Portfolio Analysis Using Single Index Model

ANTON ABDULBASAH KAMIL

School of Mathematical Sciences
Universiti Sains Malaysia
11800 USM, Minden, Penang
MALAYSIA


Abstract: - This paper focused on portfolio analysis that set-up among 10 selected stocks from
Kuala Lumpur Stock Exchange (KLSE). Two types of analysis conducted in this paper, which
is daily analysis and weekly analysis using single index model. The result shows that entrance
of 5 stocks to set-up optimal portfolio for daily analysis and only 2 stocks been selected in the
weekly analysis. Among this 2 analysis, weekly analysis provides a higher profit level with
lower risk level if compared to daily analysis.

Key-Words: - Investment return, Investment risk, Diversification, Portfolio theory, Expected
return and risk.

1 Introduction
This paper will focus on stock investment
through setting up a portfolio with
calculated expected profit and risk level.
The main objective is to select an optimal
portfolio from daily and weekly analysis
that will provide an optimal return with
certain level of risk among 10 selected
stocks in KLSE.
Single index model provide an optimal
expected return and risk through formula (1)
and (2).
Expected return,
m p p p
R R + = (1)
Expected risk,

=
+ =
N
i
e i m p p
i
X
1
2 2 2 2 2
(2)
where:

=
=
N
i
i i p
X
1

=
=
N
i
i i p
X
1

Throughout this paper, a best
combination of stocks in portfolio will be
conducted with high profit and low risk
level. Besides, the proportion of capital to
invest in selected stock will be computed
with single index model.

2 Theory
2.1 Investment Return
The return of an investment is the money
earned from the difference of the
investment result as a profit of the
investment. The expected return of an
individual stock can be written as,
m i i i
R R + = (3)
where:
=
i
R The expected return to security i
=
i
The component of security is return
that is independent of the markets
performance a random variable.
=
i
A constant that measure the expected
change in
i
R given a change in
m
R .
=
m
R The rate of return on the market
index a random variable.

2.2 Investment Risk
According to Fisher and Jordan (1994) the
investment risk is a risk of holding
securities which is generally associated
with the possibility that the return will be
less than the return that were expected. Risk
is also define as a standard deviation
around the expected return. More
dispersion or variability about a security
expected return meant the security was
riskier than the one with less dispersion.
Securities carry differing degree of
expected risk leads most investor to the
notion of holding more than one security at
a time intended to reduce the risk. As the
posibility of deviation is higher, the risk is
also higher. The risk of a security which is
the variance of a securitys return can be
written as,
2 2 2 2
i
e m i i
+ = (4)
where:
=
2
i
Variance of a securitys return.
=
2
i
e
Variance of a stock movement that
is associated with the movement of
the market index.
=
2
m
Variance in market index.

2.3 Diversification
Diversification is a process of portfolio
combination involve a few different
investment instrument. Diversification will
reduce the expected investment risk. The
influence of diversification towards the
securitys risk as follow:
i. Diversification can reduce the
unsystematic risk. The unsystematic risk
is a risk cause by the characteristic of the
industry. The unsystematic risk is also
known as unique risk.
ii. Diversification cannot reduce the
unsystematic risk but it can flatten risk
dispersion. Systematic risk which is
known as market risk is due to the overall
changes in the market.
Diversification of ones holding is
intended to reduce risk in an economy in
which every assets return are subject to
some degree of uncertainty. Effort to spread
and minimize risk is a form of
diversification. A traditional form of
diversification have concentrated upon
holding a number of security types (stock,
bonds) across industry lines (utility, mining,
manufacturing groups). The best
diversification comes through holding large
number of securities scattered across
industries.

2.4 Portfolio Theory
In this text portfolio will consist of
collection of securities, or portfolio is a
group of investment opportunity.
An investor that form a portfolio hope
to gain as much return as possible at the
lowest risk compare to the investment on
only one investment opportunity. The
portfolio theory explain the correlation
between the expected return and the risk of
the portfolio. There a few model used to
analyse the portfolio such as the Markowitz
model, factor model, and single index
model.

