Professional Documents
Culture Documents
Mean-Variance and Single-Index Model Portfolio Optimisation: Case in The Indonesian Stock Market
Mean-Variance and Single-Index Model Portfolio Optimisation: Case in The Indonesian Stock Market
Portfolio Optimisation:
Case in the Indonesian Stock Market
Adi Kurniawan Yusup*
ABSTRACT
Manuscript type: Research paper
Research aims: This study aims to compare the performance of mean-
variance and single-index models in creating the optimal portfolio.
Design/Methodology/Approach: This study creates optimal portfolios
using the mean-variance and single-index models with daily stock return
data of 38 companies listed on the LQ45 index, IDX Composite index
and Bank Indonesia’s 7-Day (Reverse) Repo Rate from January 1, 2012 to
December 31, 2019. The two models are compared using the Sharpe ratio.
Research findings: The result shows that the single-index model
dominates the Indonesian Stock Exchange (IDX), more so than the mean-
variance model. BBCA has the highest proportion for both mean-variance
and single-index portfolios.
Theoretical contribution/Originality: This study compares two popular
portfolio models in the Indonesian stock market.
Practitioner/Policy implication: This study helps investors to create
optimal portfolios using a model that is more suited to the IDX.
Research limitation/Implication: This study creates the optimal portfolio
without differentiating risk preferences (i.e., risk averse, risk moderate
and risk taker). In addition, this research only uses daily return data and
does not compare it with weekly and monthly data.
https://doi.org/10.22452/ajba.vol15no2.3
2. Literature Review
Ri = a i + b i RM + e
3. Methodology
3.1 Database
This research uses daily closing prices of stocks incorporated in
the LQ45 index, IDX Composite index and Bank Indonesia’s 7-Day
(Reverse) Repo Rate (7-DRRR), from January 2012 to December 2019.
LQ45 represents the 45 most liquid companies on the IDX based on
the following criteria: highest capitalisation, huge transaction value
and prospects for future growth (Malini, 2019). The annual closing
price of stocks are gathered from Yahoo Finance, then matched with
financial statement reports from IDX. The IHSG index is obtained
from Yahoo Finance, and the risk-free rate date is obtained from Bank
Figure
Figure 1: Research procedure
1: Research procedure
Data collection
Daily return of LQ45 stocks, IHSG index and
BI-7DRR
Model comparison
Comparing both optimal portfolios using
Sharpe ratio
𝑘𝑘
Sharpe ratio
∑𝑘𝑘𝑖𝑖=1 𝑅𝑅𝑖𝑖
𝐸𝐸(𝑅𝑅𝑖𝑖 ) = (1) (1)
𝑘𝑘
Where:
Where:
Ri Rate of return of stock i
Ri k RateNumber
of returnofofrate of return
stock i data
E(Ri) Expected return of stock i
k Number of rate of return data
E ( RVariances
i)
and
Expected standard
return of stock i deviation measure the risk of an
individual stock. Variance measures the averaged squared difference
between actual and average return, while standard deviation is the
square root of variance (Bradford & Miller, 2009). Larger variances
and standard deviation indicate greater 7 volatility. Variance and
standard deviation are calculated as follows:
2
å ( Ri - E ( Ri ))
(2)
s i= i= 1
(2)
k
si = s 2
(3) (3)
i
Where:
Ri Rate of return of stock i
k Number of rate of return data
E(Ri) Expected return of stock i
2
S i Variances of stock i
Si Standard deviation of stock i
Where:
Ri Rate of return of stock i
E(Ri) Expected return of stock i
Rj Rate of return of stock j
Sij Covariance of stock i and j
n Number of stocks as sample
Where:
Ri Rate of return of stock i
E(Rp) Expected return on the portfolio
wi Weight of stock i in the portfolio
n Number of stocks included in portfolio
E(Ri) Expected return of individual stock i
å1. w = 100%
1. 1.n n (12)i
å wå =w100%
n
= 100%
1. å w = 100%
i= 1
i
i= 1
(12)
i= 1
(12) (12)
i
i= 1
i (12)
2.
