Risk and Return: The CAPM model

1
 Risk pricing

The two basic questions that every

risk and return model in finance
tries to answer are:
• How do you measure risk?
• How do you translate this risk measure
into a risk premium (risk pricing)

2
 A good risk and return model should…
1. It should come up with a measure of risk that applies to all
assets and not be asset-specific.
2. It should clearly delineate what types of risk are rewarded
and what are not, and provide a rationale for the
delineation.
3. It should come up with standardized risk measures, i.e., an
investor presented with a risk measure for an individual
asset should be able to draw conclusions about whether the
asset is above-average or below-average risk.
4. It should translate the measure of risk into a rate of return
that the investor should demand as compensation for
bearing the risk.
5. It should work well not only at explaining past returns, but3
also in predicting future expected returns.
 The Capital Asset Pricing Model

Uses variance of actual returns around an expected return

as a measure of risk.
Specifies that a portion of variance can be diversified
away, and that is only the non-diversifiable portion that
is rewarded.
Measures the non-diversifiable risk with beta, which is
standardized around one.
Translates beta into expected return -
Expected Return = Riskfree rate + Beta * Risk
Premium
Works as well as the next best alternative in most cases.
4
 The Mean-Variance Framework

 Investors who buy an asset expect to earn returns over the time horizon
that they hold the asset. Their actual returns over this holding period
may be very different from the expected returns, and it is this difference
between actual and expected returns that is a source of risk.
=>For example, expected returns of T-bills
 An investor buys the stocks of HPG. This investor, having done his
research, may conclude that he can make an expected return of 15%
over one-year holding period. The actual return over this period will
almost certainly not be equal to 15%; it might be much greater or much
lower.

5
Calculating risk and return using probabilities

Stock A Stock B
PA RA(%) PB RB(%)
0.05 10 0.08 12
0.25 15 0.24 18
0.4 22 0.45 28
0.25 25 0.18 32
0.05 30 0.05 38

6
Calculating risk and return using probabilities

 Mean return of a stock

μ= σni=1 PiRi
 Standard deviation

σ= ෍ Pi (R i −μ)2
i=1

7
 The Mean-Variance Framework

.Thevariance on any investment measures the disparity between

actual and expected returns

Low Variance Investment

High Variance Investment

Expected Return
 How risky is HAX? A look at the past…

30.00%

20.00%

10.00%

0.00%
Mar-17 Jun-17 Sep-17 Dec-17 Apr-18 Jul-18 Oct-18 Feb-19 May-19 Aug-19

-10.00%

-20.00%

-30.00%

9
-40.00%
Calculating risk and return using historical data

HAX SVC
Jan-18 11.32% -23.20%
Feb-18 -5.15% 2.67%
Mar-18 -3.36% 1.82%
Apr-18 -31.62% -2.05%
May-18 -13.78% -0.78%
Jun-18 7.48% -3.55%
10
Calculating risk and return using historical data

 Mean return of a stock

σni=1 𝑅𝑖
μ =
𝑛
 Standard deviation

σni=1(𝑅𝑖 −μ)2
σ=
𝑛

11
 The Effects of Diversification

Firm-specific risk can be reduced, if not eliminated, by

increasing the number of investments in your portfolio
(i.e., by being diversified). Market-wide risk cannot. This
can be justified on either economic or statistical grounds.
On economic grounds, diversifying and holding a larger
portfolio eliminates firm-specific risk for two reasons
(a)Each investment is a much smaller percentage of the portfolio,
muting the effect (positive or negative) on the overall portfolio.
(b)Firm-specific actions can be either positive or negative. In a
large portfolio, it is argued, these effects will average out to
zero. (For every firm, where something bad happens, there will
be some other firm, where something good happens.)

12
 The Effects of Diversification

The variance in a portfolio is partially determined by the variances of the

individual assets in the portfolio and partially by how they move together
The correlation coefficient measure how two asset move together. It is
ranged from -1 to 1
Consider a portfolio of two assets. Asset A has an expected return of μA
and a variance in returns of σ2A, while asset B has an expected return of
μB and a variance in returns of σ2B. The correlation in returns between the
two assets, which measures how the assets move together, is ρAB

13
 The Effects of Diversification

If ρAB = 1: No firm-specific risk can be diversified away

If ρAB = -1: All firm-specific risk will be diversified away
If ρAB = 0: No correlation between the two securities’ returns

14
 The Effects of Diversification

The expected return of the portfolio:

μ𝑝𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜=𝑤𝐴 ∗μ𝐴 +(1−𝑤𝐴 )∗μ𝐵
The variance of the portfolio:
σ2 𝑝𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜=𝑤𝐴 2σ2 +(1−𝑤 ) 2 σ2 +2𝑤 (1−𝑤 )σ σ ρ
A 𝐴 B 𝐴 𝐴 𝐴 𝐵 AB

where wA = Proportion of the portfolio in asset A

15
 The Effects of Diversification

Year S T
2016 6.6% 24.5%
2017 5.6% -5.9%
2018 -9% 19.9%
2019 12.6% -7.8%
2020 14% 14.8%