2.5 Single Index Model
The single index model is also known as
the market model. In this moel the portfolio
risk depend on the sensitivity of the
security associated to the changes of the
portfolio market return. The portfolio
analysis converge on two parameters which
is the expected return and the portfolio risk.
Besides this the portfolio analysis
also calculate the correction or variance of
each pairs of possibility securities that form
portfolio. If the total of stocks collected in
the potfolio increase, the covariance that is
calculated will also increase. This model is
a model that analyse the movement of the
stocks cause by the market index.

2.5.1 Expected Return And Risk Of
An Individual Security
Most of the stocks prices tend to increase
when the market goes up and when the
market goes down, the stocks prices tend to
decrease. This suggested that one reason
the security expected return might be
correlated is due to the common response to
the market changes. A useful measurement
of this correlation can be obtained by
relating the return on a stock to the return
on a stock market index that can be writen
as:
m i i i
R a R + = (5)
where:
=
i
R The return of security i.
=
m
R The return on market index
=
i
a The component of security is return
that is independent of the markets
performance.
Equation breaks the return on a stock
into two component, that part of it is due to
the market and another part is independent
of the market.
i
measures how sensitive a
stocks return is to the return on the market.
The term
i
a represents component of
return insensitive to (independent) the
return on market. Let
i
denote the
expected value of
i
a and let
i
e represent
the random element of
i
a . Then,
i i i
e a + = (6)
where
i
e has an expected value of zero.
The return on a stock equation can be
written as,
i m i i i
e R R + + = (7)
where:
i
= Component of security is return that
is independent of the markets
performance.
i
e = Residual.
i
R is dependent variable while
m
R is
independent variable. Both
m
R and
i
e are
random variable. Each of them have a
probability distribution and a mean and
standard deviation which is
m
and
i
e
.
When these components were added
together, it will be equal to the total return.
It is convenent to have
i
e uncorrelated with
m
R which mean that,
0 )] )( 0 [( ) , ( = =
m m i m i
R R e E R e Cov .
(8)
If
i
e is uncorrelated with
m
R , it
implies that equation (7) describes the
return on any security is independent of the
return on the market changes. The
estimation of
i
,
i
and
ei
are often
obtain from time series regression analysis.
Regression analysis guarantees that
m
R
and
i
e will be uncorrelated, at least over
the period that the equation has been fit. So
i
e is independent of
j
e for all value of I
and j, or E( 0 ) =
j i
e e is the key assumption
in the single index model. This implies that
the only reason stocks vary together,
systematically is because of a common
comovement with the market.
Basic equation:
i m i i i
e R R + + = (for
all stocks i = 1, , N)
By construction: E(
i
e ) = 0 (for all stocks i
= 1, , N)
By assumption:
1. Index unrelated to unique return:
0 ) ( [ =
m m i
R R e E (for all stocks i =
1, , N)
2. Securities only related through
common response to market:
0 ) ( =
j i
e e E (for all pairs of stocks i =
1,, N and j = 1,,N but j i )
By definition:
1. Variance
i i i
e e E e
2 2
) ( = =
2. Variance
2
m m
R =
The expected return, standard deviation
and covariance in the single index model
are used to represent the jointmovement of
securities as the following result:
1. The mean return,
m i i i
R R + =
2. The variance of a securitys return,
2 2 2 2
ei m i i
+ =
3. The covariance of returns between
securities i and j,
2
m j i ij
=
The expected return are divided into
two component which is the independent
component,
i
(unique part) and a market
related part,
m i
R . Likewise, a securitys
variance has the same two part which is
unique risk,
2
i
e
and market related risk. In
contrast, the covariance depends only on
the market risk. The single model index
implied that the only reason securities move
together is a common response to the
market movements. The total parameter
that should be find is 3N + 2. For the
portfolio with 10 stocks, there are 32
parameter that should be find.