2. wi ³ 0 (13) (13)
2. 2. w02.³ 0
(no short positionwiis³ available)
i wi ³ 0is available) (13)(13) (13)
(no short position
(no short position
(no short is available)
(noposition
short is available)
position is available)
3.5 Single-index Model
5 Single-index
3.5 Single-indexModel
ModelModel
3.5 Single-index
3.5.1 Calculating and Using Linear Regression
5.13.5.1
Calculating
Calculating
3.5.1and and and
Calculating Using Linear
Using
and Regression
Linear
Using Regression
Linear Regression
To find for each stock in single-index model, we regress the return of each stock on
o find
To findand and
To find for each
for stock
each in
stock single-index
in model,
single-index we model,
model, regress theregress
we regress return of each
the return stockstock
of each on on
the market and
return.
88 for Asian
each Journal
The regression stock in
of single-index
equation is as follows:
Business we
and Accounting 15(2), 2022 the return of each stock o
e market return.
the market
the Thereturn.
return.
market regression
The equation
regression isequation
equation
The regression asisfollows:
as follows:
is as follows:
3.5 Single-index Model
3.5.1 Calculating α and β Using Linear Regression
To find α and β for each stock in single-index model, we regress the
return of each stock on the market return. The regression equation
is as follows:
Ri = a i + b i RM + e (14) (14)
Where:
Ri Return of a stock
RM Market return
αi Individual stock’s alpha (intercept), part of return that is not
affected by market
βi Individual stock’s beta (slope), part of return that is affected
by market
e Residual
s 2 (ei ) = s i2 - bi 2s M2 (16)(16)
Where:
s 2 (ei ) = sUnsystematic
2 2 2
i - b i s M risk of stock i
(16)
s 2 (ei ) = s i2 - bi 2s Variances
2
M (total risk) of stock i (16)
bi2 Beta
Beta square
square of
of stock
stock ii
s M2 Variances of
Variances of market
marketindex
index
Excess return is the difference between stock return and the risk-
free rate. In this case, the risk-free rate is 7-DRRR. Excess return to
beta (ERB) measures the relative premium return of a stock toward
unit risk that could not be diversified. ERB is calculated as follows:
Where:
E(Ri) Expected return of stock i
Rf Risk-free rate
bi Beta of stock i
ERBi Excess return to beta of stock i
n
[ E ( Ri ) - R f ]b i
s M2 å
s 2 (ei )
Ci = i= 1
æ
n ç b2 ÷
ö (18)
1 + s M å çç 2 ÷
2 i
i = 1 ç s (ei )
÷
÷
è ø
Where: n
[ E ( Ri ) - R f ]b i
s M2 å Variances
2 of market
s (e )
Ci = E(Ri) i= 1 Expected ireturn of stock i
Rf æ 2 ö
n ç b rate
Risk-free
bi 1 + s Beta çof stock
÷
å i÷
2 i
ç s2(e ) ÷risk of stock i
S2(ei) M Unsystematic
i= 1 ç i ÷
è ø
3.5.3 Determining Optimal Portfolio Stock Candidate and Calculating
Proportion
The prerequisite of a stock included into an optimal portfolio is that
its ERB value is higher than C* which can be written as follows:
ERBi > C * (19) (19)
zi C*
ERBi Z-value
Excess of stock
return
Cut-off
Cut-off to beta
point
point of stock
i (maximum i
b value
(maximum value of Cii))
of C
z = i
( ERB - C*) (20) (20)
bi
C* Cut-off
Beta i
point i(maximum
of stock s 2 (value
i
ei ) of Ci)
s 2Based
(e ) on
i
those z-values,
Unsystematic thestock
risk of proportion
i of each stock in optimal portfolio can be calc
Where:
as:
Based
ERBion those
Where: z-values,
Excess thebeta
return to proportion
of stock iof each stock in optimal portfolio can be calculat
Zi Z-value of stock i
as: bi b i Beta of stock i
zi i=
zC* Z-value
Cut-off -of Cstock
( ERBi point *)(maximum
i value of Ci) (20)
s 2 (ei ) Unsystematic risk of stock i
