Scribd
Upload
60 views3 pages

Notes For Finance 604 & 612 Prepared by Jessica A. Wachter

This document provides an optional proof of the security market line. It begins by defining the security market line equation and beta. It then considers an economy with multiple individuals who maximize expected utility by investing in risky securities or the risk-free asset. By taking the first order condition and applying Stein's Lemma, it shows that the capital asset pricing model must hold. Specifically, it derives that the expected return on any security minus the risk-free rate is equal to its covariance with the market return divided by the market variance, in line with the security market line.

Uploaded by

Mohamed Al-Wakil
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
60 views3 pages

Notes For Finance 604 & 612 Prepared by Jessica A. Wachter

This document provides an optional proof of the security market line. It begins by defining the security market line equation and beta. It then considers an economy with multiple individuals who maximize expected utility by investing in risky securities or the risk-free asset. By taking the first order condition and applying Stein's Lemma, it shows that the capital asset pricing model must hold. Specifically, it derives that the expected return on any security minus the risk-free rate is equal to its covariance with the market return divided by the market variance, in line with the security market line.

Uploaded by

Mohamed Al-Wakil
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Optional Proof of the Security Market Line1

The security market line states that for any asset with return Ri ,

E[Rj ] = Rf + βj E[RM − Rf ]

where Rf is the return on the riskfree asset, RM is the return on the market port-
folio, and βi is the regression coefficient from a regression of Ri on RM . That is,
βj = Cov(Rj , RM )/Var(RM ).
Suppose that there are I individuals in the economy. Let W0i > 0 be the initial wealth
of individual i. Let Xij be the proportion of W0i invested in security j. The total wealth
in the economy is W0m = ΣIi=1 W0i . Let W̃i denote the value of individual i’s porfolio.
That is, if i chooses a portfolio with return Rp , W̃i = W0i (1 + Rp ). Each individual
maximizes expected utility E[ui (W̃ )]. We assume that the individual prefers more to
less and is risk averse. This implies that u0i > 0 and u00i < 0.
The individual solves
max ui (W̃ )

subject to !
X X
W̃ = W0i Xij Rj + (1 − Xij )Rf
j j

the first order condition for an optimum is

E[u0i (W̃i )(Rj − Rf )] = 0

for every j. By the definition of covariances, this is the same as saying:

E[u0i (W̃i )]E[Rj − Rf ] = −Cov(u0i (W̃i ), Rj ) (1)


1
Notes for Finance 604 & 612 prepared by Jessica A. Wachter.

1
We assume that returns are normally distributed. The following mathematical results
holds:

Result 1 Stein’s Lemma


Suppose x and y are joint normal. Then

Cov(g(x), y) = E(g 0 (x))Cov(x, y)

Using Stein’s Lemma and (1), we can prove the CAPM.


Proof:
Equation (1), together with Stein’s Lemma implies that

E[u0i (W̃i )]E[Rj − Rf ] = −E[u00i (W̃i )]Cov(W̃i , Rj )

Eu00 (W̃ )
Let Qi = − Eui0 (W̃ i) . We can think of Qi as a measure of investor’s risk aversion. Divide
i i

and sum up both sides over i:


 
I 1
Σi=1 E[Rj − Rf ] = ΣIi=1 Cov(W̃i , Rj ) = Cov(W̃M , Rj )
Qi

Note that

Cov(W̃M , Rj ) = W0m Cov(1 + RM , Rj ) = W0m Cov(RM , Rj )

⇒
 
I 1
Σi=1 E[Rj − Rf ] = W0m Cov(RM , Rj )
Qi
 
I 1
Σi=1 E[RM − Rf ] = W0m Var(RM )
Qi
⇒
Cov(RM , Rj )
E[Rj − Rf ] = E[RM − Rf ]
Var(RM )
2

2
The last term is the risk premium on the market. Note that we can also derive an
expression:
1
E[RM − Rf ] = P 1 W0m Var(RM )
Qi

the first term on the right hand side is the “harmonic mean” of risk aversion. The
premium goes up with risk aversion and variance.

You might also like