Mean-Variance and Single-Index Model

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.

Keywords: Mean-variance Model, Optimal Portfolio, Single-index Model

JEL Classification: G11

* Corresponding authors. Adi Kurniawan Yusup is a doctoral degree student of Widya

Mandala Catholic University, Surabaya, Indonesia, e-mail: adikurniawanyusup@hotmail.
com

https://doi.org/10.22452/ajba.vol15no2.3

Asian Journal of Business and Accounting 15(2), 2022 79

1. Introduction
The pandemic created significant problems in the business world.
Most sectors worldwide, including in Indonesia, had to stop their
operations temporarily to minimise the spread of Covid-19. The
Central Bureau of Statistics of Indonesia reported that the country’s
GDP shrank by 5.32% in the second quarter of 2020, 3.49% in the
third quarter and 2.19% in the last quarter (Bank Indonesia, 2020).
The Indonesian government rolled out several policies to boost
its GDP, including easy fiscal and monetary policies to increase
consumption, investment, government spending, and export, as the
four components of GDP (Aryusmar, 2020). As a result, many became
interested in investing with the relaxation of investment policies. The
Indonesia Stock Exchange (IDX) accounted for 3.88 million investors
at the end of 2020 (Depository, 2020), and exceeded its annual target
of 3 million last year.
Return is one of the factors that motivates investors as a reward
for their courage in taking risks. Many investors invest in stocks to
get a high return instantly (Hamka et al., 2020). They should use
surplus funds after financing their primary need to invest because
there is a risk of loss. Moreover, investing in stocks needs long
periods to achieve high returns (Partono et al., 2017). However,
Indonesia’s Financial Service Authority (2019) showed that financial
literacy in the country was at 38.03%, which is considered low.
This is reflected in many using loans and primary-need money for
investment. In addition, low financial literacy could increase the
information asymmetry that increases herding behaviour (Din et al.,
2021), that is, investing is the same stocks as others with the hope
of high returns (Fransiska et al., 2018). Nevertheless, this behaviour
causes investors to pick the wrong stocks for their portfolio, which
leads to losses. Therefore, building an investment portfolio is a critical
issue for investors.
Generally, investors are risk-averse (Reilly & Brown, 2003). They
expect high returns with low risk, whereas investment theory states
that there is a linear relationship between risk and return, i.e., high
risk, high return (Rui et al., 2018). Hence, risk should be reduced to
get the desired return with lower risk. A fundamental approach to
minimising risk is diversification (Lee et al., 2020). Diversification is
the allocation of a fixed portfolio across different types of securities
and asset classes that limit exposure to any source of risk (Bodie et
al., 2014). The optimal portfolio for the most returns with the lowest
risk is located in the efficient frontier curve (Fama & French, 2004).
However, creating an optimal portfolio is a challenge, especially

80 Asian Journal of Business and Accounting 15(2), 2022

for new investors. Investors need to work with a lot of information,
e.g., historical price, trends and firm performance. They also need
analytical tools to get the right composition of stocks in their
portfolio.
One approach used to obtain an optimal portfolio is modern
portfolio theory (MPT) or mean-variance analysis. MPT was
established by Markowitz (1952), and attempts to maximise portfolio
return for any given risk, or equivalently, minimise risk for any
given amount of return. Both ways will produce the same result in
creating an efficient portfolio. The basic concept behind MPT is that
assets in a portfolio should not be viewed in isolation but should
be evaluated by how it affects the portfolio’s risk and return on the
whole (Omisore et al., 2012). MPT seeks to reduce total variance as a
component of risk of portfolio return by combining different assets
with returns that are not positively correlated. This theory assumes
that investors are rational and markets are efficient.
The Markowitz model provides several risk-return combinations
which investors can choose from based on their preferences (Putra &
Dana, 2020; Rigamonti, 2020). However, implementing the Markowitz
model is very time-consuming, because it requires a lot of estimations
to fill the covariance matrix. The model does not provide guidelines
for forecasting the security risk premium, that is important for the
construction of an efficient frontier of risky assets (Singh & Gautam,
2014). Sharpe (1963), in trying to simplify the Markowitz model,
came up with a single-index model that reduces computational
requirements and data. This model solves the portfolio optimisation
problem by showing a linear relationship between risk and return—
which is simpler to understand—and considers that the relationship
between securities occurs only through their individual relationship
with some indices of business activity. As a result, covariance data
is reduced from (n2 – n)/2 under the Markowitz model to only n
measures of each security as it relates to the index (Chitnis, 2010).
Nevertheless, single-index models may not be the most efficient with
respect to volatility.
There are limited studies on comparing the mean-variance and
single-index models in creating a portfolio. Furthermore, those
studies show inconsistent results. Ozkan and Cakar (2020) show that
the mean-variance model performs better in the markets of developed
countries, while the single-index model performs better in developing
countries. On the other hand, Chasanah et al. (2017), whose research
sample is a developing country market, find that optimal portfolio
formation with the mean-variance model is more dominant than the

