2.

Mean-variance portfolio theory

(2.1) Markowitzs mean-variance formulation
(2.2) Two-fund theorem
(2.3) Inclusion of the riskfree asset

2.1

Markowitz mean-variance formulation

Suppose there are N risky assets, whose rates of returns are given by the random
variables R1 , , RN , where
Rn =

Sn(1) Sn (0)
, n = 1, 2, , N.
Sn(0)

Let w = (w1 wN )T , wn denotes the proportion of wealth invested in asset n,

N
X
with
wn = 1. The rate of return of the portfolio is
n=1

RP =

N
X

wnRn .

n=1

Assumptions

1. There does not exist any asset that is a combination of other assets in the
portfolio, that is, non-existence of redundant security.

= (1 1 1) are linearly independent, otherwise

2. = (R1 R2 RN ) and
RP is a constant irrespective of any choice of portfolio weights.

The first two moments of RP are

P = E[RP ] =

N
X

E[wn Rn] =

n=1

N
X

wnn, where n = Rn,

n=1

and
2 = var(R ) =
P
P

N X
N
X

wiwj cov(Ri, Rj ) =

i=1 j=1

N X
N
X

wiij wj .

i=1 j=1

Let denote the covariance matrix so that

2 = w T w .
P

For example when n = 2, we have

(w1

w2 )

11 12
21 22

w1
w2

= w1212 + w1w2(12 + 21) + w2222.

Remark
2 . In mean-variance
1. The portfolio risk of return is quantified by P
analysis, only the first two moments are considered in the portfolio model. Investment theory prior to Markowitz considered
the maximization of P but without P .

2. The measure of risk by variance would place equal weight on

the upside deviations and downside deviations.
3. In the mean-variance model, it is assumed that i, i and ij are
all known.

Two-asset portfolio
Consider two risky assets with known means R1 and R2, variances
12 and 22, of the expected rates of returns R1 and R2, together
with the correlation coefficient .
Let 1 and be the weights of assets 1 and 2 in this two-asset
portfolio.
Portfolio mean: RP = (1 )R1 + R2, 0 1
2 = (1 )2 2 + 2(1 ) + 2 2 .
Portfolio variance: P
1 2
1
2

We represent the two assets in a mean-standard deviation diagram

(recall: standard deviation = variance)

As varies, (P , RP ) traces out a conic curve in the R plane.

With = 1, it is possible to have = 0 for some suitable choice of
weight. In general, putting two assets whose returns are negatively
correlated has the desirable effect of lowering the portfolio risk.
6

In particular, when = 1,
P (; = 1) =

(1 )212 + 2(1 )12 + 222

= (1 )1 + 2.

This is the straight line joining P1(1, R1) and P2(2, R2).
When = 1, we have
P (; = 1) =

[(1 )1 2]2 = |(1 )1 2|.

When is small (close to zero), the corresponding point is close to

P1(1, R1). The line AP1 corresponds to
P (; = 1) = (1 )1 2.
1
The point A (with zero ) corresponds to =
.
1 + 2
1
The quantity (1 )1 2 remains positive until =
.
1 + 2
1
When >
, the locus traces out the upper line AP2.
1 + 2
7

Suppose 1 < < 1, the minimum variance point on the curve that
represents various portfolio combinations is determined by
2
P
= 2(1 )12 + 222 + 2(1 2)12 = 0

set
giving
12 12
= 2
.
1 212 + 22

Mathematical formulation of Markowitzs mean-variance analysis

N X
N
1 X
minimize
wiwj ij
2 i=1 j=1

subject to

N
X

wiRi = P and

i=1

N
X

wi = 1. Given the target expected

i=1

rate of return of portfolio P , find the portfolio strategy that mini2.

mizes P
Solution
We form the Lagrangian
L=

N X
N
1 X

2 i=1 j=1

wiwj ij 1

N
X

i=1

wi 1 2

where 1 and 2 are Lagrangian multipliers.

N
X

i=1

wi R i P

10

We then differentiate L with respect to wi and the Lagrangian multipliers, and set the derivative to zero.

N
X
L
=
ij wj 1 2Ri = 0, i = 1, 2, , N.
wi
j=1

(1)

N
X
L
=
wi 1 = 0;
1
i=1
N
X
L
=
wiRi P = 0.
2
i=1

(2)
(3)

From Eq. (1), the portfolio weight admits solution of the form

w = 1(11 + 2)
where

1 = (1

1 1)T and = (R1

(4)

R2 RN )T .