b i zii
ERBi BetaExcess
of stock i
return to beta of wii = i nn
stock
(2
(2
Based C* Cut-off point (maximum
2 on those z-values, the proportion of
s (ei ) zeach
value
Unsystematic risk of stock
i i
wi =
ofstock
n
å
zCi) in optimal portfolio can be calculat
ii
ii==11 (21)
as:
Z-value of stock i
Based
ERB i on those
Excess z-values,
return to beta the proportion
of stock
i= 1
å
i zi of each stock in optimal
portfolio can be calculated as:
Where: Cut-off point (maximum value of Ci)
C*
Beta of stock i
Where:w z
i =
inwoptimal
i
wii i
Unsystematic risk of stock Weight
Weight of
of stock
stock ii in n portfolio
optimal portfolio
(21) (21)
Based on those z-values, the proportion å of each
zi stock in optimal portfolio can be calculat
w
Excess return to beta z Weight
of istock
i
i i of stock
S-value
S-value of
of i in
stock
stock optimal
i
i portfolio
i= 1
as:
zi nn value
Cut-off point (maximum S-valueCiof
) stock
ofNumber
Number of i
of stocks
stocks that
that meet
meet criteria
criteria (19)
(19)
Where:
Where:
n w Number of stocks
Weight thati in
of stock meet criteria
optimal (19)
portfolio
i
3.5.4 Z
CalculatingS-value of
Expected stock i
Return z
ose z-values, the wproportion
i n
i
of Number
each
Weight of stock
stock iin
inoptimalwand
optimal
of stocks that
Standard
i =meet
i
portfolio
portfolio can
n criteria
Deviation of Portfolio
be calculated
(19) (21)
3.5.4The zexpected
Calculating
i
return ofofstock
Expected
S-value portfolio
Return
i andusing åi= 1single-index
the
Standard zi Deviationmodel is calculated using the fol
of Portfolio
3.5.4 Calculating
The equation:
expected return ofExpected
portfolioReturn
usingand
theStandard Deviation
single-index of Portfolio
model is calculated using the followi
n Number of stocks that meet criteria (19)
The expected return of portfolio using the single-index model is
equation:
Where:
calculated zi the following equation:
wi =using
n E(Rpp) = a pp + b ppE(RMM ) (21) (2
(2
3.5.4 Calculating Expected Return and Standard Deviation of Portfolio
Where: åi= 1 i
wi Weight z of stock i in optimal portfolio
E(Rp ) = a p + b p E(RM ) (22) (22)
The expected
z return of portfolio
S-value of stock iusing the single-index model is calculated using the followi
Where:i
E( Rpp) Expected
Where: Expected return
return of
of portfolio
portfolio
equation:
n E(Rp) Number
Expected return
of stocks thatofmeet
portfolio
criteria (19)
E( Rp ) Expected return of portfolio nn
a
ap pp Portfolio alpha
Portfolioalpha
Portfolio that
alpha that is
that is obtained
is obtained from pp å wiia ii
from
obtained from a =
Weight of stock i in optimal portfolio E ( R ) = a + b E (aRM=) n wii=a=11 (22)
ap
3.5.4 Calculating Expected
Portfolio Return
alpha p obtained
and
that is p from
Standard p Deviation
p åi= 1 ofnni Portfolio
i
pp = å iswcalculated
Where:
S-value of stock i bbp from bmodel iib ii
Portfolio
Portfolio
Portfoliobeta
beta
beta that
that is
that is
is obtained
obtained
obtained from
The expected
ppreturn of portfolio using the single-index n using the followi
åi= 1 i i
ii==11
Number of stocks E b (
that R
p E(R )
meet criteria
Expected
Portfolio(19)
return
beta of
that portfolio
is obtained from b = w b
equation: Ep ( MR) MM ) Expected
Expected
market return
Expected market
market return
return
p
n
Portfolio risks consist of systematic risk and unsystematic risk. Therefore, total portfol
p( RM ) Portfolio
aEPortfolio Expected market
risks consist
alpha return
thatofissystematic
obtained from risk = å unsystematic
a p and wi a i risk.