Asian Journal of Business and Accounting 15(2), 2022 81

single-index model. Nyokangi (2016) shows that the choice between
mean-variance and single-index models depends on the time span
of study. On the other hand, Yuwono and Ramdhani (2017) find no
significant return difference between the mean-variance model and
single-index model.
Therefore, this paper aims to construct portfolios using the mean-
variance and single-index models, and compare the performance of
each in the IDX. The sample used in this research are stocks listed in
the LQ45 index, the most liquid stocks in the IDX. This research is
expected to give insight to investors—especially new investors—on
constructing an optimal portfolio.

2. Literature Review

2.1 Portfolio Theory (Markowitz Theory)

Asset allocation for the sake of diversification in financial markets
has become an important issue for investors. Investors expect asset
allocation to give them the highest return with a particular risk, or
the lowest with a certain expected return. Diversification is a risk
reduction technique by spreading investment in different securities
and asset classes (Bodie et al., 2014; Jayeola et al., 2017). In the
traditional approach of diversification, the greater the number of
securities invested in portfolio, the lower the risk (Lekovic, 2018).
Based on the law of large numbers, this theory is supported by
some its proponents, like Hicks (1935), Williams (1938), and Leavens
(1945). Nevertheless, portfolio efficiency decreases as the number of
portfolios increases because it results in excessive diversification. This
can reduce portfolio risk, but ignores the correlation between returns
on different assets (Francis & Kim, 2013).
Simple diversification was rejected by Markowitz (1952), who
was awarded the Nobel Prize in 1990 for his essay “Portfolio
Selection” and his book Portfolio Selection: Efficient Diversification
(1959). He initiated the first quantitative theory of portfolio selection
and management, currently known as MPT. This is a framework
to create an investment portfolio based on the maximisation of
expected returns and the simultaneous minimisation of investment
risk (Fabozzi et al., 2002; Rodríguez et al., 2021). This theory uses
return and risk in a coherent framework (Chen & Pan, 2013) and
considers covariances among the stocks in the portfolio. This model
is also called the mean-variance (MV) portfolio theory. Return
is based on historical data while risk is obtained from standard
deviation of returns. The Markowitz model provides a solution to