11

To determine 1 and 2, we apply the two constraints

T

w = 11 1 + 21 1.
P = T 1w = 1T 11 + 2T 1.
1 =

(5)
(6)

Write a = 1 11, b = 1 1 and c = T 1, we have

1 = 1a + 2b and P = 1b + 2c.
aP b
c bP
Solving for 1 and 2 : 1 =
and 2 =
, where

2
= ac b .
Note that 1 and 2 have dependence on P , which is the target
mean prescribed in the variance minimization problem.

12

Assume 6= h1, and 1 exists. Since is positive definite, so

a > 0, c > 0. By virtue of the Cauchy-Schwarz inequality, > 0.
The minimum portfolio variance for a given value of P is given by
T

(111 + 21)
a2
2bP + c
P
= 1 + 2P =
.

The set of minimum variance portfolios is represented by a parabolic

2 plane. The parabolic curve is generated by
curve in the P
P
varying the value of the parameter P .
2
P

= w

w = w

13

Alternatively, when P is plotted against P , the set of minimum

variance portfolio is a hyperbolic curve.
What are the asymptotic values of
dP
dP

so that

dP
?
dP
lim

2
dP dP
=
2 d
dP
P

=
2P
2a

2b
P q

=
a2
P 2bP + c
aP b

dP
=
.
lim
dP
a

14

Summary
aP b
c bP
and 2 =
, and the optimal

weight w = 1(11 + 2).

Given P , we obtain 1 =

To find the global minimum variance portfolio, we set

2
dP
2aP 2b
=
=0
dP

2 = 1/a. Correspondingly, = 1/a and

so that P = b/a and P
1
2 = 0. The weight vector that gives the global minimum variance
is found to be

11
wg =
=
a

11

11

15

Another portfolio that corresponds to 1 = 0 is obtained when P

c
is taken to be . The value of the other Lagrangian multiplier is
b
given by
 
a cb b

1
.

b
The weight vector of this particular portfolio is
2 =

=
wd =
b

Also, d2 =

 2
 
c
a b 2b cb + c

c
= 2.
b

16

Feasible set
Given N risky assets, we form various portfolios from these N assets.
We plot the point (P , RP ) representing the portfolios in the R
diagram. The collection of these points constitutes the feasible set
or feasible region.

17

Consider a 3-asset portfolio, the various combinations of assets 2

and 3 sweep out a curve between them (the particular curve taken
depends on the correlation coefficient 12).
A combination of assets 2 and 3 (labelled 4) can be combined with
asset 1 to form a curve joining 1 and 4. As 4 moves between 2 and
3, the curve joining 1 and 4 traces out a solid region.

18

Properties of feasible regions

1. If there are at least 3 risky assets (not perfectly correlated
and with different means), then the feasible set is a solid twodimensional region.
2. The feasible region is convex to the left. That is, given any two
points in the region, the straight line connecting them does not
cross the left boundary of the feasible region. This is because the
minimum variance curve in the mean-variance plot is a parabolic
curve.

19

Minimum variance set and efficient funds

The left boundary of a feasible region is called the minimum variance
set. The most left point on the minimum variance set is called the
minimum variance point. The portfolios in the minimum variance
set are called frontier funds.
For a given level of risk, only those portfolios on the upper half
of the efficient frontier are desired by investors. They are called
efficient funds.
A portfolio w is said to be mean-variance efficient if there exists
2 2 , except itself. That is,
no portfolio w with P P and P
P
you cannot find a portfolio that has a higher return and lower risk
than those for an efficient portfolio.

20

2.2

Two-fund theorem

Two frontier funds (portfolios) can be established so that any frontier portfolio can be duplicated, in terms of mean and variance, as
a combination of these two. In other words, all investors seeking
frontier portfolios need only invest in combinations of these two
funds.
Remark
Any convex combination (that is, weights are non-negative) of efficient portfolios is an efficient portfolio. Let i 0 be the weight
h
i
b
i
i
of Fund i whose rate of return is Rf . Since E Rf for all i, we
a
have
n
n
h
i
X
X
b
b
i
iE Rf
i = .
a
i=1
i=1 a
21

Proof
1 ), 1 , 1 and w 2 = (w2 w2 )T , 2, 2 are two
Let w1 = (w11 wn
n
1 2
1
1 2
known solutions to the minimum variance formulation with expected
2 , respectively.
rates of return 1
and

P
P

n
X

j=1
n
X
i=1
n
X

ij wj 1 2Ri = 0,

i = 1, 2, , n

(1)

wi r i = P

(2)

wi = 1.

