2.

Risk Management Irrelevance Propositions

FIN 500Q: Quantitative Risk Management

Basic Questions

• Can firm value be enhanced by using risk management to

reduce:
1. its idiosyncratic (diversifiable) risk?
2. its systematic (undiversifiable) risk?

• ⇒ No, under benchmark conditions described in this

module: irrelevance propositions.

• ⇒ Yes, under realistic violations of irrelevance propositions

discussed in the next module.

1
Mean and Variance of a Portfolio Return

• Portfolio P with weights wi in asset i

• RP : Return on portfolio P

• Ri : Return on asset i

• Mean and variance of portfolio returns:

N
X
E[RP ] = wi · E[Ri ],
i=1
XN N X
X
var[RP ] = wi2 · var[Ri ] + wi · wj · cov[Ri , Rj ].
i=1 i=1 j̸=i

2
Diversification

• Diversification is a very important tool for risk management.

• The power of diversification derives from the law of large

numbers: good luck in some asset returns offsets bad luck
in others.

• Diversification allows us to reduce the variance of overall

portfolio returns while keeping expected returns intact.

3
Single Asset Case

• Single asset example: investment of $500, which provides a

return of +100% (doubles) or −50% (halves), each with
probability 0.5. The mean return is 25%.

• The variance of the investor’s net return is

0.5 · (1 − 0.25)2 + 0.5 · (−0.5 − 0.25)2 = 0.5625.

• The variance of the investor’s dollar position is simply

5002 · 0.5625 = 140 625.

4
Two Uncorrelated Assets

• Now consider splitting the $500 investment into a portfolio

of two uncorrelated investments that have the same
distribution of returns.

• Choose portfolio weights w1 = w2 = 0.5.

• The mean return is still 25 percent.

• Now, there are four equally likely outcomes:

Description Prob. Dollar Outcome Return Outcome (%)
Double-Double 0.25 1000 100
Double-Half 0.25 625 25
Half-Double 0.25 625 25
Half-Half 0.25 250 -50

5
Two Uncorrelated Assets

• The variance of the portfolio return is

0.25 · (1 − .25)2 + 0.25 · (0.25 − 0.25)2 + 0.25 · (0.25 − 0.25)2
+ 0.25 · (−0.50 − 0.25)2 = 0.5 · 0.752 = 0.28125,
and the variance of the dollar position is
5002 · 0.28125 = 70 312.5.

• So by diversifying into two assets, the variance halved. This

is a special case because the correlation between the two
assets is zero.

• Alternatively, applying formula with w1 = w2 = 0.5,

σ12 = σ22 = 0.5625, and σ12 = ρ12 · σ1 · σ2 = 0:
var[RP ] = 0.52 · 0.5625 + 0.52 · 0.5625 = 0.28125.
6
N Identically Distributed Assets with Correlation ρ

• Consider a portfolio with weight wi = 1/N in each of N

assets, where each asset has variance σ 2 and correlation ρ
with every other asset.

• Then:
1 2 N −1 2 1
σP2 = σ + ρσ = (σ 2 − ρσ 2 ) + ρσ 2 .
N N N
• The first term represents diversifiable or idiosyncratic risk in
each asset. As N becomes large, this term tends to zero.

• The second term represents undiversifiable or systematic

risk.

7
Intuition for CAPM

• When ρ = 0, the portfolio variance tends to zero.

• When ρ > 0, not all the risk can be diversified away.

0.2

0.18

0.16

0.14

0.12

0.1

0.08

0.06

0.04

0.02
0 5 10 15 20 25 30

8
Intuition for CAPM

• Capital markets will price the risk that cannot be diversified.

• Assets having positive correlation with the market portfolio

must offer a higher expected return than the riskless rate.

• This is the basic intuition of the CAPM.

9
Logic of CAPM

• CAPM: assume that investors only care about the mean and
variance of their portfolio returns.

• Investors’ wealth is held in appropriate amounts of the

market portfolio, which offers a random return RM , and the
riskless asset, which provides a non-random return RF .