g Expected Returnisand theStandard Deviation(of RpPortfolio
)unsystematic
= a p +isbthe ( )
variance
Therefore, sum total portfolioEand
of systematic variance p E R
risk.
sum
M
i=of
1 systematic and (22)
n
urn of portfolio using the single-index
unsystematic risk. model is calculated using the following
p = å wi b i
Where: bp Portfolio beta that is obtained from b13 13
i= 1
( R ) Expected
E Portfolio
Portfolio riskof
risk
return = Systematic
= Systematic risk
portfolio risk13 ++ Unsystematic
Unsystematic risk
risk
E ( RMp ) Expected market sreturn2 2 2 2 (23) (23)
P = b p s M + s (e p ) n
Ea(pRp ) = aPortfolio
p + b p E ( R
alpha
M ) that is obtained from a p = å w(22)
ia i
i= 1
Where: Asian Journal of Business and Accounting n15(2), 2022 91
b
Expected return of 2pportfolio 13 b p =
Portfolio beta that is obtained from å wi b i
s Portfolio variance i= 1
2 = b 2s 2risk
Portfolio
Portfolio
Portfolio riskrisk
risk =s Systematic
== Systematic
Systematic
P pp risk
risk
2 M
+++Unsystematic
2 (e )
+s Unsystematic
Unsystematic
2 pp riskrisk
risk
P2 M2
s P2P2s=sPPb==pp2b2sbpM2M2ps+Ms sM+22+
s(e2s()e p()e p )
2 2 2
(23
(2
= Systematic risk + Unsystematic risk pp
(23)
s P2 = b p 2s M2 + sWhere:
2
(e p )
Where:
Where:
Where:
Where:
Where:
22
s
Where:
2 Portfolio variance
Portfolio variance
ssP2PPP2P2s P Portfolio
Portfolio
Portfolio
Portfolio
Portfolio
variance
variance
variance
variance
variance nn
=w å
n
e b Portfolio beta that is obtained from =
bb=ppnn å wni bw wi ii bb ii
bbpppp b p Portfolio
Portfolio beta
beta that is
is obtained
obtained from
from b
å p i= 1 iå
Portfolio
Portfolio
Portfolio
Portfoliobeta
beta
beta that is
thatthat
beta obtained
is obtained
is from
from
obtained b
from
pp = p
b = iibi=ii 1 w b
p n =1 i i
at is obtained from b p s= 2222så 2 wi b i Variances
ii==11
i= 1
Variances
Variances of
of
of market
market
market
ssM Variances of market
M M Variances
Variances of
ofmarket
market
M2i= 1
M
Variances of market
M n nn
ee pp å
nn
rket
å
2
ss 2(s(eepppp()e)p )Portfolio
2
22
2 Portfolio unsystematic risk that is is
obtained from 22s (e 2
(e pp )ss =p22)((å
= ))w= w22i sn 2 (weis) 22
Portfolio
Portfolio
Portfolio
Portfolio unsystematic
unsystematic
unsystematic risk
unsystematic
unsystematic riskrisk
that
risk
that that
is
that
is obtained
obtained
obtained
is
obtained from
from
obtained
from sfrom
from =iis (eii )wii s
s (ep ) Portfolio
ematic risk that is obtained from s (e p ) = å wi s (ei )
2 unsystematic
n
2 risk that is obtained from s (iie==1p1 i)= 1= ii== 11 wi s å
i= 1
i= 1
3.63.6
3.6
3.6
Model
Model
Model
Model
Comparison
Comparison
Comparison
Comparison
3.6
3.6 Model
Model Comparison
Comparison
3.6
Two
Two Model
Two optimal
optimal
optimal Comparison
portfolios,
portfolios,
portfolios, using
using
using thethe
the mean-variance
mean-variance
mean-variance andand
and single-index
single-index
single-index models,
models,
models, with
with
with the
the
Two
Two optimal
Two optimal portfolios,
optimal portfolios,
portfolios, usingusing
using the the mean-variance
the mean-variance
mean-varianceand and single-index
single-index models,
andsingle-index models, w w
Two
the mean-variance expected
expectedoptimal
expectedand returns
returns
returns are
areare
portfolios,
single-index compared
compared
compared using
models, using
usingusing
the
with the
the the Sharpe
mean-variance
the same
Sharpe
Sharpe ratio.
