variables. Therefore, correlation measures
between two variables.

Correlation and Diversification

So why is correlation important to diversification? Correlation (and covariance) measure how two assets move toget
two assets, the greater the diversification benefit. If you think of two assets with a negative correlation, as one asset
return below its average. This will smooth out the returns of a portfolio of these two assets. However, if the assets h
its mean, the other asset will also have a return above its mean, so there is less benefit to diversification. For an app
manufacturers and would be expected to have a high correlation because many of the firm specific risks that would
share firm specific risk with Microsoft, so we would expect GM and Microsoft to have a lower correlation than GM a
benefit.
 average, variance, and standard deviation for historical returns. Unfortunately, Excel does not
our own equations.

(6) (7)
Squared Value Product
of Deviation (2) × (5)
0.140625 0.0351563
0.005625 0.0014063
0.015625 0.0039063
0.105625 0.0264063
Variance = 0.0668750

is found under the Math & Trig tab. The function looks like this:

d using the caret key (^). For example, we could have entered an equation as H13^(1/2).
 (6) (7)
Squared Value Product
of Deviation (2) × (5)
0.000025 0.0000063
0.021025 0.0052563
0.030625 0.0076563
0.001225 0.0003063
Variance = 0.0132250

the return deviations, multiply this product by the probability of the state of the economy, and

Probability of
State of the
Economy times
Product of the
Deviations
0.000469
-0.002719
-0.005469
0.002844
-0.004875

eviations, the correlation between Supertech and Slowpoke is:

qual probabilities, both calculations are often done using historic market data. When using
orrelation for you.
 under More Functions, Statistical. Both functions use similar inputs, namely the arrays that

he second data array. It is irrelevant which data array you select first. That is, the correlation

ther. If two variables tend to vary together (that is, when one of them is above its expected
e covariance and correlation between the two variables will be positive. On the other hand,
ow its expected value, then the covariance and correlation between the two variables will be

ovariance is an unstandardized number. A large covariance can arise because the variance of
variables. Thus, the only interpretation we can take from the covariance is the direction, either
tion is to -1, the stronger the negative relationship between the variables, and the closer the
bles. Therefore, correlation measures both the direction and magnitude of the relationship

measure how two assets move together. All else the same, the lower the correlation between
h a negative correlation, as one asset has a return above its average, the other asset will have a
two assets. However, if the assets have a positive correlation, as one asset has a return above
benefit to diversification. For an application, think of GM and Ford. Both are auto
of the firm specific risks that would affect GM also affect Ford. However, GM is less likely to
have a lower correlation than GM and Ford, and therefore have a greater diversification
 Chapter 11 - Section 4
The Return and Risk for Portfolios

In the textbook, the equation for the standard deviation of a portfolio is presented. Given the following information
deviation of the portfolio?

Stock A Stock B
Expected return 9% 14%
Standard deviation 19% 55%
Weight of stock 30% 70%

Correlation 0.10

The expected return and standard deviation of the portfolio are:

Expected return: 12.50%

Standard deviation: 39.48%

Of course, we could be interested in examining the opportunity set for the two assets. To see this, we can create a ta
expected return and standard deviation of the two assets for various portfolio weights is:

Expected Standard
Weight of Stock A Return Deviation

0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
 85%
90%
95%
100%

So what does the opportunity set for these two assets look like? Below, you will see. To examine how a change in the
correlation in the cell above.

Opportunity Set of Two Assets

1%
99 %
3
93 %
5
87 %
7
81 %
9
75 %
1
70 %
Total Return on Portfolio

3
64 %
5
58 %
7
52 %
9
46 %
1
41 %
3
35 %
5
29 %
7
23 %
9
17 %
1
12 %
63
5% 10% 210% 410% 610% 810% 1010%

Risk (Standard Deviation of Portfolio's Return)

So how do we find the minimum variance portfolio? The best way is to use Solver. Try this for yourself and see if you
portfolio is about 92.93%
en the following information concerning two stocks, what is the expected return and standard

To see this, we can create a table for various portfolio weights and then graph the results. The
is:
 examine how a change in the correlation will affect the shape of the opportunity set, change the

ets

810% 1010% 1210%

urn)

his for yourself and see if you don't agree that the weight of Stock A in the minimum variance
 Chapter 11 - Section 8
Market Equilibrium

In this section, you will learn how beta is estimated. Before we begin that discussion, we want to start with a graph o
month end values for the S&P 500, a common proxy for the market as a whole, and the adjusted closing price for Am
estimating beta, 60 monthly returns is a commonly used number of historical returns. Since we are going to be using
much data as possible. However, the further back in time we go, the less the company is like the current company. Fo
more than 100 years. But is AT&T in its current form actually comparable to AT&T in 1930? Not really. For this and ot
standard when estimating beta.