2.5.2 The Portfolio Expected Return
and Risk
Beta on a portfolio,
p
is define as a
weighted average of the individual
i
s on
each stock in the portfolio where the
weights are the fraction of the portfolio
invested in each stock. Then

=
=
N
i
i i p
X
1
(9)
The Alpha of a portfolio,
p
is define as

=
=
N
i
i i p
X
1
(10)
Therefore the expected return of a portfolio
is written as,
m p p p
R R + = (11)
Given the expected return of any portfolio
are,

=
=
N
i
i i p
R X R
1
(12)
Then by replace
i
R in the equation above

= =
+ =
N
i
N
i
m i i i i p
R X X R
1 1
(13)
Given the variance of a portfolio is written
as

= =

=
+ =
N
i
N
i
N
i j
j
ij j i i i p
X X X
1 1 1
2 2 2
(14)
Therefore, by replacing the result above for

= =
=
+
+ =
N
i
N
i j
j
N
i
e i m j i j i
N
i
m i i p
i
X X X
X
1 1 1
2 2 2
1
2 2 2 2



(15)
If the portfolio p is taken to be the
market portfolio, then the expected return
on p must be
m
R . From the equation (11)
the only value of
p
and
p
that
guarantee
m p
R R = for any choice of
m
R
is equal to zero and
p
equal to one. Thus,
the Beta on the market is one and stocks are
thought of as being more or less risky than
the market according to whether their Beta
is larger or smaller than 1. In double
summation , j i if i = j then the terms
would be,

=
=
N
i
N
i j
j
m i i m j i j i
X X X
1 1
2 2 2 2
. But
this are exactly the terms in the first
summation. Thus, the variance on the
portfolio can be written as,

= = =
+ =
N
i
N
j
N
i
e i m j i j i p
i
X X X
1 1 1
2 2 2 2
.
(16)
Thus, the risk of the investors portfolio
could be represented as,

=
+ =
N
i
e i m p p
i
X
1
2 2 2 2 2
(17)
Assume for a moment that an investor
forms a portfolio by placing equal amounts
of money into each of N stocks. The risk of
this portfolio can be written as,

+ =

=
N
i
e m p p
i
N N
1
2 2 2 2
1 1
(18)
Because N show the total of stocks in
the portfolio, then the larger the N, the
value in the second part of equation (18)
will be smaller. This part show the residual
risk or unsystematic risk, then the larger the
total of stocks in the portfolio the
contribution of the unsystematic risk will be
decrease. The other risk will vanished
although the total of stocks in the portfolio
is getting larger because the risk is
correlated to the Beta of the portfolio.
When the residual risk is closer to zero,
then the portfolio risk will be,
[ ] [ ]
i i m m p m p p
X = = =
2 / 1
2 2

(19)
Because the value of
m
is the same
for every stock, then the contribution of the
stock risk to the portfolio risk consist of
many stock depend on
i
. Total of the
individual risk are the same as equation (4).
The effect of
2
i
e
to the portfolio risk will
reduce when it is closer to Zero by
increasing the stock, then the
2
i
e
is known
as diversifiable risk or unsystematic risk
(risk that can be reduce by diversification).
The effect of
2
2 m
i
to the portfolio
risk will not change although the total
stocks has increase is known as systematic
risk or non diversifiable risk. Because
2
m

is a constant and unsystematic risk cannot
vanish by increasing the total stock, then
the stock Beta,
i
always been used as a
risk measurement for certain stock. The
value of
2
i
e
can be calculated by,
2 2 2 2
m i i e
i
= (20)

3 Data And Methodology
There are 2 types of data collected for the
process of analysis in this paper. First,
composite index and price of 10 selected
stocks was collected daily according to
trading day of KLSE. The duration of data
is from October 15
th
, 2002 until March 18
th
,
2003, that is 100 trading days in KLSE.
Selection of 10 stocks for this papers
analysis was according to the most volume
traded in KLSE from March 11
th
, 2003
until March 15
th
, 2003.
Secondly, interest rates was collected
that was release by Bank Negara Malaysia
at March 18
th
, 2003. The rates is 3%
annually.
Data analysis model;
i. Stocks profit:
t
t i
i
P
P P
R
1

