Calculating efficient portfolios

Lecturer: Dr Panagiotis Koutroumpis

School of Economics and Finance
 Calculating the efficient frontier - an example

v We consider a world with four risky assets having the following expected returns and
variance-covariance matrix:
 Calculating the efficient frontier - an example

v Each cell of the column vector labelled 𝑬(𝒓) minus constant contains the mean return
of the given asset minus the value of the constant 𝑐 (in this case 𝑐 = 4%). We use this
column to find the second envelope portfolio below.

v We separate our calculations into two parts: In the next subsection we calculate two
portfolios on the envelope of the feasible set. In the subsequent subsection we calculate
the efficient frontier.
 Calculating two envelope portfolios

v By Proposition 2, we have to find two efficient portfolios to identify the whole efficient
frontier. By Proposition 1 each envelope portfolio solves the system 𝑆𝑧 = 𝐸(𝑟) − 𝑐 for 𝑧.

v To identify two efficient portfolios, we use two different values for 𝑐. For each value of 𝑐,
23
we solve for 𝑧 and then set 𝑥1 = ∑ to find an efficient portfolio.
5 25

v We solve this system for 𝑐 = 0 and 𝑐 = 0.04, separately and obtain the two portfolios.
 Computing two efficient portfolios
A B C D E F G H
1 CALCULATING THE EFFICIENT FRONTIER Here we compute the efficient
Expected
Variance-covariance matrix
Expected minus portfolio in 2 steps.
returns constant
2 E(r) E(r) - c
3 0.40 0.03 0.02 0.00 0.06 0.02 <-- =F3-$B$8
4
5
0.03
0.02
0.20
0.00
0.00
0.30
-0.06
0.03
0.05
0.07
0.01
0.03
<--
<--
=F4-$B$8
=F5-$B$8
In step 1 we compute a portfolio
6
7
0.00 -0.06 0.03 0.10 0.08 0.04 <-- =F6-$B$8 𝑧:

z = S {E ( r ) - c}
8 Constant 0.04
9 -1
10 Computing an envelope portfolio with constant = 0
Envelope
z portfolio
11 x
12 0.1019 <-- {=MMULT(MINVERSE(A3:D6),F3:F6)} 0.0540 <-- =A12/SUM($A$12:$A$15)
13 0.5657 0.2998
14
15
0.1141
1.1052
0.0605
0.5857 In step 2, we normalize 𝑧 to sum
16
17
Sum 1.0000 <-- =SUM(F12:F15)
to 1 , computing two efficient
portfolios, 𝑥 and 𝑦.
18 Computing an envelope portfolio with constant = 0.04
Envelope
z portfolio
19 y
20 0.0330 <-- {=MMULT(MINVERSE(A3:D6),G3:G6)} 0.0423 <-- =A20/SUM($A$20:$A$23)
21 0.1959 0.2514
22 0.0468 0.0601
23 0.5035 0.6462
24 Sum 1.0000 <-- =SUM(F20:F23)
 Calculating two envelope portfolios

v To complete the basic calculations, we compute the means, standard deviations, and
covariance of returns for portfolios 𝑥 and 𝑦:
 Some excel formulas (see page 230 of the textbook)
v The transpose vectors of 𝑥 and of 𝑦 are inserted using the array function 𝑻𝒓𝒂𝒏𝒔𝒑𝒐𝒔𝒆.
This now enables us to calculate the mean, variance, and covariance as follows:

v 𝑬(𝒙) uses the array formula 𝑴𝑴𝒖𝒍𝒕(𝒕𝒓𝒂𝒏𝒔𝒑𝒐𝒔𝒆_𝒙, 𝒎𝒆𝒂𝒏𝒔). Note that we could have
also used the function 𝑺𝒖𝒎𝑷𝒓𝒐𝒅𝒖𝒄𝒕(𝒙, 𝒎𝒆𝒂𝒏𝒔).

v 𝑽𝒂𝒓(𝒙) uses the array formula 𝑴𝑴𝒖𝒍𝒕(𝑴𝑴𝒖𝒍𝒕(𝒕𝒓𝒂𝒏𝒔𝒑𝒐𝒔𝒆_𝒙, 𝒗𝒂𝒓_𝒄𝒐𝒗), 𝒙).

v 𝑺𝒊𝒈𝒎𝒂(𝒙) uses the formula 𝑺𝒒𝒓𝒕(𝒗𝒂𝒓_𝒙).

v 𝑪𝒐𝒗(𝒙, 𝒚) uses the array formula 𝑴𝑴𝒖𝒍𝒕(𝑴𝑴𝒖𝒍𝒕(𝒕𝒓𝒂𝒏𝒔𝒑𝒐𝒔𝒆_𝒙, 𝒗𝒂𝒓_𝒄𝒐𝒗), 𝒚).

v 𝑪𝒐𝒓𝒓(𝒙, 𝒚) uses the formula 𝒄𝒐𝒗(𝒙, 𝒚)/(𝒔𝒊𝒈𝒎𝒂_𝒙 ∗ 𝒔𝒊𝒈𝒎𝒂_𝒚).