82 Asian Journal of Business and Accounting 15(2), 2022

overcome trade-off problems between risk and return in selecting
a combination of assets in an optimal portfolio. The investor has a
choice of either maximising return with any acceptable risk or to
minimise risk with any acceptable level of return. Both options use
the same mathematical process.
Efficient frontier refers to several alternative portfolios that give
the highest return on certain risks or lowest risk on certain returns
(Fabozzi et al., 2002; Letho et al., 2022). Nevertheless, not all portfolios
on the efficient frontier are optimal. An optimal portfolio is located
on the efficient frontier chosen by investors based on their individual
preferences shown by utility. Theoretically, an optimal portfolio
is defined as a tangency point between the efficient frontier and
the investor’s indifference curve, which presents the utility of the
investor. A higher indifference curve has larger certainty equivalents
and larger expected utility (N. Kumar, et al., 2014).
However, this theory has several limitations. Mean-variance
portfolios find portfolio weight based on first moment and
the parametric representation of second moment, assuming
no predictable time variation. It assumes constant investment
opportunities, and its static mean-variance solution is myopic and
does not consider events beyond the present (Jones, 2017). MPT
also assumes continuous pricing, a world in which markets are free,
societies are stable, and investors are rational wealth maximisers
(Curtis, 2004). In fact, this assumption is contradicted by observations
of investors who get swept up in herd behaviour, whereby the
investor tries to follow the decisions made by others and undervalues
information available in the market (Bikhchandani & Sharma, 2000).
They do not consider risk and return in choosing their portfolio.
Moreover, these models need complex statistics-based mathematical
modelling and formulas to support the concept and theoretical
assumptions (Mangram, 2013). This is problematic when it is used
to build portfolios from large stock groups, with many estimations
needed for asset selection.

2.2 Single-index Model

The single-index model (SIM) concept was developed by Sharpe
in 1963. This model simplifies Markowitz’s mean-variance model
(Mahmud, 2019) to overcome its complexity. This model is no longer
based on inter-asset correlation, but instead on the assumption that
co-movement between stock return is due to movement in market
returns, or to be more specific, returns of a broad market index, RM
(Frankfurter et al., 1976; Hadiyoso et al., 2015). The single-index

Asian Journal of Business and Accounting 15(2), 2022 83

model specifies two sources of uncertainty for return: systematic and
unique uncertainty. Systematic uncertainty, or beta (β), is uncertainty
from the market or macroeconomics, while unique uncertainty, alpha
(α), is uncertainty from the company or industry itself. Hence, a
single-index model is written as follows:

Ri = a i + b i RM + e

From the equation, the difference between expected return (ai

+ biRM) and actual return (Ri) is symbolised by e,, residual return.
An important concept in the single-index model is β terminology.
This refers to sensitivity towards market return and is usually
predicted using historical data (B. R. Kumar & Fernandez, 2019).
The higher the β, the more sensitive the stock is to market return.
The basic assumption used in the single-index model is that stocks
are correlated only if they have the same response toward market
return. Therefore, the covariance between two stocks only can only
be calculated based on the similarity of their response toward market
returns.
The optimal portfolio of the single-index model is obtained by
sorting the value of excess return to β (ERB) for each stock from the
largest to the smallest (Elton et al., 1976). The negative ERB value or
positive ERB that is caused by negative excess return and negative β
has to be removed. Those ERB values are compared with the value
of cut-off point of the shares (C*). C* is the largest value Ci of each
individual stock. Stocks with larger ERB value compared to the C*
are included in the optimal portfolio and the weights are calculated,
while others are taken out.

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

84 Asian Journal of Business and Accounting 15(2), 2022

Indonesia’s 7-DRRR report.
A total of 38 companies on the LQ45 index are used as the
research sample. These are the only firms on the LQ45 index with
enough information in the examined period. The other seven LQ45
companies do not have complete data—some issued stocks after
2012, and some had their stocks suspended, resulting in no trade
for several days. When those stocks were traded again, the price
could differ significantly from the last trading period, showing high
volatility (Garcíaa et al., 2015). These 38 companies comprise various
sectors, as listed in Appendix 1.

3.2 Research Procedure

Figure 1 shows the steps of constructing a portfolio using both the
mean-variance and single-index models and comparing the results.

Figure
Figure 1: Research procedure
1: Research procedure

Data collection
Daily return of LQ45 stocks, IHSG index and
BI-7DRR

Expected return and standard deviation

Calculating rate of return, expected return and
standard deviation of each stock as research
sample and IHSG index

Mean-variance model Single-index model

1. Creating covariance matrix
2. Calculating expected return and
standard deviation of several possibility
of portfolios
3. Creating efficient frontier by finding
several portfolio possibilities that give
maximum return given certain amount
of risk (standard deviation)
1. Calculating and using linear
regression
2. Calculating unsystematic risk, excess
return to beta (ERB) and Ci
3. Determining optimal portfolio stock
candidate and calculating the
proportion
4. Calculating expected return and
standard deviation of optimal portfolio