ratio.
ratio. The
TheThe
and Sharpe
Sharpe
Sharpe ratio
single-index
ratio
ratio is
is is commonl
models,
commonly
commonly uu
models, returns
expected with thearesame expected
compared usingreturns
the are compared using the ratio is com
expected
measure returns
portfolio are compared
performance using
(Zakamulin, the Sharpe
Sharpe
2008).
ratio.
ratio.
This
The
ratio
Sharpe
Theisportfolio
Sharpe
computed ratio
as is com
using the SharpeSharpe
expected ratio.
returns
ratio.portfolio
measure
measure The
portfolioThe
Sharpe Sharpe
are ratio
performance
performanceratio
compared is commonly
using used
is commonly
(Zakamulin,
(Zakamulin, used
the2008).
Sharpe
to This
2008). to
This measure
ratio.
ratioThe
ratio is Sharpe as
is computed
computed as thethe
ratio
the is exce
com
excess
excess
measure
performance
measure
relative portfolio
portfolio
toratio
the performance
(Zakamulin,
performance
risk-free rate2008).
divided(Zakamulin,
This
(Zakamulin,
by ratio
risk is 2008).
computed
2008). by
adjustment This
This asratio
the
ratio
using is computed
excess
is computed
asset as
as the
the
(Zakamulin, 2008).
returnThis
relative
relative
measure to
to the
the istocomputed
risk-free
risk-free
portfolio
relative therate
rate as
divided
divided
performance
risk-free the(Zakamulin,
excess
by
by
rate risk
risk return
adjustment
adjustment
divided 2008).
by risk by
by using
using
This asset
asset
ratio
adjustment is return
return
return volatility
by volatility
volatility
computed th(
as(Ga
(Ga
relative
relative to
to the
the risk-free
risk-free rate
rate divided
divided by
by risk
risk adjustment
adjustment by
by using asset return volat
2009).
using
ed by risk adjustment
2009).
2009).
relative
asset
Theby
The
to
The higher
return
using
higher
higher
the the
risk-free
the Sharpe
thevolatility
asset return
Sharpe
Sharpe ratio
ratio
ratio
divided is, is,
(Gatfaoui,
ratevolatilityis, the better
2009).
(Gatfaoui,
the
the
by better
better performing
The
performing
performing
risk adjustment
higher theusing
the
by
the
the assetis.
portfolio
Sharpe
portfolio
portfolio
using assetis.return
is. volat
return volat
ratio is,
2009). The thehigher
better theperforming
Sharpe the portfolio
ratio is, the is. performing the portfolio is.
the better
better
2009).
tio is, the better The
performing higher the Sharpe
the portfolio is. ratio is, performing the portfolio is.
2009). The higher the Sharpe ratio is, the better performing the portfolio is.
E ( RE ()R-p )R- R f
Sharpe ratio = Sharpe ratio = pp ff (23)
Sharpe
Sharpe ratio
ratio == s P )- R (23
(2
E ( Rp ) - R f EPP(( R
sE Rp ) - R f
harpe ratio = Where: Sharpe ratio
ratio =
Sharpe(23) = E ( Rp ) - R f
p f
s P E(Rp) expected return of portfolio
Sharpe ratio = ss PP
Where:
RF
Where:
Where: risk-free rate sP
Sp standard deviation of portfolio
E(ER(ppR
Where:
Where: ) p )expected
expected
expected return
return
return of of portfolio
of portfolio
portfolio
Suppose
Where:
R the Sharpe ratio of mean-variance model is higher
of portfolio risk-free rate
REFF( RF ) risk-free
risk-free rate
rateIn that case, it is concluded that the mean-
than single-index model.