= and

=
=
N
t
it i
R
N
R
1
1

Markets profit:
1
1

=
t
t t
m
IK
IK IK
R
and

=
=
N
t
mt m
R
N
R
1
1

ii. Alpha and beta value of each stock:
[ ]

=
=

=
N
t
m m
N
t
m m i i
i
R R
R R R R
1
2
1
) (
) )( (
and
m i i i
R R =
iii. Unsystematic risk (variance):

=
N
t
m mt m
R R
N
1
2 2
) (
1
1

Stocks profitable risk (variance):

=
N
t
i it i
R R
N
1
2 2
) (
1
1

Residual risk (variance):
2 2
1
2 2
m i e
i
=
iv. Excess return to beta ratio (ERB):

i
f i
R R
ERB

=
v. Cut-off rate ) (
i
C :

=
2
2
2
2
2
1
) (
i
i
e
i
m
e
i f i
m
i
R R
C

vi. Stock selection into portfolio is
according to value of ERB and . *
i
C If
the value of ERB > *
i
C , then the
stock selected into portfolio or vice
versa.
vii. Capital proportion ( )
i
X for each
selected stock in portfolio:

=
=
N
i
i
i
i
Z
Z
X
1
where

= *
2
i
i
f i
e
i
i
C
R R
Z
i

viii. Coefficient variation:
p
p
R
CV

=

4 Result
4.1 Daily Analysis

Table 1. Market and Stock Return, Alpha, Beta and Variance (Daily Analysis)

Stock
) 10 (
3
x
R
i

) 10 (
5
x
R
m

) 10 (
..
3
x
I Var

) 10 (
.
5
x
m Var

) 10 (
.
5
x
R R Kov
m i

Beta I Alpha I
) 10 (
3
x

) 10 (
.
3
x
e Var
i

IOI 0.234 8.3 0.467 7.19 2.712 0.377 0.203 0.457
MAYBANK 0.020 8.3 0.296 7.19 2.719 0.378 -0.011 0.286
AMMB -1.620 8.3 0.515 7.19 7.023 0.976 -1.701 0.446
BAT 0.510 8.3 0.055 7.19 0.783 0.109 0.501 0.054
SIME 0.465 8.3 0.187 7.19 3.762 0.523 0.422 0.167
COMMERZ -0.200 8.3 0.341 7.19 6.238 0.867 -0.272 0.287
PBB -0.420 8.3 0.147 7.19 2.481 0.345 -0.449 0.138
AFFIN -4.290 8.3 0.415 7.19 1.809 0.251 -4.311 0.410
KEMAS 0.738 8.3 1.514 7.19 11.100 1.541 0.610 1.343
GENTING 1.382 8.3 0.294 7.19 7.032 0.977 1.301 0.225

From table 1, there are 6 stocks provide
positive return with highest value recorded
by Genting which is 0.001382. On the other
hand, the lowest return recorder by Affin
with -0.00429. The level of markets return
is 0.000083 and market risk level is
0.0000719. For unsystematic risk level,
Kemas recorded the highest value of
0.0001343. Here it shows that profit level
and risk level is positively correlated but
with just a small correlation coefficient
value.

Table 2. ERB,
i
C , and Capital Investment Proportion ) (
i
X

Stock ERB
) 10 (
4
x C
i

ERB>< *
i
C
i
Z
i
X
IOI 0.000402 0.0882* >* 0.3216 0.0194
MAYBANK -0.000160 0.0292 <
AMMB -0.001740 -2.184 <
BAT 0.003928 -1.652 >* 7.8832 0.4752
SIME 0.000732 -0.868 >* 2.2361 0.1348
COMMERZ -0.000330 -1.161 <
PBB -0.001460 -1.748 <
AFFIN -0.017380 -2.868 <
KEMAS 0.000425 -2.344 >* 0.4731 0.0285
GENTING 0.001329 -0.0038 >* 5.6735 0.3421

In table 2, the optimal value of
i
C is
0.0882 and this value will be use as an
indicator for stock selection into optimal
portfolio. Here, there are 5 counters show
their ERB value is greater than *
i
C . These
counter are BAT, Genting, SIME, Kemas
and IOI. The largest capital investment
proportion will go to BAT with 47.52% and
the smallest proportion is 1.94% by counter
IOI. Lastly, table 3 give the value of
expected return from this optimal portfolio
with daily analysis is 0.08% and at a very
low risk level, that is 0.000062. This shows
that investment make through this portfolio
will provide a higher return, that is 10%
higher than markets return (0.008%).

