Lecture 3: CAPM in practice

Investments

FIN460-Papanikolaou

CAPM in practice

1/ 59

Overview

1. The Markowitz model and active portfolio management.

2. A Note on Estimating
3. Using the single-index model and the CAPM in the passive
portfolio problem
3.1 Using to get good covariance estimates
3.2 Using to get good expected-return estimates

4. The Black and Litterman Method.

, An Example: Global Asset Allocation

5. Conclusions

FIN460-Papanikolaou

CAPM in practice

2/ 59

Using the Markowitz model

With a reasonable number of securities, the number of

parameters that must be estimated is huge:

, For a portfolio of N (100) securities we need:

i s
E(ri )s
Cov(ri , r j )s
Total

N
N
1
N(N
1)
2
1
N(N
+ 3)
2

100
100
4950
5150

, About how much data will we need when we have 500 securities?
1000 Securities?

Means and covariances are estimated with error.

Small errors in mean or covariance estimates often lead to
unreasonable weights.

FIN460-Papanikolaou

CAPM in practice

3/ 59

Using the Markowitz model

A remedy for both of these problems is:

1. First, we use
I

the CAPM to determine what market believes expected returns

should be

a single factor model to calculate asset covariances

2. then we can combine our views with the CAPM-derived

estimates to get portfolio weights.

The key input we will need for both of these is the set of asset s.
So, first, we must consider the problem of estimating s.

FIN460-Papanikolaou

CAPM in practice

4/ 59

Estimating
1. Let ri,t , rm,t and r f ,t denote historical individual security, market
and risk free asset returns (respectively) over some period
t = 1, 2, . . . , T .
2. The standard way to estimate a beta is to use a
characteristic line regression:

ri,t r f ,t = i + i (rm,t r f ,t ) + i,t

3. To estimate this, typically we would use monthly data
4. Typically use 5 years (60 months) of data.
, Why not use more data? (10 or 20 years)
I

Parameter instability

, Can we use weekly, daily, or intraday data?

I
I

Non-synchronous prices
Bid-ask bounce

FIN460-Papanikolaou

CAPM in practice

5/ 59

Estimating - Example
Here is an example of a (12 month) characteristic-line regression for
GM Common stock (from BKM, Ch. 10):

FIN460-Papanikolaou

CAPM in practice

6/ 59

Estimating - Example
This figure illustrates this still better:

i , i , and i,t off of this scatterplot?

How do we read
FIN460-Papanikolaou

CAPM in practice

7/ 59

Estimating
There are a number of Institutions that supply s:
1. Value-Line uses the past five years (with weekly data) with the
Value-Weighted NYSE as the market.
2. Merrill Lynch uses 5 years of monthly data with the S&P 500 as
the market

FIN460-Papanikolaou

CAPM in practice

8/ 59

Estimating

Merrill uses total rather than excess returns:

ri,t = ai + bi rm,t + ei,t

So, the ai here is equal to:

ai = i + (1 i )r f
, where r f is the average risk-free rate over the estimation period
, ai is not equal to zero if the CAPM is true.
, However, bi and i will be very close, as long as cov(r f ,t , rm,t ) 0

FIN460-Papanikolaou

CAPM in practice

9/ 59

Estimating

Note also that Merrill used adjusted s, which are equal to:
Ad j

1/3 + (2/3) i

, Intuition?
I

Statistical Bias

Why 1/3 and 2/3 ?

What should these numbers be based on?

Other more advanced shrinkage techniques.

FIN460-Papanikolaou

CAPM in practice

10/ 59

Estimating for new companies

In the absence of historical data, how would we estimate for

new companies?
Standard industry practice is to use comparables:

, Find a similar company, that is traded on an exchange and use the

beta of that company.

The following approach is in the same spirit but is more powerful.

, Often a comparable company cannot be found.

, Can yield a model-predicted beta from a sample of companies
with similar characteristics.

FIN460-Papanikolaou

CAPM in practice

11/ 59

Estimating

Which characteristics to use?