• From the logic of diversification:

1. A security’s expected return depends only on its return
covariance with the market portfolio.
2. A security that has a zero covariance with the market
portfolio must offer an expected return equal to the return
of the risk-free security.

10
CAPM Derivation

• Consider the following portfolio:

• Buy $1 worth of firm i share
• Short market portfolio by amount $wi
• Long $wi in riskless asset

• The covariance of this portfolio with the market is:

cov[Ri − wi RM + wi RF , RM ] = cov[Ri , RM ] − wi var[RM ].

• Setting it to zero implies that by choosing the following

short position in the market:
cov[Ri , RM ]
wi∗ = ,
var[RM ]
we have formed a portfolio with no systematic risk.

11
CAPM Derivation (Continued)

• Result 2 implies that the portfolio’s expected return must

equal the riskless rate.

• Therefore, E[Ri − wi∗ RM + wi∗ RF ] = RF , implying the

CAPM:
cov[Ri , RM ]
E[Ri ] = RF + (E[RM ] − RF ).
var[RM ]
| {z }
wi∗

• wi∗ is known as the beta of the asset (βi ).

• Note 1: To completely hedge the systematic risk of an asset, investors
must short βi units of the market portfolio.
• Note 2: In order for investors to hold the asset, they must be rewarded.
The expected return on the asset must equal the riskless rate plus βi
times the risk premium on the market portfolio.
12
Firm Valuation Using CAPM

• Consider firm i that operates for two periods.

• Firm has no debt and is 100% equity financed.

• Random cash flow C is paid out to shareholders in the

second period. No payouts in the first period.

• For valuation of the firm we use the CAPM:

E[Ri ] = RF + βi · (E[RM ] − RF ), (1)

where Ri = V .
C−V

13
Firm Valuation Using CAPM

• Expected return on assets:

 
C−V
E = RF + βi · (E[RM ] − RF ).
V

• Or conversely,

E[C]
V = . (2)
1 + RF + βi · (E[RM ] − RF )

• The value of the firm is the expected cash flow discounted

at the appropriate discount factor by the CAPM.

14
Firm Valuation Using CAPM

• Example: consider a company with a random cash flow a

year from today.
• Mean cash flow is $250.
• RF = 0.05.
• E[RM ] = 0.12.
• βi of the cash flow return (C/V − 1) from historical (or
comparable firm) data is 1.1.

250
⇒V = 1+0.05+1.1·(0.12−0.05) = 221.828.

15
RM Irrelevance Proposition 1

• Question 1: would a firm benefit by reducing its

idiosyncratic (diversifiable) risk?

• Answer (RM Irrelevance Proposition 1): no. The shareholders

of the firm are already diversified. Any cost of hedging will
simply reduce the expected value of the firm’s cash flow,
without changing its discount factor in (2). Therefore, there
is no benefit to shareholders.

• Shareholders do not need the firm to diversify unsystematic

risk. They can hold shares in a number of companies that
have not diversified, and can diversify the unsystematic risk
costlessly. We will refer to this practice by shareholders as
homemade risk management.
16
RM Irrelevance Proposition 2

• Question 2: would a firm benefit by reducing the

systematic risk of its cash flow?

• Answer (RM Irrelevance Proposition 2): no. The cost to the

firm of reducing its systematic risk would be at the market
price of systematic risk, which could be done by
shareholders.

• Let’s see the logic behind this answer.

• First, if the firm does not hedge, its value V is given by (2).

17
Finding Hedged Firm’s Value

• Second, suppose the firm hedges away its systematic risk.

• How does it achieve this? It sells short $βi V of the market

and invests the proceeds in the riskless asset.

• This gives the firm the same portfolio as in the CAPM

derivation, with wi∗ = βi .

• The covariance of the hedged firm with the market

portfolio is zero.

18
Hedged Firm’s Value

• By Result 2, the expected return of the hedged firm is RF .

Its value is expected to grow from V to V · (1 + RF ).

• Moreover, because the hedged firm has no systematic risk,

investors discount its return at RF .

• Therefore, the value of the hedged firm is:

V1 = V · (1 + RF )/(1 + RF ) = V .

• The value of the hedged firm equals the value of the

unhedged firm!