Table 3. Return and Risk Level for Optimal Portfolio (Daily Analysis)


000083 . 0 =
m
R Var.m = 0.0000727
Stock
i
R
i
X
Alpha
) (
i
X
Beta
) (
i
X
i i
X R
Var. e
i
i i
e Var X . .
BAT 0.00051 0.4752 0.000238 0.05169 0.000242 0.000054 0.000026
GENTING 0.00138 0.3421 0.000445 0.33429 0.000472 0.000225 0.000077
SIME 0.00047 0.1348 0.000057 0.07046 0.000063 0.000167 0.000023
KEMAS 0.00074 0.0285 0.000017 0.04391 0.000021 0.001342 0.000038
IOI 0.00023 0.0194 0.000004 0.00731 0.000004 0.000457 0.000009
Jumlah 0.000761 0.50766 0.000802 0.002245 0.000173
=
p
R 0.000761 + 0.50766(0.000083) = 0.0008 = 0.08%
Var.p = 0.000062

4.2 Weekly Analysis
Data on every Tuesday is used to
process the analysis of weekly data. If
Tuesday is public holiday, then data on
the next day will choose for
replacement. Here, there are 23 data for
weekly analysis and result for weekly
analysis is shown in table 4.

Table 4. Market and Stock Return, Alpha, Beta and Variance (Weekly Analysis)

Stock
) 10 (
3
x
R
i

) 10 (
4
x
R
m

) 10 (
..
3
x
I Var

) 10 (
.
4
x
m Var

) 10 (
.
4
x
R R Kov
m i

Beta I Alpha I
) 10 (
4
x

) 10 (
.
3
x
e Var
i

IOI 1.419 4.98 2.772 6.011 6.291 1.047 8.986 2.113
MAYBANK 0.007 4.98 1.227 6.011 2.216 0.369 -1.763 1.145
AMMB -7.251 4.98 2.539 6.011 6.277 1.044 -77.702 1.883
BAT 2.222 4.98 0.099 6.011 -0.648 -0.108 22.758 0.092
SIME 1.794 4.98 0.250 6.011 2.311 0.384 16.027 0.161
COMMERZ -0.947 4.98 1.483 6.011 5.314 0.884 -13.870 1.013
PBB -2.066 4.98 0.356 6.011 1.072 0.178 -21.546 0.337
AFFIN -18.94 4.98 2.264 6.011 -3.543 -0.589 -186.507 2.055
KEMAS 1.994 4.98 4.312 6.011 10.508 1.748 11.246 2.475
GENTING 6.392 4.98 1.749 6.011 7.121 1.185 58.027 0.905

From table 4, the result is almost the
same as daily analysis with 6 counters
recorded positive return level with
highest value is 0.006392 by Genting.
The lowest return level also recorded by
Affin with -0.01894. On the other hand,
the highest unsystematic risk level
recorded by Kemas with 0.002475 and
the lowest is 0.000092 by BAT. Besides,
markets return for this weekly analysis
stood at 0.000498 with risk level at
0.0006011. If comparing relation
between return and risk level, this
weekly analysis shows a negative value
of correlation coefficient. This is good
sign that high return obtained with a
lower risk level for weekly analysis.