, Industry
, Firm Size
, Financial Leverage
, Operating Leverage
, Growth / Value

FIN460-Papanikolaou

CAPM in practice

12/ 59

Estimating for new companies - Methodology

1. Select a sample of companies to estimate the model.
2. Estimate i,t for these companies using historical return data.
3. Regress estimated betas i,t on several characteristics that can
drive betas. Example:

i,t = a0 + j INDUST RY j,i +a1 FLEVi,t +a2 SIZEi,t +a3 OLEVi,t +ui,t
j

, INDUST RY j,i is a dummy variable that take the value 1 if firm i

belongs in industry j and 0 otherwise.
4. Suppose you are asked to find the beta for new company XYZ in
the tech industry, given XYZs characteristics. Your estimate will
be

xyz,t = a0 + tech + a1 FLEVxyz,t + a2 SIZExyz,t + a3 OLEVxyz,t

FIN460-Papanikolaou

CAPM in practice

13/ 59

Estimating
Betas may change over time, thus we use short windows of data
(5-years)
Possible reasons for this are
1. changes in the firms leverage
2. changes in the type of a firms operations
3. the firm acquires targets in other industries

We can use rolling window regressions to estimate a time series

of s:

, at month t estimate using months t 60 through t 1

, at month t + 1 estimate using months t 59 through t .
Alternatively, use more sophisticated statistical models that allow
for time-variation in beta.
FIN460-Papanikolaou

CAPM in practice

14/ 59

Estimating
s can be quite volatile.
Example: AT&T

FIN460-Papanikolaou

CAPM in practice

15/ 59

Estimating
s of industries also changes over time.
Example: Oil Industry
50
Oil industry market beta
2

Real Oil Price (1980=100)

45

40

35

1.5

30

25
1
20

15
0.5
10

FIN460-Papanikolaou

CAPM in practice

2004

2002

2000

1998

1996

1994

1992

1990

1988

1986

1984

1982

1980

1978

1976

1974

1972

1970

1968

1966

1964

1962

16/ 59

The CAPM and Active Portfolio Management

To create an optimal portfolio we need to estimate the minimum

variance/efficient frontier, MVE portfolio, and the best capital
allocation line (CAL).
Where are the inputs coming from?
1. One way to do this is to ignore market opinion, and to estimate the
E(ri )s and i, j s for all assets.
I

For example, one could use past empirical data.

we have already seen that there are problems with this approach

2. An alternative approach is to accept market opinion and simply

hold the market portfolio.
I

This approach doesnt tell you what to do if you have information

that you believe has not yet been incorporated into market prices.

FIN460-Papanikolaou

CAPM in practice

17/ 59

The CAPM and Active Portfolio Management

The approach we will instead take is to
1. Calculate s for the securities we plan to hold using the methods
discussed earlier .
2. Using these s, calculate i, j s and E(ri )s, assuming the CAPM
holds exactly.
3. Then, incorporate our information by perturbing the values away
from the CAPM-calculated values.
4. Finally, using these estimates and the Markowitz portfolio
optimization tools (from Lecture 2), determine our optimal portfolio
weights.

Note that if
(i) We start with all of the assets in market portfolio
(ii) We use the unmodified i, j s and E(ri )s we get from step 2

Then the weights we calculate in step D.4 will be exactly those of

the market portfolio. Why??

FIN460-Papanikolaou

CAPM in practice

18/ 59

A single index model

An alternative to estimating i, j s is to assume that a single index
model describes returns:
1. ri,t r f ,t = i + i (rm,t r f ) + i,t
2. for two different securities i and j,

cov( i,t , j,t ) = 0.

I
I

The CAPM does not imply this.

Is this a reasonable simplification?

To get covariances/correlations, use:

i, j = cov(ri,t , r j,t ) = i j 2m i 6= j
and

i, j =

i j 2m
i j

For variances:

i,i = 2i = 2i 2m + 2,i
FIN460-Papanikolaou

CAPM in practice

19/ 59

The single index model vs the CAPM

1. The single index model is a statistical model.

, It specifies that all common movements between stocks can be

captured by a single index.

, It is also often called a single factor model.

, Can be generalized to multiple common factors.
, It is a statistical technique designed to estimate large covariance
matrices.

2. The CAPM is an economic model of expected returns.

, It specifies that the market portfolio captures systematic risk.