The correlation coefficient between the returns of two stocks = -0.389

 The variance of a portfolio…

Let consider 4 cases:

A = 80% S + 20% T
B = 60% S + 40% T
C = 40% S + 60% T
D = 20% S + 80% T
 Mean-Variance Models Measuring Market Risk
The CAPM model

The capital asset pricing model (CAPM) is based on several

assumptions:
+ Investors are rational and they are risk-averse
+ All information is freely available to investors
+ Investors hold diversified portfolios, eliminating all firm-specific
risk
+ Capital market are perfectly competitive: no transaction cost, no
taxes, numerous buyers and sellers, etc
…..
 The Market Portfolio

Assuming diversification costs nothing (in terms of transactions costs), and

that all assets can be traded, the limit of diversification is to hold a portfolio of
every single asset in the economy (in proportion to market value). This
portfolio is called the market portfolio.
Individual investors will adjust for risk, by adjusting their allocations to this
market portfolio and a riskless asset (such as a T-Bill)
Preferred risk level Allocation decision
No risk 100% in T-Bills
Some risk 50% in T-Bills; 50% in Market Portfolio;
A little more risk 25% in T-Bills; 75% in Market Portfolio
Even more risk 100% in Market Portfolio
A risk seeker Borrow money; Invest in market portfolio
Every investor holds some combination of the risk free asset and the market
portfolio.

Aswath Damodaran 19
 Mean-Variance Models Measuring Market Risk
The CAPM model

In the CAPM world, where all investors hold the market portfolio, the
risk to an investor of an individual asset will be the risk that this asset
adds to the market portfolio.
Intuitively, if an asset moves independently of the market portfolio, it
will not add much risk to the market portfolio. In other words, most of
the risk in this asset is firm-specific and can be diversified away.
In contrast, if the correlation coefficient between this asset and the
market portfolio is 1, it will add risk to the market portfolio. This asset
has more market risk and less firm-specific risk. Statistically, this
added risk is measured by the covariance of the asset with the market
portfolio
 Mean-Variance Models Measuring Market Risk
The CAPM model

Beta is a standardized measure of this covariance, obtained by dividing

the covariance of any asset with the market by the variance of the
market. It is a measure of the non-diversifiable risk for any asset can
be measured by the covariance of its returns with returns on a market
index, which is defined to be the asset's beta
 Mean-Variance Models Measuring Market Risk
The CAPM model

In particular, the expected return on an asset can be written as a

function of the risk-free rate and the beta of that asset
E(Ri ) = Rf + βi [E(Rm) – Rf]
where E(Ri ) = Expected return on asset i
Rf = Risk-free rate
E(Rm) = Expected return on market portfolio
βi = Beta of asset i
 Mean-Variance Models Measuring Market Risk
The CAPM model

To use the capital asset pricing model, we need three inputs:

 The riskless asset is defined to be an asset for which the investor
knows the expected return with certainty for the time horizon of
the analysis.
 The risk premium is the premium demanded by investors for
investing in the market portfolio, which includes all risky assets
in the market
 Stock returns over a period of time
 Limitations of the CAPM

1. The model makes unrealistic assumptions

2. The parameters of the model cannot be estimated precisely
- Definition of a market index
- Firm may have changed during the 'estimation' period'
3. The model does not work well
- If the model is right, there should be
a linear relationship between returns and betas
the only variable that should explain returns is betas
- The reality is that
the relationship between betas and returns is weak
Other variables (size, price/book value) seem to explain differences in returns better.

Aswath Damodaran 24
 Why the CAPM persists…

The CAPM, notwithstanding its many critics and limitations, has survived as
the default model for risk in equity valuation and corporate finance. The
alternative models that have been presented as better models (APM,
Multifactor model..) have made inroads in performance evaluation but not in
prospective analysis because:
• The alternative models (which are richer) do a much better job than the CAPM in
explaining past return, but their effectiveness drops off when it comes to
estimating expected future returns (because the models tend to shift and change).
• The alternative models are more complicated and require more information than
the CAPM.
• For most companies, the expected returns you get with the the alternative models
is not different enough to be worth the extra trouble of estimating four additional
betas.

Aswath Damodaran 25
 Regression Estimates of Betas

 The slope of the regression corresponds to the beta of the

stock and measures the riskiness of the stock
 The intercept of the regression provides a simple measure of
performance of the stock
Compare α and Rf (1-β)
 The R-squared: an estimate of the proportion of the risk of a
firm that could be be attributed to market risk; the balance
(1 – R2) could then be attributed to firm-specific risk
 The t-statistic or p-value

Aswath Damodaran 26