expected return
pp expected return of of portfolio
portfolio
variancesEPP s(model
P p ) standard
R dominates
standard
expected the
deviation
return
standard deviation
deviation of ofIndonesian
of of portfolio stock market more so
portfolio
portfolio
portfolio
than theRFF single-index risk-free rate
model.
risk-free rate On the contrary, if the Sharpe ratio of
RF
on of portfolio the single-index risk-free
model israte higher than the mean-variance model, it
s
Suppose
dominates PP the
Suppose
Suppose the standard
the Sharpe
standard
theIDX more
Sharpe
Sharpe deviation
ratio
deviation
than
ratio
ratio of
ofthe
of of
of portfolio
mean-variance
mean-variance
mean-variance
mean-variance model
portfolio model
model.
model is is higher
is higher
higher than
than
than single-index
single-index
single-index mom
mo
sP standard deviation of portfolio
f mean-variance that
that
that case,
model
case,
case, isithigher
ititisis is concluded
concludedthanthat
concluded thatthat
thethe
single-index
the mean-variance
mean-variance
mean-variance model
model. Inmodel
model dominates
dominates
dominates thethe
the Indonesian
Indonesian
Indonesian stocm
stock
stock
e mean-variance Suppose
moreso so
Suppose
model
more
more so thanthe
thethe
dominates
than
than the
the Sharpe
Sharpe ratio
single-index
ratio
the Indonesian
single-index
single-index of
of
model.
model. mean-variance
model.OnOn
stock
On thethe
market
the contrary,
contrary, model
contrary,
mean-variance model
if
if the
the is
if the
is higher
Sharpe
higher
Sharpe
Sharpe than
ratio
than
ratio
ratio of single-in
of of the sin
single-in
the
the single
single
Suppose the Sharpe ratio of mean-variance model is higher than single-in
model
model. On thethat case,
contrary,
model
model is is higher
itifis
is higherthe
higher than
concluded
Sharpe
than
than the
theratio
the mean-variance
thatofthe model,
themean-variance
single-index
mean-variance
mean-variance model,
model, itmodel
dominates
itit dominates
dominates thethe
dominates
the IDXIDX
IDX themore
more
more thanthan
Indonesia
than theth
the
that case, it is concluded that the mean-variance model dominates the Indonesia
thatvariance
case, itmodel.
is concluded that the mean-variance model dominates the Indonesia
-variance model, itso
variance
more sodominates
variance
more model.
than
model.
than the the
the IDX more model.
single-index
single-index than the On
model. mean-
On the
the contrary,
contrary, if
if the
the Sharpe
Sharpe ratio
ratio of
of th
th
more so than the single-index model. On the contrary, if the Sharpe ratio of th
model
model is is higher
higher than
than the
the mean-variance
mean-variance model,
model, it
it dominates
dominates the
the IDX
IDX more
more tt
model is higher than the mean-variance model, it dominates the IDX more t
variance
variance model.
model.