, However, it does allow securities i and j to be correlated on top of
what their covariance with the market portfolio implies, for
instance if they are in the same industry. The only requirement is
that these correlations wash out as we add more stocks.

FIN460-Papanikolaou

CAPM in practice

20/ 59

The CAPM and Active Portfolio Management

To get the markets beliefs about expected returns, use the
CAPM:

E(ri ) r f = i (E(rm ) r f )
Remember, this is estimated by the regression

ri,t r f ,t = i + i (rm,t r f ) + i,t

, Given our estimates i , i and E(rm,t r f ), if we calculate
i + i E(rm,t r f )
it will be the same as estimating E(ri,t r f ,t ) using historical
averages. Why?

, The CAPM estimate of E(ri,t r f ,t ) is obtained by imposing

i = 0.
FIN460-Papanikolaou

CAPM in practice

21/ 59

The CAPM and Active Portfolio Management

Then, to construct the efficient frontier based on the Single Index
Model (for 100 securities) we need estimates of the following:

rf
E(rm )
2m
i
i
2,i

1
1
1
N
N
N

1
1
1
100
100
100

Total

3N + 3

303

This is considerably smaller than the 5150 we had before.

Where are the E(ri )?
Where are the i, j = Cov(ri , r j )?
What will the efficient portfolio weights be if we assume all i are
zero? (and include all assets in our analysis)
FIN460-Papanikolaou

CAPM in practice

22/ 59

Example (cont)
Monthly data for GE, IBM, Exxon (XON), and GM:
Excess Returns
mean
std
IBM
XON
GM
GE
VW-Rf
Rf

3.22%
1.41%
0.64%
2.26%
1.67%
0.36%

8.44%
4.03%
7.34%
5.86%
4.02%
0.05%

alpha

beta

std

R2ad j

1.31%
0.42%
-1.06%
0.53%

1.14
0.59
1.02
1.04

7.13%
3.28%
6.14%
4.15%

28.5%
33.7%
30.0%
49.9%

, VW is the Value-Weighted index of all NYSE, AMEX, and

NASDAQ common stocks.

, Rf is the (nominal) 1-month T-Bill yield, 4.394%/year

(0.359%/month)

FIN460-Papanikolaou

CAPM in practice

23/ 59

Example (cont)

For expected returns, lets not impose the CAPM just yet.
To get the correlation structure, we use:

i, j =

IBM
XON
GM
GE

i j 2m
i j

IBM
1
0.32
0.30
0.39

FIN460-Papanikolaou

XON
0.32
1
0.33
0.42

i 6= j

GM
0.30
0.33
1
0.40

CAPM in practice

GE
0.39
0.42
0.40
1

24/ 59

Example (cont)

Plugging the (1) expected returns, (2) return standard

deviations, and (3) correlation matrix into the Excel spreadsheet
we get the following weights for the tangency portfolio:
Weight
IBM
XON
GM
GE

29.6 %
49.7 %
-21.4 %
42.4 %

It seems unreasonable that we should hold such extreme

portfolio positions.
, The equilibrium arguments we used in developing the CAPM
suggest that the market knows something we dont about future
expected returns!

FIN460-Papanikolaou

CAPM in practice

25/ 59

Example (cont)
Lets use the CAPM as a way of getting around this problem.
This is equivalent to setting i = 0 for all securities, or using the
regression equation:

E(ri ) = r f + i [E(rm ) r f ]
and the past (average) return on the market to get equilibrium
estimates of the expected returns.
Stock
IBM
XON
GM
GE

CAPM E(rie )
1.91 %
0.99 %
1.70 %
1.73 %

FIN460-Papanikolaou

CAPM in practice

26/ 59

Example (cont)

With this equilibrium set of expected returns, we now get the

portfolio weights:
weights
IBM
XON
GM
GE

13.6%
33.3%
16.4%
36.6%

, Why are the weights different?

, Why are these not the market weights? When would these be the
actual market weights?

, Is this the portfolio you would want to hold, given that you were
constrained to hold these four assets?

