BFW1001 Foundations of Finance

Lecture 3
Risk, Return and Cost of Capital
Dr. Keshab Shrestha, CFA
Professor of Banking and Finance
keshab.shrestha@monash.edu

Contents
1 Expected Return and Risk 2

2 Behavior under Uncertainty 4

3 Valuation under Uncertainty 5

4 Cost-of-Capital for Common Stock 7

4.1 Estimation of CAPM Beta . . . . . . . . . . . . . . . . . . . . . . . . 8

1
 BFW1001 Foundations of Finance Lec. 3

In this lecture, we discuss the valuation of financial securities where the associated
future cash flows are uncertain.
In Lecture 1, we considered the fair value (V0 ) today of a financial security that
promises to pay Vn in n periods. The fair value is given by
Vn
V0 = (Eq. 4, Lec. 1)
(1 + r)n
How do we value the security if the future payment Vn is random?

We may know the minimum, maximum or expected values of Vn .

But, we do not know the actual value of the future payment with certainty.

When the future payments are known with certainty, we use the above formula.
This formula comes from Lecture 1 where we assumed that there is no uncertainty
associated wit the future cash flows. The interest rate (r) used in no-risk or no-
uncertainty environment is referred to as risk-free interest rate. In order to explicitly
reveal this, we replace r with rf to indicate the interest rate is risk-free. Then, the
above formula, for certainty case, is given by:

Vn
V0 = (1)
(1 + rf )n
As r is e↵ective 1-period rate, so rf is the e↵ective 1-period risk-free rate.

1. Expected Return and Risk

In order to deal with the uncertain cash flows, we need to define the (random) return,
expected return and risk.1
Consider a financial security that is currently trading at P0 – this is the current
price and is known today. In one period, it can be sold at P̃1 . We use the symbol
‘tilde’ (⇠) on the top of P1 to indicate that P1 is random because the price in one
period is not known with certainty.
Then, we define the rate of return over one period as follows:

P̃1 P0
R̃ = (2)
P0
1
This course requires some background in Business Statistics. You need to know the concepts like
expected value, variance, standard deviation, covariance and correlation, etc. In Statistics Units, the
random variable is usually denoted by X which could mean anything, e.g., rain fall, temperature,
sales, heights etc. But, in Finance, it usually means return (future return). Note that historical
returns are known but the future return is not known or is random.

Prof. Keshab Shrestha, CFA 2

BFW1001 Foundations of Finance Lec. 3

P0 P̃1

Time = 0 1

Figure 1: Security prices at two di↵erent times.

Note that if we purchase one unit of the security at P0 today and sell it in one
period at P̃1 , our profit is P̃1 P0 and the random rate of return is given by equation
(2).2 Here, we are assuming that there are no other cash flows from holding the unit
of the security other than the one we receive when selling the security in one period.
For example, no dividends or coupons.3
Also, note that return is random because the price in one period is random.4 Since
the return is random, we can define its expected value and standard deviation. The
return standard deviation is known as volatility.
Consider a financial security that is currently trading at 10 per share. In one
period, it is expected to sell at 12, 10 or 9 per share with 30%, 40% and 30%
probabilities respectively. We can calculate the expected one period return (µ) and
volatility ( ) of this stock as follows:

End-of-period Probability Return ⇡i ⇥ Ri (Ri µ)2 ⇡i (Ri µ)2

Price ⇡i Ri
12 0.3 0.2000 0.0600 0.0289 0.00867
10 0.4 0.0000 0.0000 0.0009 0.00036
9 0.3 -0.1000 -0.0300 0.0169 0.00507
Column Sum 0.0300 0.01410

2
Note that if P̃1 P0 is negative, then it will be a loss and the return will be negative. The
return is defined as selling price (P̃1 ) minus the buying price (P0 ) the whole thing is divided by the
buying price (P0 ).
3
If there are dividends, we can include the dividends as part of the price at time t = 1.
4
Even though the past returns are known and can be computed from the past prices, the future
return is not known with certainty simply because we do not know the price of the security that will
prevail in one period. However, we use the past returns to characterize the probability distribution
of the return. This is similar to a situation where we do not know whether a head or tail will show
in a coin toss, but we know the sequence of heads and tails in the past tosses. We can use the past
outcomes of a coin toss to compute the probability of head.