Model comparison
Comparing both optimal portfolios using
Sharpe ratio

3.3 Expected Return and Standard Deviation of Individual Stock

The expected return of individual stock is the arithmetic mean of daily return during the
research
period (Benniga, 2006) asoffollows:
Asian Journal Business and Accounting 15(2), 2022 85

𝑘𝑘
 Sharpe ratio

3.3 Expected Return and Standard Deviation of Individual Stock

3.3 Expected
The expected return ofReturn and Standard
individual Deviation
stock is the of Individual
arithmetic Stockreturn during the
mean of daily
The
research expected
period return
(Benniga, of as
2006) individual
follows: stock is the arithmetic mean of
daily return during the research period (Benniga, 2006) as follows:

∑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

3.4 Mean-variance Model

3.4.1 Creating Covariance Matrix

Covariance measures how returns on two risky assets move in
tandem (Bodie et al., 2014). If the covariance is positive, two assets
move together while they vary inversely if the covariance is negative.
Covariance is calculated as follows:

86 Asian Journal of Business and Accounting 15(2), 2022

 s ij = å Pr( scenario)[ Ri - E ( Ri )][ R j - E ( R j )] (4)
(4)
scenarios
n2 - n
Total covariance data =
2 si = s 2i (5) (5)
Total covariance data =

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

The covariance matrix is a tool to calculate the standard deviation

of a stock portfolio that is used to quantify risk. The format of the
covariance matrix is as follows:

é var 2 cov( x1 , x2 ) cov( x1 , xn ) ùú

ê x1
ê cov( x , x ) var 2 x2 cov( x2 , xn ) úú
Covariance matrix =
ê 2 1
(6)
ê ú
ê ... ... ú
ê 2 ú
ê cov( xn , x1 ) cov( xn , x2 ) ............. var xn ú
ë û

3.4.2 Calculating Expected Return and Standard Deviation of Portfolio

The expected return of portfolio is the weighted average of individual
stock return (Rachmat & Nugroho, 2013) and can be calculated using
the following expression.
n
E ( RP ) = å wi E ( Ri ) (7) (7)
i= 1

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

Risk in portfolio management is measured through uncertainty

of the returns. Correlation, coefficients, weights of each security and
variance of its stocks are factors influencing portfolio risk (Aliu et al.,
2017). Risk can be divided into two types: systematic risk (market