FIN460-Papanikolaou

CAPM in practice

27/ 59

Example (cont)

However, there may be times when we think that the market is a

little wrong along one or more dimensions (a very dangerous
assumption!)
How can we combine our views with what the market expects?
1. First, suppose that I think that the market has underestimated
the earnings that IBM will announce in the next month, and that
IBMs expected return is 2% higher than the market expects.
2. Also, I have no information on the other three securities that would
lead me to think that they are mispriced,
3. and I believe that the past betas, and residual std devs are good
indicators of the relative future values.

FIN460-Papanikolaou

CAPM in practice

28/ 59

Example (cont)
In this case, we would use the same variance and covariance
inputs, but would change the expected returns. The new portfolio
weights would be:
Stock
IBM
XON
GM
GE

E(rie )
3.91%
0.99%
1.70%
1.73%

weights
54.1%
17.7%
8.7%
19.5%

compared to the old allocation

IBM
XON
GM
GE
FIN460-Papanikolaou

E(rie )

weights

1.91%
0.99%
1.70%
1.73%

13.6%
33.3%
16.4%
36.6%

CAPM in practice

29/ 59

Example (cont)

Alternatively, suppose that I think the risk () of Exxon is

increasing.
1. I guess that Exxons will rise from 0.59 to 0.8.
2. I also expect that Exxons idiosyncratic risk will not change.

First, I should recalculate almost everything using the equations:

E(ri ) = r f + i [E(rm ) r f ]
2i = 2i 2m + 2,i
i, j =

FIN460-Papanikolaou

i j 2m
i j

CAPM in practice

30/ 59

Example (cont)
The new correlations we come up with are:
IBM
XON
GM
GE

IBM
1.00
0.44
0.30
0.39

XON
0.44
1.00
0.45
0.57

GM
0.30
0.45
1.00
0.40

GE
0.39
0.57
0.40
1.00

as opposed to the old correlation matrix of:

IBM
XON
GM
GE

IBM
1
0.32
0.30
0.39

FIN460-Papanikolaou

XON
0.32
1
0.33
0.42

GM
0.30
0.33
1
0.40

CAPM in practice

GE
0.39
0.42
0.40
1

31/ 59

Example (cont)

If, I believe that these are the new correlations, but that the
market still believes that the past correlations represent the future
(and will not discover this information over the next several
months) then I would use the old expected returns, giving new
portfolio weights of:

E(rie )
IBM
XON
GM
GE

old
weight

1.91%
0.99%
1.70%
1.73%

FIN460-Papanikolaou

13.6%
33.3%
16.4%
36.6%

CAPM in practice

new
weight
17.0%
16.7%
20.5%
45.7%

32/ 59

Example (cont)
If, I believe that market knows that the of Exxon is higher, and
the expected return on Exxon is higher now to compensate for
the increased risk:

E(rie )
IBM
XON
GM
GE

old
weight

1.91%
0.99%
1.70%
1.73%

13.6%
33.3%
16.4%
36.6%

new
weight

E(re )

1.91%
1.34%
1.70%
1.73%

9.2%
54.8%
11.1%
24.8%

One other alternative is that I dont believe that the market yet
knows that the risk of Exxon is higher, but will discover this in the
next few months.
1. What will happen as the market finds out?
2. What should I do in this case?

FIN460-Papanikolaou

CAPM in practice

33/ 59

Forming views

Where is our information coming from?

1. It could come from public sources of information that may or may
not have been impounded in prices yet.
2. It could be private, that is it could come from our own analysis of
fundamentals.

Nevertheless, we may not be 100% confident that our information

is accurate.
For instance, in the previous example, if I am only somewhat
confident of my belief that the market will not adjust the price of
Exxon properly, I might want to adjust the portfolio weights only
part-way.

FIN460-Papanikolaou

CAPM in practice

34/ 59

Forming views - Example

Consider the Morningstar report on AIG:

FIN460-Papanikolaou

CAPM in practice

35/ 59

Forming views - Example

1. Market price: P0 = 58.95

2. Your view:
2.1 V0 = (1 + a)P0 = 65.00 a = 10.26%
2.2 rCAPM = [E(P1 ) V0 ]/V0
3. You think the actual return will be:

r = [E(P1 ) P0 ]/P0
= [E(P1 ) V0 +V0 P0 ]/P0
= rCAPM (V0 /P0 ) + [V0 P0 ]/P0
= rCAPM + a + rCAPM a
rCAPM + a

FIN460-Papanikolaou

CAPM in practice

36/ 59

Forming views - Example

If you were 100% confident in your view, then you would also
think that AIGs expected return will be higher than the CAPM
implied return by an amount a = 10.26%.
This is a very large number and will likely lead to extreme
portfolio allocations.
Suppose that you are only 10% confident in your view, then you
might use a = 0.1 10.26% = 1.26%.
NEXT: The Black and Litterman model gives a systematic
framework that enables you to incorporate your views with what
the markets views in forming the optimal portfolio.