Prof. Keshab Shrestha, CFA 3

 BFW1001 Foundations of Finance Lec. 3

The one period random return can take the following three possible values:

12 10 10 10 9 10
R1 = = 0.20, R2 = = 0.0 and R3 = = 0.10
10 10 10
The expected return (µ) can be computed as follows:

3
X
µ= ⇡i Ri = (0.3 ⇥ 0.20) + (0.4 ⇥ 0.00) + (0.3 ⇥ 0.10) = 0.0300 = 3.0%
i=1

The variance of the return can be computed as follows:

3
X
2
= ⇡i (Ri µ)2 = 0.3 ⇥ 0.0289 + 0.4 ⇥ 0.009 + 0.3 ⇥ 0.169 = 0.01410
i=1

Then, the volatility or return standard deviation is given by5

p p
volatility = = + 2 = + 0.01410 = 0.118743 = 11.8743%

In Finance, the volatility is used as one of the ways to measure the risk of
the security. The other measure of risk is the so-called value-at-risk (VaR)
which you may learn in BFW2401 Commercial banking and finance.

2. Behavior under Uncertainty

Since in the situation where the return is random, the most logical strategy is to work
with the expected return and return standard deviation or volatility (see Figure
2).
Assumption 1: We assume that individuals prefer higher expected return to lower
expected return, ceteris paribus (Latin phrase for ‘all other things being equal’).
What about the volatility? Based on the preference to risk, we can categorize
individuals into the following three categories:

1. Risk-averse: This type of individuals dislike risk - prefer lower volatility to

higher volatility. For this type of individuals to take higher risk, they need to
be compensated with a higher expected return.
5
Sometimes, I get the question “Is volatility the return standard deviation or return variance?”.
It is the return standard deviation. Another clue is that volatility is expressed in percentage term
whereas the unit of variance will be percentage squared.

Prof. Keshab Shrestha, CFA 4

 BFW1001 Foundations of Finance Lec. 3

2. Risk-neutral: This type of individuals do not care about the risk (volatility).
They neither like nor dislike risk or volatility. Volatility is irrelevant for such
individual. They only care about the expected return.

3. Risk-loving or Risk-seeking: This type of individuals like risk. They are

willing to pay for taking risk - they are willing to accept lower expected return
in order to take higher risk.

It is important to note that all three types of individuals prefer higher expected
return compared to lower expected return ceteris paribus (see Assumption 1).
Assumption 2: In Finance, we assume that individuals are, in general, risk-averse.

Expected Return (µ)

For risk-averse individual the following are true:

i. C is preferred to A
C
E ii. D is preferred to B
µA iii. Do not know if A is preferred to E
A
D iv. Preferred direction is North-West
B

A Volatility ( )

Figure 2: Risk-Return Space

Take a look at Figure 2. By investing both ‘A’ and ‘B’, we may be able to reach
point ’D’ which is associated with lower risk (volatility) compared either ‘A’ or ‘B’.
Such process of reducing risk by investing on more than one security is known as
diversification.

3. Valuation under Uncertainty

Now, let us consider the valuation of a financial security that promises to pay a
random amount Ṽn in n periods. Since, we do not know the actual amount Ṽn with
certainty, we cannot use the equation (1):

Prof. Keshab Shrestha, CFA 5

BFW1001 Foundations of Finance Lec. 3

Ṽn
V0 6=
(1 + rf )n
In order to value uncertain cash flows, we modify the above equation in two ways:

1. Replace the random value Ṽn with its expected value, i.e., E[Ṽn ].

2. Replace the e↵ective 1-period risk-free interest rate (rf ) with what is known as
the e↵ective 1-period cost-of-capital (k) that reflects the risk associated with
the future cash flow.

Then, the resulting valuation equation, which provides fair value of the security, is
given by:

h i
E Ṽn
V0 = (3)
(1 + k)n

It is important to note the following:

In the certainty case (Lecture 1), we showed that if the fair value is di↵erent
from the actual market price, there are ways to make free-money without any
uncertainty or risk.

But, under uncertain situation (e.g., situation with random future payments),
when the fair value is di↵erent from market price, This is known as disequi-
librium condition.

Under uncertainty, if there is a disequilibrium in the market (i.e., if the fair

value is di↵erent from the actual market price), there is no guarantee that we
can follow a strategy to make guaranteed free-money.

– This explains asset bubbles, where the di↵erence between the fair value
and market value is large.
– If you come up with a strategy that will make money when the gap between
these two prices closes or shrinks, there is no guarantee that such strategy
will work.
– The gap may even increase, rather than shrinking, indicating a larger mis-
pricing.
– Therefore, it is hard to tell when the bubble will burst.

The valuation under uncertainty is based on equilibrium concept.