variance model. 14 14
14
14
92 Asian Journal of Business and Accounting 15(2), 2022
14
4. Results
Portfolio 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
96
ACES 4.23 4.38 5.01 5.97 6.42 6.94 7.24 7.62 9.27 10.91 11.69 12.42 13.75 14.79 15.59 15.72 16.53 16.88
ADRO 1.77 1.88 1.76 1.76 1.67 1.56 1.42 1.11 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
AKRA 4.30 4.41 3.92 3.43 3.10 2.71 2.41 1.99 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
ANTM 2.98 3.13 2.32 1.48 0.92 0.32 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
ASII 5.54 5.72 4.02 2.04 0.85 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
BBCA 18.99 19.61 21.37 24.29 25.67 27.05 27.56 28.44 30.61 30.97 30.83 30.35 28.39 24.16 20.12 19.41 13.83 10.48
BBNI 2.99 3.21 3.10 3.15 3.00 2.80 2.65 2.21 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
BBRI 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.14 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
BBTN 1.82 1.95 1.81 1.78 1.68 1.61 1.53 1.34 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
BMRI 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
BSDE 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
CPIN 0.00 0.00 0.00 0.31 0.80 1.34 1.62 1.98 3.53 4.93 5.58 6.33 8.08 10.17 11.92 12.22 14.23 15.23
CTRA 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
ERAA 0.00 0.00 0.06 0.09 0.71 0.96 1.10 1.28 1.94 2.20 2.29 2.38 2.46 2.26 2.03 1.99 1.63 1.40
EXCL 3.33 1.35 2.79 2.09 1.63 1.13 0.79 0.30 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
GGRM 5.06 5.06 3.65 1.80 0.78 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
ICBP 5.06 5.17 5.48 5.97 6.20 6.40 6.41 6.39 5.84 4.54 3.86 2.76 0.00 0.00 0.00 0.00 0.00 0.00
INCO 0.08 0.01 0.10 0.11 0.08 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
97
Number of
26 26 27 26 27 23 21 21 17 15 15 13 11 10 9 9 8 8
stocks used
4.3 Single-index Model
The initial step to determine optimal portfolio using single-index
model is calculating α and β using regression Equation 14. Using the
result of α and β, unsystematic risk and excess return to beta (ERB)
can be found by consecutively using Equations 16 and 17. Table 3
shows α, β, unsystematic risk and ERB of each 38 individual stocks.
Alpha shows unique stock return that is not affected by market
return. The highest α is owned by WIKA, which is 0.01. β shows
sensitivity of the stock toward market return. Positive β means an
increase in market return will result in an increase in stock return.
Meanwhile, a negative β means an increase in market return will
result in a decrease in stock return. SMRA has the highest beta,
0.9131. This means when market return increases for 1%, SMRA’s
return will increase 0.9131%. Of 38 stocks analysed, only ASII has
negative β during the research period. It means when market return
increases, ASII return will decrease.
Unsystematic risk is risk that is related to particular stock or
security. From Table 3, WIKA has the highest unsystematic risk of
0.1848. BBCA is stock that has the lowest unsystematic risk. This
shows that most of its risk is affected by the market and it is good
sign for the investor to invest in this company. ERB is the comparison
between excess return and β. The highest ERB of these 38 stocks is
CPIN, at 0.3819. Stocks that have negative or positive ERB due to
negative excess return and negative beta have to be taken out from
the list because they are ineligible for the establishment of an optimal
portfolio. Of 38 stocks, eight stocks have to be taken out. The seven
stocks with negative ERB are PTBA, UNTR, PGAS, GGRM, ITMG,
ANTM and EXCL, while the one stock with positive ERB, negative β
and negative excess return is ASII.
5. Conclusion
This research highlights the difficulties faced by new investors in
creating optimal portfolios. Two well-known portfolio models that
can be used are mean-variance and single-index. Both models offer
investors the ability to create portfolios with maximum return on any
desired level of risk, or minimum risk with any desired of level of
return. However, the mean-variance model is more complex because
it has a larger number of covariance data to process. In addition, this
research shows that the single-index model dominates the Indonesian
stock market compared to the mean-variance model.
References
Aliu, F., Pavelkova, D., & Dehning, B. (2017). Portfolio risk-return
analysis: The case of the automotive industry in the Czech
Republic. Journal of International Studies, 10(4), 72-83. https://doi.
org/10.14254/2071-8330.2017/10-4/5
Aryusmar, D. (2020). The effect of the household consumption,
investment, government expenditures and net exports on
Indonesia’s GDP in the Jokowi-JK era. Journal of Critical Reviews,
57(8). https://doi.org/10.17762/pae.v57i8.1318
Bank Indonesia. (2020). Pertumbuhan ekonomi Indonesia triwulan IV
2020 melanjutkan perbaikan. https://www.bi.go.id/id/edukasi/
Documents/Infografis-Pertumbuhan-Ekonomi-Indonesia-
Triwulan-IV-2020.pdf
Benniga, S. (2006). Principles of Finance with Excel. Oxford: Oxford
University Press.