FIN460-Papanikolaou

CAPM in practice

37/ 59

Black-Litterman model - Global Asset Allocation

Focus on asset allocation between country indices.
Use an international version of the CAPM.
Calculate the equilibrium expected returns based on estimated
variances and covariances. This is appropriate, since
covariances can be estimated accurately, especially using daily
data, for a small number of assets.
The portfolio manager can specify any number of market views"
in the form of expected returns, and a variance (measure of
uncertainty) for each of the views.
, If the manager holds no views, she will hold the market portfolio.
, If her views are high variance (low certainty), she will hold close to
the equilibrium portfolio.

, When her views are low variance, she will move considerably
away from the market portfolio.

FIN460-Papanikolaou

CAPM in practice

38/ 59

Black-Litterman model - Global Asset Allocation

First, lets look at the sort of portfolio allocation we get if we use
historical returns and volatilities as inputs:

FIN460-Papanikolaou

CAPM in practice

39/ 59

Black-Litterman model - Global Asset Allocation

also, we use historical correlations:

FIN460-Papanikolaou

CAPM in practice

40/ 59

Black-Litterman model - Global Asset Allocation

If you use our procedures from Lecture 2 and calculate and optimal
portfolio, with p = 10.7%, you will get portfolio weights of:

FIN460-Papanikolaou

CAPM in practice

41/ 59

Black-Litterman model - Global Asset Allocation

CAPM Based Estimates
assuming market risk premium=7.15%

FIN460-Papanikolaou

CAPM in practice

42/ 59

Black-Litterman model - Global Asset Allocation

There are N assets in the market, returns are normally
distributed:

R N(, )
is an unknown quantity. The CAPM estimate says
N(, )
What this means is that the investor is uncertain about what the
expected returns are. He will form his best guess or in Bayesian
terminology his posterior beliefs, by combining information from
two sources:
, The first, his prior , will be based on the CAPM. The variance of
his prior, , denotes how much confidence the investor has on
the CAPM. High values of means that he attaches less weight to
the CAPM.

, The second will be his own views, as we will see next.

FIN460-Papanikolaou

CAPM in practice

43/ 59

Black-Litterman model - Global Asset Allocation

Suppose that we believe that Germany will have an expected

return of 4%
Black-Litterman allows you to express this view as

ger = q + ,

N(0, )

Here, q = 4% is your belief about Germanys expected return.

The variable allows for a margin of error in your view.
Your confidence in your view is represented by , i.e. the
variance in your margin of error.
The higher is, the less confident we are.

FIN460-Papanikolaou

CAPM in practice

44/ 59

Black-Litterman model - Global Asset Allocation

The Black-Litterman model also allows you to form relative views.

Suppose that we believe that Germany will outperform France by

5%
Black-Litterman allows you to express this view as

P = Q + ,

N(0, )

Here, P is a matrix, for instance if there are only two countries

and you only form a relative view, it will equal


P=

1 1
0 0

As before, is the variance of your views.

FIN460-Papanikolaou

CAPM in practice

45/ 59

Black-Litterman model - Global Asset Allocation

Given our views and the CAPM prior, our best guess about what
expected returns are is is going to be a weighted average of the
CAPM expected returns, and our views, Q.

= [()1 + P0 1 P]1 [()1 + P0 1 Q]

In the case of one asset with prior and one view, q, it simplifies
to

1/(2 )
1/
+q
2
1/ + 1/( )
1/ + 1/(2 )

FIN460-Papanikolaou

CAPM in practice

46/ 59

Forming posterior beliefs - Example

Suppose that you would like to for your best guess about GEs
expected excess return going forward.