Prof. Keshab Shrestha, CFA 6

 BFW1001 Foundations of Finance Lec. 3

Similarly, the financial security with multiple uncertain future cash flows can be
valued as follows:
h i h i h i
E C̃1 E C̃2 E C̃n
V0 = + + · · · + (4)
(1 + k) (1 + k)2 (1 + k)n

where C̃t is the random cash flows due in t periods and in the formula we replace it
by its expected value.

4. Cost-of-Capital for Common Stock

One of the ways to value common stock is to discount its future cash flows, which
are dividends. As mentioned in previous section, the cost of capital should reflect
the risk associated with the dividends. Specially for stock, the e↵ective 1-period
cost-of-capital (k) is given by the following formula:6

k = rf + (µm rf ) (5)

where
⇣ ⌘
Cov R̃m , R̃ h i
= ⇣ ⌘ and µm = E R̃m
V ar R̃m
where

rf is the e↵ective 1-period risk-free interest rate.

R̃m is the 1-period random return on the stock market (i.e., market portfolio),
which includes all the stocks traded in the market.

R̃ is the 1-period random return on the common stock.

Cov(R̃m , R̃) is the covariance between the return on the stock (R̃) and the
return on the market portfolio (R̃m ).

Finally, V ar(R̃m ) is the variance of the 1-period return on the market portfolio.

Above equation is known as Capital Asset Pricing Model (CAPM) and

the beta ( ) is known as CAPM beta.
6
There is a solid theoretical foundation for the formula. However, the discussion of the theory is
beyond the scope of this Unit.

Prof. Keshab Shrestha, CFA 7

 BFW1001 Foundations of Finance Lec. 3

When we are dealing with more than one common stock, we use stock-i to refer to
the ith stock. In this case, the e↵ective 1-period cost-of-capital (ki ) for the ith stock
is given by

ki = rf + i (µm rf ) (6)

where ⇣ ⌘
Cov R̃m , R̃i
m,i
i = ⇣ ⌘ = 2
V ar R̃m m

where R̃i is the 1-period random return on the ith stock, m,i is the covariance between
the return on the market portfolio (R̃m ) and the return on the ith stock (R̃i ) and, as
2
before, m is the variance of the return on the market portfolio.
Note that the volatility of the return on the market portfolio is return standard
2
deviation of market portfolio ( m ), which is equal to the square-root of m .
The di↵erence between the expected return on the market (µm = E[R̃m ]) and the
risk-free rate (rf ), i.e., (µm rf ), is known as the risk-premium on the market.
Note the following:

If there is no risk then beta will be zero in which case ki = rf and E[Ṽn ] = Vn .
Then, we end up with equation (1) as a special case.

Since the risk-free rate and return on the market portfolio does not change when
we go from one stock to another, di↵erent stocks have di↵erent cost-of-capital
simply due to di↵erence in betas.

– That means, if the two stocks have the same beta, the cost-of-capital are
the same for the two stocks.
– This does not mean the two stocks are the same.

4.1. Estimation of CAPM Beta

The question is “where do we get the beta?”. The covariance and variance are defined
in terms of probabilities. In Section 1, we only considered three possible returns. But,
in real world there infinite possible returns. Furthermore, the probabilities are now
know.

The CAPM beta can be computed only if we know Cov(R̃m , R̃i ) and V ar(R̃m )
– these are theoretical quantities.

But, in actual practice, we do not know these quantities.

Prof. Keshab Shrestha, CFA 8

BFW1001 Foundations of Finance Lec. 3

Therefore, we estimate Cov(R̃m , R̃i ) with the sample covariance between the
historical returns on the stock market and the stock.

Similarly, V ar(R̃m ) is estimated by the sample variance of the historical returns

on the stock market.

We can compute the covariance between the returns on the stock and returns on the
market using the Excel function
“=COVARIANCE.S(range1, range2)”,
where ’range1’ consists of historical return on the market and ’range2’ consists of
historical return on stock. Similarly, the variance of return on the market portfolio
can be computed using the Excel function
“=VAR.S(range1)”.

When we refer to return on the stock market portfolio, this portfolio is supposed
to consist of all the stocks traded in the stock market.

However, sometimes we use stock market index that does not necessarily consists
of all the stocks traded.

For example, for Malaysian market, we use FTSE Bursa malaysia KLCI
Index to represent stock market. This index consists of the largest 30 companies
by market capitalization on Bursa Malaysia’s Main Board.

In the United States, sometimes we use S&P500 Index that consists of 500 lead-
ing U.S. companies and captures approximately 80% of the coverage available
by market value.

We will use Excel in the class to estimate the CAPM beta for Maybank Sdn.
Berhad.

Prof. Keshab Shrestha, CFA 9