Table 5. ERB,
i
C , and Capital Investment Proportion ) (
i
X

Stock ERB
) 10 (
3
x C
i

ERB><
i
C
i
Z
i
X *
i
Z *
i
X
AFFIN 0.033119 -0.052 >* 9.188165 0.479261
GENTING 0.004909 1.079* >* 5.011201 0.261388 5.011201 0.501955
SIME 0.003165 -0.797 >* 4.972173 0.259352 4.972173 0.4980453
KEMAS 0.000811 0.120 <
IOI 0.000805 0.191 <
PBB -0.01482 -1.224 <
BAT -0.01525 -2.005 <
MAYBANK -0.00155 0.102 <
COMMERZ -0.00172 -0.950 <
AMMB -0.0075 -1.426 <

In table 5, the optimal value of
i
C is
0.001079 and there are 3 ERB values
greater than 0.001079. Although this 3
countries is going to form the optimal
portfolio, but there is one counter provide
negative expected return value and this
counter should not been choosen. The
counter that provide negative expected
return is Affin. This is because the and
value of Affin is negative that cause its
expected return become negative. So there
are just 2 counters will be selected into
optimal portfolio for weekly analysis.
These 2 counters are Genting and SIME
with capital investment proportion 50.2%
and 49.8% respectively.

Table 6. Return and Risk Level for Optimal Portfolio (Weekly Analysis)


000498 . 0 =
m
R Var m = 0.000601
Stock Alpha I Beta I
i
R
i
X
Alpha(
i
X )
Beta (
i
X )
i i
X R
Var. e
i
i
X Var. e
i
GENTING 0.005803 1.184661 0.006392 0.502 0.002913 0.5947 0.003209 0.000905 0.000454
SIME 0.001603 0.384462 0.001794 0.498 0.000798 0.19146 0.000893 0.000161 0.00008
Jumlah 0.003711 0.78616 0.004102 0.001066 0.000534
=
p
R 0.003711 + 0.78616 (0.000498) = 0.004103 = 0.41%
Var. p = 0.000639

Tabel 6, show the expected return for
this weekly optimal portfolio is 0.41% at
0.000639 risk level. The expected return for
this portfolio is even higher than weekly
markets return that is 0.000498. This
means that investment through this
combination of stock will provide a higher
return level and at a lower risk level.

4.3 Coefficient Variation
Comparison have made between daily dan
weekly analysis to choose the analysis that
provide high return and low risk through
coefficient variation value.
Daily analysis portfolio:
8425 . 9
0008 . 0
000062 . 0
= = =
p
p
R
CV


Weekly analysis portfolio:
1610 . 6
004103 . 0
000639 . 0
= = =
p
p
R
CV


Weekly analysis showed a lower CV
value that is 6.1610 while daily analysis is
9.8425. This indicate that portfolio of
weekly analysis will be choosen and will
provide a higher return with certain risk
level if compared to daily analysis.


5 Conclusion
Criteria of choosing stock into optimal
portfolio not limited to just comparing
Excess Return to Beta Ratio (ERB) and
Cut-off Rate Optimal ( ) *
i
C , but have to
consider the value of expected return of
each stock.
There are 5 counters forming the
optimal portfolio of daily analysis with
expected return from this portfolio is 0.08%
at 0.000062 risk level. These 5 counters are
BAT, Genting, SIME, Kemas and IOI with
capital investment proportion 47.52%,
34.21%, 13.48%, 2.85%, and 1.94%
respectively.
There are just 2 counters selected into
optimal portfolio of weekly analysis with
expected return stood at 0.41% with risk
level of 0.000639. The 2 counters are
Genting and SIME with capital investment
proportion 50.2% and 49.8% respectively.
Coefficient variation value for daily
analysis is 9.8425 and if compared to
weekly analysis (6.1610), it shows that
portfolio in weekly analysis provides higher
return level with lower risk level. This
mean the best portfolio is from weekly
analysis that is combination of counter
Genting and SIME.

References:
[1] Elton, E. J., & Gruber, M. J., Modern
Portfolio Theory and Investment Analysis,
Fourth Edition, John Wiley and Sons, New
York, 1994.
[2] Fisher, D. E., & Jordan, R. J., Security
Analysis and Portfolio Management, Fifth
Edition, Prentice Hall, New Jersey, 1994.