19
Why Didn’t the Firm’s Value Change?

• Why didn’t the firm’s value change?

• Firm sells its systematic risk in the market, and earns a

negative risk premium for the sale.

• Having gotten rid of the systematic risk, the firm is now

discounted at the riskless rate.

• However, the lower discounted rate is exactly offset by the

loss in cash flow.

• By just selling risk in the market the firm is unable to affect

its market value.

20
Homemade RM Again

• Conclusion: firm’s shareholders would not vote to pursue a

costly risk-reducing strategy for the firm if the purpose is to
either reduce the firm’s nonsystematic or systematic risk.

• By hedging, in a sense, the firm is buying insurance in the

market for its systematic risk, and the price it pays for the
insurance is the market price.

• If it was unhedged, investors would be providing

themselves the insurance, charging the firm by valuing it at
a higher discount rate.

21
Gold Mining Firm

• So far, we hedged the firm value by shorting the market

index.

• Another technique is to short forward contracts in the

commodity a firm produces.

• Consider a gold mining firm that will produce a unit of gold

T periods from now.

• Current spot price of gold is S0 , while its price T periods

from now is random: ST .

• Gold price risk has a systematic component – its beta is β.

• Let the forward price of gold be F0 .

22
Gold Mining Firm’s Choices

• Firm’s possible choices:

1. Unhedged: wait for T periods and sell the gold in the spot
market at ST .
2. Hedged: immediately short a forward contract for F0 .
Settlement takes place at T . Value of firm at T will be
ST + (F0 − ST ) = F0 .

• Under which strategy is the firm more valuable?

23
Forward Price of Gold

• Assume continuous compounding.

• By no arbitrage:
F0 = S0 · eRF T . (3)

• Why is this the forward price?

• An investor can sell gold currently at the price S0 , invest the
proceeds at the riskless rate, and obtain S0 · eRF T at T .
• Alternatively, she can sell gold at the forward price F0 ,
which she receives at maturity.
• Each strategy has a riskless payoff. Therefore, the payoffs
must be the same, otherwise markets permit an arbitrage.

24
Expected Spot Price at Maturity

• The CAPM with continuous compounding implies

E[ST ] = S0 · e(RF +β(E[RM ]−RF ))T . (4)

• This is just saying that to hold gold, investors must be

compensated for the systematic risk of gold, and therefore,
the expected price of gold must increase at the CAPM
discount rate.

25
Gold Mining Firm Value Under Two Strategies

• Unhedged firm: expected payoff at maturity is E[ST ].

Discounted at CAPM-based rate, value of firm is
E[ST ] · e−(RF +β(E[RM ]−RF ))T = S0 from (4).

• Hedged firm: total payoff at maturity is

ST + (F0 − ST ) = F0 , a sure profit. Therefore, investors
discount the hedged firm’s profit at the riskless rate, and
current value is F0 · e−RF T = S0 from (3).

• In either case, the value of the firm is the same.

• Once again investors do not need firm to hedge. They can

hold the risky firm and short forward contracts themselves.

26
Expected Profit from Short Forward

• Expected profit at maturity from shorting forward at time 0

is F0 − E[ST ].

• (3) implies S0 = F0 · e−RF T . Substituting for S0 in (4) implies

E[ST ] = F0 · eβ(E[RM ]−RF )T . (5)

• Expected profit from the short forward is

F0 · (1 − eβ(E[RM ]−RF )T ).
• Notice that if β > 0, then the expected profit from the short
forward is negative – its an expected loss.
• Why? By shorting a forward contract, an investor is selling
systematic risk (measured by beta) to the market, and the
market charges for absorbing the risk.
• Conversely, longing a forward contract on a commodity with
β > 0 will deliver an expected profit.
27
Firing Policy

• Question: would you fire a trader who on average makes

losses on his trading?

• You shouldn’t necessarily.

• If trader is reducing the firm’s systematic risk, we should be

expecting a loss.

• Lowering the systematic risk lowers the firm’s discount

factor (cost of financing for the firm), and therefore will
leave the value of the firm unchanged.

• If losses are made for other reasons, then may want to

consider firing. . .

28