Asian Journal of Business and Accounting 15(2), 2022 87

 n n
s 2 pn = nån ån wi w j s ij (8)
s 2åp =å å
s 2 pwå=
2 n n
s p= i= 1 jw
j s iijwunsystematic
= 1w
j s ijw w s
risk or non-diversifiable risk)
i= 1 j =
i and
i=11 j = 1
åi= 1 åj= 1 i j ij (8)
risk (diversifiable (8) (8)
risk) (Wagdi & Tarek, 2019). Risk of the portfolio proxied by standard (9)
s p = s 2p
deviation can be calculated as2 follows: (9) (9)
s p =s p =
2
s pss =p s 2p (9)
p
n n
Where:
Where:
Where:2Where:
s 2p = å å wi w j s ij (8) (8)
s p Variance of the portfolio i= 1 j = 1
s 2 p s 2 p Variance ofVariance
the
s 2Variance ofportfolio
the portfolio
of the portfolio
wi p
Weight of stock i in the portfolio s p = s 2p (9) (9)
wi wi Weight of stock
Weight of i in
stock the
i in portfolio
the portfolio
wi WeightWeight of stockofjstock in the portfolio
in thei portfolio
wj
wj wj Weight
Where: of stock
Weight Weightj inofthe
of stock inportfolio
the jportfolio
j stock in the portfolio
n wj Number 2 of stocks included in portfolio
Number s
Numberofp stocks Variance
Variance
included
of stocks of
ofinthethe portfolio
inportfolio
portfolio
n n n Covariance Number of included
between stocks
stock includedportfolio
i and stock in portfolio
j
s ij w Weight of stock i in the portfolio
s ij s ij Covariance
Covariance
i between
betweenstock i
stockand i stock
and j
stock j
sp wj Covariance
s ij Standard Weightbetween
deviation of stockstock
of portfolio i and
j in the stock j
portfolio
sp sp Standardn deviation
Standard ofdeviation
Number
deviation
Standard portfolio
of of stocks included in portfolio
portfolio
of portfolio
s pSij Covariance between stock i and stock j
S p Standard deviation of portfolio
3.4.3 Creating Efficient Frontier Curve
4.3 Creating Efficient Frontier Curve
3.4.3 Creating Efficient Frontier Curve
3.4.3 Creating Efficient Frontier Curve
An approach usedCreating
3.4.3 in this research to find Curve
Efficient Frontier optimal portfolios located in efficient frontier curve
n approach used
An approach in
used
An approach this
in research
this
used to
research
inused
this find
to
research optimal
find optimal
to find portfolios
portfolioslocated in efficient
located infrontier
in efficient curve
frontier curve curv
is generating
Anportfolio
approach with maximum
in thisreturn
research foroptimal
any
to find portfolios
given standard
optimal located
deviation.
portfolios efficient
located frontier
generating portfolio
is generating with
inportfolio
efficient
is generating maximum
with maximum
frontier
portfolio return
with curve
maximum for
return any
for any
is generating
return given for standard
given
any givendeviation.
standard
portfolio deviation.
with
standard maximum
deviation.
return for any given standard deviation. n
Max E ( RP )n = ån wi En( Ri ) (10)
MaxMax (=RPå) E
E ( RPE)Max = (wå = 1(w
iE R E ) ( R )
Ri P ) =i i å iwi E ( Ri ) (10) (10) (10)
i= 1 i= 1
(10)
i= 1
Some restrictions used to find Max E ( RP ) are as follows:
Some restrictions
Some usedrestrictions
restrictions to find
used Maxto
toused
find Max ( REPto
Efind )( Rare
P )Eas ( follows:
are Ras follows:
P ) are as
SomeSomerestrictions used Max find Max E(Rp) are as follows:
follows:
n n
Given standard deviation: s p =n nn n w ws
n i nj ijå å (11)
Given standard
Given deviation:
standard
Given standard
Given
s
deviation:
deviation: =s
standard deviation:
p p = i= 1w jw s å åå å å å
s p = i j iij j s ijwi w j s ij
= 1w w
i= 1 j i==11 j = 1
(11)(11)
(11) (11)
i= 1 j = 1
1. n

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

3.5.2 Calculating Unsystematic Risk, Excess Return to Beta (ERB) and Ci

Unsystematic risk is risk specific to individual company (firm-specific
risk) (Masry & Menshawy, 2017). There are two methods to calculate
unsystematic risk: direct and indirect. Using the direct method,
unsystematic risk uses residuals of a factor model, such as CAPM
and the Fama and French (1993) model. Using the indirect method,
unsystematic risk can be calculated using the following formula:

Unsystematic risk = Total risk – systematic risk (15)

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:

Asian Journal of Business and Accounting 15(2), 2022 89

 E ( Ri ) - R f
ERBi = (17) (17)
bi

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

To determine which stocks are included in a portfolio, the

cut-off point (C*) has to be determined. The cut-off point value is
obtained from the highest amount of Ci of each individual stock. Ci
is calculated as follows:

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)

Among all stocks that fulfil criteria (19), we have to calculate

z-value to determine the stock weight in the optimal portfolio. The
formula to calculate z-value with no short position available is as
follows:

90 Asian Journal of Business and Accounting 15(2), 2022

 zi b ii Z-value Betaofof
Beta ofstock
stocki ii s (ei )
stock
b i s 22(eii)BetaUnsystematic
of stock i
Unsystematic risk
risk of
of stock
stock ii
Where:
s 2 (eERB
) ii Unsystematic
ERB Excess risk of
Excess return
return to stock
to beta ofi stock
beta of stock ii
i