To begin, we would like to graph the returns of Amazon.com stock against the returns of the S&P 500. In this case, w

Performance of Amazon.com Stock and the S&P 500: Sing

40%

Ch a r
30%

20%
Amazon.com Return

f(x) = 0.9301086911x + 0.0213561093 10%

0%
-20% -15% -10% -5% 0% 5% 10%
-10%

-20%

-30%

-40%
S&P 500 Return

Notice that we have added a trend line in this graph. This trend line is called the characteristic line. The slope of this
market returns. The slope of this line is the beta of the stock.

RWJ Excel Tip

To add a trend line to a chart, do the following:
 1) Click anywhere in the chart. This displays the Chart Tools, adding the Design, Layout, and Format tabs.
2) On the Layout tab, in the Analysis group, click Trend line.
3) You can use any of the predefined options. Note that on the chart there is an equation. We went to More Op
the graph. We will have more to say about this equation later.

The equation in the graph above is a linear regression. We can use the trend line option on a graph to estimate a line
linear regression as well as give us more statistical information about the regression estimate.

RWJ Excel Tip

To estimate a linear regression, go to the Data tab, Data Analysis, and select Regression, then OK.

The input box for our linear regression looks like this:
The Y input range is the dependent variable, in this case the stock returns, and the X input range is the independent
the data and selected the Labels box, which will put a label on the output for the variables. Finally, we selected the C
confidence interval. The output for this regression is below.

SUMMARY OUTPUT

Regression Statistics
Multiple R 0.4763651752
R Square 0.2269237802
Adjusted R Square 0.2135948798
Standard Error 0.0952868398
Observations 60

ANOVA
df SS MS F
Regression 1 0.1545793713 0.1545793713 17.0249438688
Residual 58 0.526615747 0.0090795818
Total 59 0.6811951183

Coefficients Standard Error t Stat P-value

Intercept 0.0213561093 0.0123026699 1.735892249 0.0878929575
S&P 500 0.9301086911 0.2254191761 4.1261294052 0.0001193752

More Regression

If you are just interested in the slope and intercept for a regression, Excel has functions that will calculate these valu

Beta (slope): 0.93011

Intercept: 0.02136

RWJ Excel Tip

Both the SLOPE and INTERCEPT functions are located under More Functions, Statistical. The inputs for each function
the inputs we used for our results.
 sion, we want to start with a graph of actual stock returns. On the next tab, you will find the
and the adjusted closing price for Amazon.com stock over a 60 month period. When
turns. Since we are going to be using a statistical process to estimate beta, we would like as
mpany is like the current company. For example, with AT&T, we could get stock prices for
&T in 1930? Not really. For this and other reasons, 60 monthly returns has become relatively

eturns of the S&P 500. In this case, we used a scatter plot which resulted in the graph below.

ck and the S&P 500: Single Index Model

40%
lin e
c te ristic
C h a ra
30%

20%

10%

0%
0% 5% 10% 15% 20%
-10%

-20%

-30%

-40%
&P 500 Return

characteristic line. The slope of this line represents how the stock's returns respond to the
sign, Layout, and Format tabs.

is an equation. We went to More Options and selected the box to display the equation on

e option on a graph to estimate a linear regression, but Excel has a tool that will estimate a
sion estimate.

ression, then OK.

he X input range is the independent variable, or market return. We included the row above
e variables. Finally, we selected the Confidence Interval box and asked for a 90 percent

Significance F
0.0001193752

Lower 95% Upper 95% Lower 95.0% Upper 95.0%

-0.0032703601 0.0459825788 -0.0032703601 0.0459825788
0.478883185 1.3813341973 0.478883185 1.3813341973

nctions that will calculate these values separately, SLOPE and INTERCEPT.

atistical. The inputs for each function are the Y values and the X values. Below, you will see
Return Data