(3)

i=1

It suffices to show that w1 + (1 )w2 is a solution corresponds

2
to the expected rate of return 1
P + (1 )P .
22

1. w1 + (1 )w2 is a legitimate portfolio with weights that sum

to one.
2. Eq. (1) is satisfied by w1 + (1 )w2 since the system of
equations is linear.
3. Note that
n h
X

i
1
2
wi + (1 )wi Ri

i=1
n
X

wi1Ri + (1 )

i=1
2
= 1
P + (1 )P .

n
X

wi2Ri

i=1

23

Proposition
Any minimum variance portfolio with target mean P can be uniquely
decomposed into the sum of two portfolios

c bP
where A =
a.

wP = Awg + (1 A)wd

Proof
For a minimum-variance portfolio whose solution of the Lagrangian
multipliers are 1 and 2, the optimal weight is

wP = 111 + 21 = 1(awg ) + 2(bwd).

Observe that the sum of weights is
c P b
P a b
ac b2
1a + 2b = a
+b
=
= 1.

We set 1a = A and 2b = 1 A, where

1 =

c P b

and

ab
2 = P
.

24

Indeed, any two minimum-variance portfolios can be used to substitute for wg and wd. Suppose

wu = (1 u)wg + uwd
wv = (1 v)wg + v wd
we then solve for wg and wd in terms of wu and wv . Then

wP = 1awg + (1 1a)w d
1a + v 1
1 u 1a
=
wu +
wv ,
vu
vu
where sum of coefficients = 1.

25

Example
Mean, variances and covariances of the rates of return of 5 risky
assets are listed:
Security
1
2.30
2
0.93
3
0.62
4
0.74
5
0.23

covariance
0.93 0.62
0.74
1.40 0.22
0.56
0.22 1.80
0.78
0.56 0.78
3.40
0.26 0.27 0.56

0.23
0.26
0.27
0.56
2.60

Ri
15.1
12.5
14.7
9.02
17.68

26

Solution procedure to find the two funds in the minimum variance

set:
1. Set 1 = 1 and 2 = 0; solve the system of equations
5
X

j=1

ij vj1 = 1,

i = 1, 2, , 5.

The actual weights wi should be summed to one. This is done

by normalizing vk1s so that they sum to one
vi1
1
wi = P n
.
1
v
j=1 j

1
After normalization, this gives the solution to wg , where 1 =
a
and 2 = 0.

27

2. Set 1 = 0 and 2 = 1; solve the system of equations:

5
X

ij vj2 = Ri,

j=1

i = 1, 2, , 5.

Normalize vi2s to obtain wi2.

After normalization, this gives the solution to wd, where 1 = 0
1
and 2 = .
b
The above procedure avoids the computation of a =
T
and b = 1 1.

1T 11

28

security
1
2
3
4
5
mean
variance
standard deviation

v1
0.141
0.401
0.452
0.166
0.440

v2
3.652
3.583
7.284
0.874
7.706

w1
w2
0.088
0.158
0.251
0.155
0.282
0.314
0.104
0.038
0.275
0.334
14.413 15.202
0.625
0.659
0.791
0.812

* Note that w1 corresponds to the global minimum variance point.

29

We know that g = b/a; how about d?

1

c
T
T
d = w d =
= .
b
b

c b

Difference in expected returns = d g = =

> 0.
b a
ab
1

c
Also, difference in variances = d2 g2 = 2 = 2 > 0.
b
a
ab

30

What is the covariance of portfolio returns for any two minimum

variance portfolios?
Write
u = w T R and Rv = w T R
RP
u
v
P

where R
so that

= (R1 RN )T . Recall that

gd = cov

=
=

1
a

1 R,

1T 1 = 1
ab

11
1
wg =
and wd =
a
b

1
b

1
b

since

b = 1 1.

31

In general,
u
v
cov(RP
, RP
) = (1 u)(1 v)g2 + uvd2 + [u(1 v) + v(1 u)]gd
uvc
u + v 2uv
(1 u)(1 v)
=
+ 2 +
a
b
a
1
uv
=
+
.
2
a
ab
In particular,

cov(Rg , RP ) = wT
g w P =

11wP
a

1
= = var(Rg )
a

for any portfolio wP .