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

4.1 Expected Return and Standard Deviation

Table 1: Expected return, variances and standard deviation of individual

stock and market

Securities Expected return (%) Variances (%) Standard deviation (%)

ACES 0.098 0.064 25.343
ADRO 0.034 0.081 28.504
AKRA 0.041 0.053 23.123
ANTM 0.003 0.072 26.746
ASII 0.016 0.039 19.698
BBCA 0.085 0.022 14.739
BBNI 0.056 0.038 19.570
BBRI 0.080 0.039 19.665
BBTN 0.056 0.054 23.292
BMRI 0.062 0.038 19.426
BSDE 0.044 0.059 24.361
CPIN 0.101 0.084 29.049
CTRA 0.076 0.084 28.952
ERAA 0.089 0.117 34.250
EXCL 0.019 0.075 27.388
GGRM 0.013 0.043 20.687
ICBP 0.067 0.047 21.753
INCO 0.051 0.093 30.447
INDF 0.046 0.036 18.955
INKP 0.146 0.102 31.967
INTP 0.034 0.057 23.839
ITMG -0.029 0.069 26.361
JPFA 0.081 0.088 29.617
JSMR 0.031 0.036 19.063
KLBF 0.064 0.039 19.739
MEDC 0.074 0.099 31.492
MNCN 0.055 0.087 29.452
PGAS 0.018 0.069 26.205
PTBA 0.022 0.073 27.011
PTPP 0.100 0.075 27.431
PWON 0.093 0.072 26.810
SMGR 0.031 0.052 22.907
SMRA 0.065 0.081 28.459
TBIG 0.075 0.052 22.701
TLKM 0.069 0.031 17.606
UNTR 0.019 0.057 23.817
UNVR 0.061 0.036 18.924
WIKA 1.023 18.480 429.887
Market Index 0.031 0.009 9.502

Asian Journal of Business and Accounting 15(2), 2022 93

 The first step to determine optimal portfolio using mean-variance
and single models is calculating expected returns, variance and
standard deviation of each stock. Table 1 presents daily expected
return, standard deviation and variances of 38 LQ45 stocks and the
IDX Composite.
Table 1 shows that 37 stocks have positive expected return, while
one stock, ITMG, has negative expected return. The five shares with
the highest expected return are WIKA with 1.02% returns, INKP with
0.15%, and CPIN, PTPP and ACES with 0.10. Meanwhile, stocks with
higher risk, shown by the higher variances, are WIKA with 18.48%
variance, ERAA with 0.12%, INKP and MEDC with 0.10%, and INCO,
JPFA and MNCN with 0.09%.
Aside from that, the IDX Composite, as Indonesia’s market index,
shows a daily return 0.031%, a variance of 0.009% and standard
deviation of 0.950%. The risk-free rate, calculated from the average
of 7-DRRR during the research period, is 6.02% per year. Since this
research uses daily return, the annual risk-free rate is divided by 241,
the average number of days in a year for the research period. Hence,
the risk-free rate used is 0.025%.

4.2 Mean-variance Model

As the number of stocks analysed is 38 stocks, the number of
382 - 38
covariance data that has to be calculated is = 703 data.
2
Based on these covariance data, the maximum return portfolio at any
given desired risk level can be calculated. Table 2 shows 18 portfolio
compositions along the efficient frontier line. 18 portfolios are
generated using 38 available stocks and changing their weightings.
Portfolio 1 is the minimum-variance portfolio. This expected
daily return of this portfolio is 0.056%, the variance is 0.856% and the
standard deviation is 9.253%, with 26 stocks. BBCA has the highest
proportion (18.99%) in this minimum-variance portfolio, as it is the
least risky among the 38 stocks selected. The highest return per unit
risk is shown by Portfolio 15. Its return to risk ratio is 0.99892, which
means 0.99892% return obtained for every 1% risk borne.
As shown in Table 2, as the return is expected to increase, the
number of stocks used to create portfolios with the lowest standard
deviation decreases. To create portfolios 1 to 8, more than 20 stocks
are selected. To create portfolios 9 to 13, between 10 and 20 stocks are
selected, while for portfolio 14 to 18, less than 10 stocks are selected.
This result has the same characteristics with a previous study
conducted by Garcíaa et al. (2015). ACES, BBCA, INKP and WIKA are