, GE has a beta of 1.2 and a standard deviation of 0.33. Assuming

the equity premium is 6%, this gives you (in excess of the risk-free
rate)
capm

GE = 1.26% = 7.2%

capm

and var(GE ) = 0.332 = 0.1

, You estimated the historical average return of GE to be 8.3% over

the last 10 years. This gives you another estimate or view:

hist
GE = 8.3%

and var(hist
GE ) =

FIN460-Papanikolaou

CAPM in practice

0.33

10

2

47/ 59

0.01

Forming posterior beliefs - Example

Your best guess will be:

GE

1
capm
var(GE )

1
hist
capm
GE + var(hist ) GE
GE

1
capm
var(GE )
1
0.1

1
var(hist
GE )

1
7.2% + 0.01
8.3%
1
0.1

1
+ 0.01

It is a weighted average of your prior (the CAPM) and your view

(the historical average return).
The relative weights depend on the precision (the inverse of the
variance) of the signals.
Your confidence in the CAPM determines .
FIN460-Papanikolaou

CAPM in practice

48/ 59

Forming posterior beliefs - Example

Your posterior belief as a function of
0.084

0.082

GE

0.08

0.078

0.076

0.074

0.072
0

0.5

FIN460-Papanikolaou

1.5

CAPM in practice

2.5

49/ 59

Black-Litterman model - Global Asset Allocation

1.4
prior (CAPM)
best guess
view

1.2

0.8

0.6

0.4

0.2

0
5

FIN460-Papanikolaou

bar

CAPM in practice

50/ 59

Black-Litterman model - Global Asset Allocation

Finally, the optimal portfolio weights will be:

w = wMKT + P0
where P are the investors view portfolios, and is a
complicated set of weights. These weights have the following
properties:
1. The higher the expected return on a view, q, the higher the weight
attached to that view, .
2. The higher the variance of a view, , the lower the absolute value
of the weight attached to the view.

If you hold no view on an asset, then your optimal allocation will

be the market weight.

FIN460-Papanikolaou

CAPM in practice

51/ 59

Black-Litterman model - Global Asset Allocation

Suppose that we believe that Germany will outperform France
and UK.

FIN460-Papanikolaou

CAPM in practice

52/ 59

Black-Litterman model - Global Asset Allocation

The optimal allocations according to Markowitz is:

FIN460-Papanikolaou

CAPM in practice

53/ 59

Black-Litterman model - Global Asset Allocation

Our view is that Germany will outperform France and UK.
However, our best guess, , will be shifted for all countries.

FIN460-Papanikolaou

CAPM in practice

54/ 59

Black-Litterman model - Global Asset Allocation

The Black-Litterman model internalizes the fact that assets are
correlated. Your views about Germany vs France and UK
translate into implicit views about other countries.

FIN460-Papanikolaou

CAPM in practice

55/ 59

Black-Litterman model - Global Asset Allocation

Lets add a view: Canada will outperform US by 3%:

FIN460-Papanikolaou

CAPM in practice

56/ 59

Black-Litterman model - Global Asset Allocation

weights depend
How bullish is the view, i.e. (Q)
How confident is the investor in the view ().

FIN460-Papanikolaou

CAPM in practice

57/ 59

Black-Litterman model vs Markowitz

Black Litterman
1. The optimal portfolio equals the (CAPM) market portfolio plus a
weighted sum of the portfolios about which the investor has views.
2. An unconstrained investor will invest first in the market portfolio,
then in the portfolios about which views are expressed.
3. The investor will never deviate from the market weights on assets
about which no views are held.
4. This means that an investor does not need to hold views about
each and every asset!

Markowitz
1. Need to estimate vector of expected returns for all assets.
2. Estimation error often leads to unrealistic portfolio positions
3. If we change the expected return for one asset, this will change
the weights on all assets.

FIN460-Papanikolaou

CAPM in practice

58/ 59

Summary

In order to use the CAPM in our mean-variance analysis, we

need estimates of . Regression analysis yields that.
The CAPM assumes that all investors hold the same views so all
of them hold the market-portfolio.
The Black-Litterman model is one way to bring the information
from the CAPM in our asset allocation decision.

FIN460-Papanikolaou

CAPM in practice

59/ 59