For any Portfolio u, we can find another Portfolio v such that these
two portfolios are uncorrelated. This can be done by setting
1
uv
+
= 0.
2
a
ab

32

The mean-variance criterion can be reconciled with the expected

utility approach in either of two ways: (1) using a quadratic utility
function, or (2) making the assumption that the random returns are
normal variables.
Quadratic utility
b 2
x ,
2
where a > 0 and b > 0. This utility function is really meaningful
only in the range x a/b, for it is in this range that the function is
increasing. Note also that for b > 0 the function is strictly concave
everywhere and thus exhibits risk aversion.
The quadratic utility function can be defined as U (x) = ax

33

mean-variance analysis

maximum expected utility criterion

based on quadratic utility

Suppose that a portfolio has a random wealth value of y. Using the

expected utility criterion, we evaluate the portfolio using


b 2
E[U (y)] = E ay y
2
b
= aE[y] E[y 2]
2
b
b
2
= aE[y] (E[y]) var(y).
2
2
Note that we choose the range of the quadratic utility function such
b
that aE[y] (E[y])2 is increasing in E[y]. Maximizing E[y] for a
2
given var(y) or minimizing var(y) for a given E[y] is equivalent to
maximizing E[U (y)].
34

Normal Returns
When all returns are normal random variables, the mean-variance
criterion is also equivalent to the expected utility approach for any
risk-averse utility function.
To deduce this, select a utility function U . Consider a random
wealth variable y that is a normal random variable with mean value
M and standard deviation . Since the probability distribution is
completely defined by M and , it follows that the expected utility
is a function of M and . If U is risk averse, then
E[U (y)] = f (M, ),

f
with
>0
M

and

f
< 0.

35

 Now suppose that the returns of all assets are normal random
variables. Then any linear combination of these assets is a normal random variable. Hence any portfolio problem is therefore
equivalent to the selection of combination of assets that maximizes the function f (M, ) with respect to all feasible combinations.
For a risky-averse utility, this again implies that the variance
should be minimized for any given value of the mean. In other
words, the solution must be mean-variance efficient.
Portfolio problem is to find w such that f (M, ) is maximized
with respect to all feasible combinations.

36

2.3

Inclusion of the riskfree asset

Consider a portfolio with weight for a risk free asset and 1 for
a risky asset. The mean of the portfolio is
RP = Rf + (1 )Rj

(note that Rf = Rf ).

The covariance f j between the risk free asset and any risky asset
is zero since
E[(Rj Rj ) (Rf Rf ) = 0.
|
{z
}
zero

2 is
Therefore, the variance of portfolio P

2 = 2 2 +(1 )2 2 + 2(1 )
P
fj
j
f
|{z}
|{z}
zero
zero

so that P = |1 |j .

37

The points representing (P , RP ) for varying values of lie on a

straight line joining (0, Rf ) and (j , Rj ).

If borrowing of risk free asset is allowed, then can be negative. In

this case, the line extends beyond the right side of (j , Rj ) (possibly
up to infinity).

38

Consider a portfolio with N risky assets originally, what is the impact

of the inclusion of a risk free asset on the feasible region?
Lending and borrowing of risk free asset is allowed
For each original portfolio formed using the N risky assets, the new
combinations with the inclusion of the risk free asset trace out the
infinite straight line originating from the risk free point and passing
through the point representing the original portfolio.
The totality of these lines forms an infinite triangular feasible region
bounded by the two tangent lines through the risk free point to the
original feasible region.

39

No shorting of the riskfree asset

The line originating from the risk free point cannot be extended
beyond points in the original feasible region (otherwise entail borrowing of the risk free asset). The new feasible region has straight
line front edges.

40

The new efficient set is the single straight line on the top of the
new triangular feasible region. This tangent line touches the original
feasible region at a point F , where F lies on the efficient frontier of
the original feasible set.

b
. This assumption is reasonable since the risk free
a
asset should earn a rate of return less than the expected rate of
return of the global minimum variance portfolio.
Here, Rf <

41

One fund theorem

Any efficient portfolio (any point on the upper tangent line) can be
expressed as a combination of the risk free asset and the portfolio
(or fund) represented by F .
There is a single fund F of risky assets such that any efficient
portfolio can be constructed as a combination of the fund F and
the risk free asset.
Under the assumptions that
every investor is a mean-variance optimizer
they all agree on the probabilistic structure of the assets
unique risk free asset
Then everyone will purchase a single fund, which is the market
portfolio.
42

Now, the proportion of wealth invested in the risk free asset is

1

N
X

wi .

i=1

Modified Lagrangian formulation

2
P
1
= wT w
2
2

minimize

wT + (1 wT 1)r = P .

subject to

1 T
Define the Lagrangian: L = w w + [P r ( r1)T w]
2
N
X
L
=
ij wj ( r1) = 0,
wi
j=1

L
=0

giving

i = 1, 2, , N

(1)

( r1)T w = P r.

(2)

43

Solving (1): w = 1( r1). Substituting into (2)

P r = ( r1)T 1( r1) = (c 2rb + r2a).
By eliminating , the relation between P and P is given by the
following pair of half lines
T

2 = w w = (w r w
P

1)

= (P r) = (P r)2/(c 2rb + r2a).