Date S&P 500 Amazon.com

12/3/2007 1468.36 $ 92.64 S&P 500 Amazon
1/2/2008 1378.55 $ 77.70 -6.12% -16.13%
2/1/2008 1330.63 $ 64.47 -3.48% -17.03%
3/3/2008 1322.70 $ 71.30 -0.60% 10.59%
4/1/2008 1385.59 $ 78.63 4.75% 10.28%
5/1/2008 1400.38 $ 81.62 1.07% 3.80%
6/2/2008 1280.00 $ 73.33 -8.60% -10.16%
7/1/2008 1267.38 $ 76.34 -0.99% 4.10%
8/1/2008 1282.83 $ 80.81 1.22% 5.86%
9/2/2008 1166.36 $ 72.76 -9.08% -9.96%
10/1/2008 968.75 $ 57.24 -16.94% -21.33%
11/3/2008 896.24 $ 42.70 -7.48% -25.40%
12/1/2008 903.25 $ 51.28 0.78% 20.09%
1/2/2009 825.88 $ 58.82 -8.57% 14.70%
2/2/2009 735.09 $ 64.79 -10.99% 10.15%
3/2/2009 797.87 $ 73.44 8.54% 13.35%
4/1/2009 872.81 $ 80.52 9.39% 9.64%
5/1/2009 919.14 $ 77.99 5.31% -3.14%
6/1/2009 919.32 $ 83.66 0.02% 7.27%
7/1/2009 987.48 $ 85.76 7.41% 2.51%
8/3/2009 1020.62 $ 81.19 3.36% -5.33%
9/1/2009 1057.08 $ 93.36 3.57% 14.99%
10/1/2009 1036.19 $ 118.81 -1.98% 27.26%
11/2/2009 1095.63 $ 135.91 5.74% 14.39%
12/1/2009 1115.10 $ 134.52 1.78% -1.02%
1/4/2010 1073.87 $ 125.41 -3.70% -6.77%
2/1/2010 1104.49 $ 118.40 2.85% -5.59%
3/1/2010 1169.43 $ 135.77 5.88% 14.67%
4/1/2010 1186.69 $ 137.10 1.48% 0.98%
5/3/2010 1089.41 $ 125.46 -8.20% -8.49%
6/1/2010 1030.71 $ 109.26 -5.39% -12.91%
7/1/2010 1101.60 $ 117.89 6.88% 7.90%
8/2/2010 1049.33 $ 124.83 -4.74% 5.89%
9/1/2010 1141.20 $ 157.06 8.76% 25.82%
10/1/2010 1183.26 $ 165.23 3.69% 5.20%
11/1/2010 1180.55 $ 175.40 -0.23% 6.16%
12/1/2010 1257.64 $ 180.00 6.53% 2.62%
1/3/2011 1286.12 $ 169.64 2.26% -5.76%
 2/1/2011 1327.22 $ 173.29 3.20% 2.15%
3/1/2011 1325.83 $ 180.13 -0.10% 3.95%
4/1/2011 1363.61 $ 195.81 2.85% 8.70%
5/2/2011 1345.20 $ 196.69 -1.35% 0.45%
6/1/2011 1320.64 $ 204.49 -1.83% 3.97%
7/1/2011 1292.28 $ 222.52 -2.15% 8.82%
8/1/2011 1218.89 $ 215.23 -5.68% -3.28%
9/1/2011 1131.42 $ 216.23 -7.18% 0.46%
10/3/2011 1253.30 $ 213.51 10.77% -1.26%
11/1/2011 1246.96 $ 192.29 -0.51% -9.94%
12/1/2011 1257.60 $ 173.10 0.85% -9.98%
1/3/2012 1312.41 $ 194.44 4.36% 12.33%
2/1/2012 1365.68 $ 179.69 4.06% -7.59%
3/1/2012 1408.47 $ 202.51 3.13% 12.70%
4/2/2012 1397.91 $ 231.90 -0.75% 14.51%
5/1/2012 1310.33 $ 212.91 -6.27% -8.19%
6/1/2012 1362.16 $ 228.35 3.96% 7.25%
7/2/2012 1379.32 $ 233.30 1.26% 2.17%
8/1/2012 1406.58 $ 248.27 1.98% 6.42%
9/4/2012 1440.67 $ 254.32 2.42% 2.44%
10/1/2012 1412.16 $ 232.89 -1.98% -8.43%
11/1/2012 1416.18 $ 252.05 0.28% 8.23%
12/3/2012 1402.43 $ 245.18 -0.97% -2.73%
Chapter 11 - Master it!

