Transcript Risk
Lecture 6: CAPM & APT
• The following topics are covered:
–
–
–
–
CAPM
CAPM extensions
Critiques
APT
L6: CAPM & APT
1
CAPM: Assumptions
• Investors are risk-averse individuals who maximize the
expected utility of their wealth
• Investors are price takers and they have homogeneous
expectations about asset returns that have a joint normal
distribution (thus market portfolio is efficient – page 148)
• There exists a risk-free asset such that investors may borrow or
lend unlimited amount at a risk-free rate.
• The quantities of assets are fixed. Also all assets are
marketable and perfectly divisible.
• Asset markets are frictionless. Information is costless and
simultaneously available to all investors.
• There are no market imperfections such as taxes, regulations,
or restriction on short selling.
L6: CAPM & APT
2
Derivation of CAPM
• If market portfolio exists, the prices of all assets must adjust until all are
held by investors. There is no excess demand.
• The equilibrium proportion of each asset in the market portfolio is
–
wi
m arket value of the individual asset
m arket value of all assets
(6.1)
• A portfolio consists of a% invested in risky asset I and (1-a)% in the
market portfolio will have the following mean and standard deviation:
–
–
~
~
~
E ( R p ) aE( Ri ) (1 a) E ( Rm )
~
( R p ) [a 2 i2 (1 a) 2 m2 2a(1 a) im ]1 / 2
(6.2)
(6.3)
• A portfolio consists of a% invested in risky asset I and (1-a)% in the
market portfolio will have the following mean and standard deviation:
• Find expected value and standard deviation of R p with respect to the
percentage of the portfolio as follows.
~
E ( R p )
a
~
~
E ( Ri ) E ( Rm )
L6: CAPM & APT
3
Derivation of CAPM
~
( R p )
a
1 2 2
[a i (1 a) 2 m2 2a(1 a) im ]1/ 2 [2a i2 2 m2 2a m2 2 im 4a im ]
2
• Evaluating
the two equations where a=0:
~
E ( Rp )
a
~
( E p )
a 0
~
~
E ( Ri ) E ( Rm )
a
a 0
m2
1 2 1 / 2
( m ) (2 m2 2 im ) im
2
m
• The slope of the risk-return trade-off:
~
E ( R p ) / a
~
( R p ) / a
a 0
~
~
E ( Ri ) E ( Rm )
( im m2 ) / m
• Recall~ that the slope of the market line is:
E ( Rm ) R f
;
m
• Equating the above two slopes:
~
E ( Rm ) R f
~
~
E ( Ri ) E ( Rm )
m
( im m2 ) / m
~
~
E ( Ri ) R f [ E ( Rm ) R f ] im2
m
L6: CAPM & APT
4
Extensions of CAPM
1.
2.
No riskless assets
Forming a portfolio with a% in the market portfolio and (1-a)% in the
minimum-variance zero-beta portfolio.
The mean and standard deviation of the portfolio are:
3.
–
–
4.
E(Rp ) aE(Rm ) (1 a)E(Rz )
~
( R p ) [a 2 m2 (1 a) 2 z2 2a(1 a)rzm z m ]1 / 2
The partial derivatives where a=1 are:
–
–
5.
E ( R p )
a
( R p )
a
;
E ( Rm ) E ( R z )
;
1
[a 2 m2 (1 a) 2 z2 ]1 / 2 [2a m2 2 z2 2a z2 ]
2
Taking the ratio of these partials and evaluating where a=1:
–
E( R p ) / a
( R p ) / a
E( Rm ) E ( Rz )
m
Further, this line must pass through the point E(R
is E ( Rz ) . The equation of the line must be:
6.
m
–
E ( R p ) E ( Rz ) [
E ( Rm ) E ( Rz )
m
), ( Rm )
and the intercept
] p
L6: CAPM & APT
5
Extensions of CAPM
• The existence of nonmarketable assets
– E.g., human capital; page 162
• The model in continuous time
– Inter-temporal CAPM
• The existence of heterogeneous expectations and taxes
L6: CAPM & APT
6
Empirical tests of CAPM
• Test form -- equation 6.36
– the intercept should not be significantly different
from zero
– There should be one factor explaining return
– The relationship should be linear in beta
– Coefficient on beta is risk premium
• Test results – page 167
• Summary of the literature.
L6: CAPM & APT
7
Roll (1977)’s Critiques
• Roll’s (1977) critiques (page 174)
• The efficacy of CAPM tests is conditional on the
efficiency of the market portfolio.
• As long as the test involves an efficient index, we are
fine.
• The index turns out to be ex post efficient, if every
asset is falling on the security market line.
L6: CAPM & APT
8
Arbitrage Pricing Theory
• Assuming that the rate of return on any security is a linear function of k
factors:
Ri E( Ri ) bi1F1 ... bik Fk i
Where Ri and E(Ri) are the random and expected rates on the ith asset
Bik = the sensitivity of the ith asset’s return to the kth factor
Fk=the mean zero kth factor common to the returns of all assets
εi=a random zero mean noise term for the ith asset
• We create arbitrage portfolios using the above assets.
•
• No wealth
n
w 0
i 1
i
-- arbitrage portfolio
• Having no risk and earning no return on average
L6: CAPM & APT
9
Deriving APT
• Return of the arbitrage portfolio:
R w R
n
p
i 1
i
i
wi E ( Ri ) wi bi1F1 ... wi bik Fk wi i
i
i
i
i
• To obtain a riskless arbitrage portfolio, one
needs to eliminate both diversifiable and
nondiversifiable risks. I.e.,
1
wi , n , wi bik 0 for
n
i
all factors
L6: CAPM & APT
10
Deriving APT
Rp wi E( Ri )
i
w E( R ) 0
i
i
i
As:
wb
i ik
0 for each k
i
How does E(Ri) look like? -- a linear combination
of the sensitivities
L6: CAPM & APT
11
APT
• There exists a set of k+1 coefficients, such that,
–
~
E(Ri ) 0 1bi1 ... k bik
(6.57)
• If there is a riskless asset with a riskless rate of
return Rf, then b0k =0 and Rf = 0
– E(Ri ) R f 1bi1 ... k bik
(6.58)
• In equilibrium, all assets must fall on the arbitrage
pricing line.
L6: CAPM & APT
12
APT vs. CAPM
• APT makes no assumption about empirical
distribution of asset returns
• No assumption of individual’s utility function
• More than 1 factor
• It is for any subset of securities
• No special role for the market portfolio in APT.
• Can be easily extended to a multiperiod framework.
L6: CAPM & APT
13
Example
• Page 182
• Empirical tests
– Gehr (1975)
– Reinganum (1981)
– Conner and Korajczyk (1993)
L6: CAPM & APT
14
FF 3-factor Model
• http://mba.tuck.dartmouth.edu/pages/faculty/ken.fren
ch/Data_Library/f-f_factors.html
L6: CAPM & APT
15