44

With the inclusion of the riskfree asset, the set of minimum variance
portfolios are represented by portfolios on the two half lines
Lup : P r = P

Llow : P r = P

ar2 2br + c

ar2 2br + c.

(3a)
(3b)

Recall that ar2 2br + c > 0 for all values of r since = ac b2 > 0.
The minimum variance portfolios without the riskfree asset lie on
the hyperbola
2 2b + c
a
P
2
P
.
P
=

45

b
When r < g = , the upper half line is a tangent to the hyperbola.
a
The tangency portfolio is the tangent point to the efficient frontier
(upper part of the hyperbolic curve) through the point (0, r).

46

The tangency portfolio M is represented by the point (P,M , M

P ),
and the solution to P,M and M
P are obtained by solving simultaneously
2 2b + c
a
P
2
P
P
=

P = r + P

c 2rb + r2a.

Once M
P is obtained, the corresponding values for M and w M are

M
r

1
P
M =
and
w
=

( r1).
M
M
2
c 2rb + r a
The tangency portfolio M is shown to be
1( r

wM =
b ar

1) ,

c br
M
=
P

b ar

and

c 2rb + r2a
2
P,M =
.
2
(b ar)

47

b
> r. Note that
When r < , it can be shown that M
P
a




b b ar
c br

=
b ar a
a
c br
b2
br
=
2+
a
a
a
ca b2

=
=
> 0,
2
2
a
a
b
b
M
so we deduce that P > > r, where g = .
a
a
b
M
P
a

b
r
a

On the other hand, we can deduce that (P,M , M

P ) does not lie on
b
the upper half line if r .
a

48

b
When r < , we have the following properties on the minimum
a
variance portfolios.
1. Efficient portfolios
Any portfolio on the upper half line
P = r + P

ar2 2br + c

within the segment F M joining the two points (0, r) and M

involves long holding of the market portfolio and riskfree asset,
while those outside F M involves short selling of the riskfree asset
and long holding of the market portfolio.
2. Any portfolio on the lower half line
P = r P

ar2 2br + c

involves short selling of the market portfolio and investing the

proceeds in the riskfree asset. This represents non-optimal investment strategy since the investor faces risk but gains no extra
expected return above r.
49

What happens when r = b/a? The half lines become

s

 

b2

P = r P c 2
b+
= r P
,
a
a
a
which correspond to the asymptotes of the feasible region with risky
assets only.
b
, efficient funds still lie on the upper half line,
a
M
though P does not exist. Recall that

Even when r =

w = 1( r1) so that

1T w = (11 r111) = (b ra).

T
When r = b/a, 1 w = 0 as is finite.
Any minimum variance portfolio involves investing everything in the
riskfree asset and holding a portfolio of risky assets whose weights
sum to zero.
50

b
When r > , the lower half line touches the feasible region with
a
risky assets only.

Any portfolio on the upper half line involves short selling of

the tangency portfolio and investing the proceeds in the riskfree
asset.
It makes sense to short sell the tangency portfolio since it has
an expected rate of return lower than the risk free asset.
51

Interpretation of the tangency portfolio (market portfolio)

One-fund theorem states that everyone will purchase a single
fund of risky assets and borrow or lend at the risk free rate.
If everyone purchases the same fund of risky assets, what must
that fund be? This fund must equal the market portfolio.
The market portfolio is the summation of all assets. If everyone
buys just one fund, and their purchases add up to the market,
then that one fund must be the market as well.
In the situation where everyone follows the mean-variance methodology with the same estimates of parameters, the efficient fund
of risky assets will be the market portfolio.

52

How does this happen? The answer is based on the equilibrium

argument.
If everyone else (or at least a large number of people) solves the
problem, we do not need to. The return on an asset depends
on both its initial price and its final price. The other investors
solve the mean-variance portfolio problem using their common
estimates, and they place orders in the market to acquire their
portfolios.
If orders placed do not match what is available, the prices must
change. The prices of assets under heavy demand will increase;
the prices of assets under light demand will decrease. These
price changes affect the estimates of asset returns directly, and
hence investors will recalculate their optimal portfolio.

53

 This process continues until demand exactly matches supply;

that is, it continues until there is equilibrium.
Summary
In the idealized world, where every investor is a mean-variance
investor and all have the same estimates, everyone buys the
same portfolio, and that must be equal to the market portfolio.
Prices adjust to drive the market to efficiency. Then after other
people have made the adjustments, we can be sure that the
efficient portfolio is the market portfolio, so we need not make
any calculations.

54