The CAPM is one of the most tested models in Finance. When beta is estimated in practice, a variation of CAPM called the market model is often use
market model, we start with the CAPM:

E(Ri) = Rf + b[E(RM) - Rf]

Since CAPM is an equation, we can subtract the risk-free rate from both sides, which gives us:

E(Ri) - Rf = b[E(RM) - Rf]

This equation is deterministic, that is, exact. In a regression, we realize that there is some indeterminate error. We need to formally recognize this in
adding epsilon, which represents this error:

E(Ri) - Rf = b[E(RM) - Rf] + e

Finally, think of the above equation in a regression. Since there is no intercept in the equation, the intercept is zero. However, when we estimate the
equation, we can add an intercept term, which we will call alpha:

E(Ri) - Rf = ai + b[E(RM) - Rf] + e

This equation, known as the market model, is generally the model used for estimating beta. The intercept term is known as Jensen's alpha and repre
return. If CAPM holds exactly, this intercept should be zero. If you think of alpha in terms of the SML, if the alpha is positive, the stock plots above th
alpha is negative, the stock plots below the SML.
 You want to estimate the market model for an individual stock and a mutual fund. First, go to finance.yahoo.com and download the adjusted prices f
months for an individual stock and a mutual fund, and the S&P 500. Next, go to the St. Louis Federal Reserve website at www.stlouisfed.org. You sho
a. FRED® database on this website. Look for the 1-Month Treasury Constant Maturity Rate and download this data. This will be the proxy for the risk-fre
using this rate, you should be aware that this interest rate is the annual interest rate, while we are using monthly stock returns, so you will need to a
month T-bill rate. For the stock and mutual fund you select, estimate the beta and alpha of the stock using the market model. When you estimate th
model, find the box that says Residuals and check this box when you do each regression. Because you are saving the residuals, you may want to save
output in a new worksheet.

1) Are the alpha and beta for each regression statistically different from zero?
2) How do you interpret the alpha and beta for the stock and the mutual fund?
3) Which of the two regression estimates has the highest R squared? Is this what you would have expected? Why?

b. In part a, you asked Excel to return the residuals of the regression, which is the epsilon in the regression equation. If you remember back to statistics
are the linear distance from each observation to the regression line. In this context, the residuals are the part of each monthly return that is not exp
market model estimate. The residuals can be used to calculate the appraisal ratio, which is the alpha divided by the standard deviation of the residua

1) What do you think the appraisal ratio is intended to measure?

2) Calculate the appraisal ratio for the stock and the mutual fund. Which has a better appraisal ratio?
3) Often, the appraisal ratio is used to evaluate the performance of mutual fund managers. Why do you think the appraisal ratio is used more often f
which are portfolios, than for individual stocks?
 market model is often used. To derive the

o formally recognize this in the equation by

ver, when we estimate the regression

as Jensen's alpha and represents the excess

e, the stock plots above the SML and if
nload the adjusted prices for the last 61
ww.stlouisfed.org. You should find the
be the proxy for the risk-free rate. When
urns, so you will need to adjust the 1-
del. When you estimate the regression
uals, you may want to save the regression

emember back to statistics, the residuals

nthly return that is not explained by the
ard deviation of the residuals.

l ratio is used more often for mutual funds,

 Master it! Solution

Mutual fund
a. Month/Year S&P 500 Stock price price Risk-free rate S&P 500 return
IBM FMAGX
 3) Since the mutual fund is a diversified portfolio, we would expect a higher R squared value. The diversification
unsystematic risk, therefore more of the return should be explained by the movement of the market.

b. 1) Since the residuals are the part of each return that is not explained by the regression line, the residuals are t
measure of the excess return per unit of systematic risk. A higher appraisal ratio is better since it is greater exce

2) The appraisal ratio for each regression estimate is:

IBM: Err:504
FMAGX: Err:504

IBM has the better appraisal ratio.

3) Fund managers, at least active fund managers, attempt to beat the market. The appraisal ratio measures the
excess return relative to the specific risk in the portfolio. Since the ratio is the return per unit of firm specific ris
that could be diversified away by holding the market portfolio. As such, it is more appropriate for a portfolio pe
measure.
 Mutual fund Market risk Stock risk Mutual fund risk
Stock return return premium premium premium
alue. The diversification of the portfolio has eliminated much of the
the market.

ine, the residuals are the systematic risk. As such, the appraisal ratio is a
since it is greater excess return per unit of risk.

sal ratio measures the performance of the manager by comparing the

unit of firm specific risk, in essence, it measures the return per unit of risk
riate for a portfolio performance measure than an individual asset risk