94 Asian Journal of Business and Accounting 15(2), 2022

the four only stocks that are always selected in 18 efficient portfolios.
Investors can choose portfolios located in the efficient frontier
because they are the best portfolios of all possible combinations
(Halicki & Uphaus, 2014). Figure 2 shows the efficient frontier
line using the 18 points in Table 2. Individual preferences and
risk tolerance affect investment decisions (Lan et al., 2018). Those
preferences are shown by the utility curve. The intersection between
the utility curve and efficient frontier is the portfolio combination
chosen by the investor.

Figure 2: Efficient frontier

0.16%
0.14%
0.12%
0.10%
0.08%
0.06%
0.04%
0.02%
0.00%
8.00% 9.00% 10.00% 11.00% 12.00% 13.00% 14.00% 15.00%

Notes: Efficient portfolio = 18

Asian Journal of Business and Accounting 15(2), 2022 95

 Table 2: 18 efficient portfolio composition (%)

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

Asian Journal of Business and Accounting 15(2), 2022

INDF 2.17 2.20 1.81 1.27 0.63 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
INKP 1.29 1.31 2.27 3.47 4.08 4.77 5.19 5.70 8.46 12.84 15.05 17.36 22.23 27.51 32.03 32.80 38.15 40.86
INTP 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
ITMG 2.91 2.94 1.58 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
JPFA 1.41 1.40 2.01 2.69 2.89 3.10 3.17 3.22 3.23 3.02 2.92 2.71 1.97 0.47 0.00 0.00 0.00 0.00

Portfolio 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
JSMR 6.48 5.96 5.66 4.50 3.70 2.74 2.01 1.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
KLBF 2.94 3.07 3.47 4.11 4.40 4.63 4.63 4.61 3.61 1.67 0.66 0.00 0.00 0.00 0.00 0.00 0.00 0.00
MEDC 2.33 2.38 2.76 3.23 3.39 3.58 3.67 3.71 3.56 2.85 2.47 1.96 0.65 0.00 0.00 0.00 0.00 0.00
MNCN 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
PGAS 1.05 1.14 0.44 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
PTBA 0.78 0.82 0.91 1.01 0.69 0.27 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
PTPP 0.00 0.00 0.00 0.42 0.75 1.10 1.28 1.55 2.67 4.28 5.09 5.84 7.02 7.49 7.63 7.63 7.49 7.27
PWON 0.00 0.00 0.00 0.00 0.28 0.67 0.87 1.12 2.28 3.78 4.53 5.21 6.11 6.13 5.94 5.89 5.29 4.77
SMGR 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
SMRA 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
TBIG 7.88 7.92 8.35 8.90 9.14 9.38 9.49 9.58 9.68 9.50 9.39 9.11 7.88 5.10 2.44 1.97 0.00 0.00
TLKM 7.02 7.23 7.72 8.48 8.68 8.88 8.93 8.91 7.89 5.55 4.36 2.48 0.00 0.00 0.00 0.00 0.00 0.00
UNTR 0.96 1.06 0.54 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
UNVR 6.57 6.65 6.99 7.48 7.67 7.82 7.76 7.61 5.79 2.18 0.36 0.00 0.00 0.00 0.00 0.00 0.00 0.00
WIKA 0.05 0.05 0.10 0.17 0.20 0.24 0.26 0.30 0.49 0.78 0.93 1.10 1.47 1.91 2.30 2.37 2.85 3.10
Several calculations
E(Rp) 0.056 0.057 0.062 0.071 0.075 0.079 0.081 0.083 0.091 0.100 0.104 0.108 0.116 0.124 0.132 0.133 0.141 0.145
Variances σp 2) 0.856 0.858 0.860 0.876 0.888 0.903 0.911 0.922 0.979 1.074 1.135 1.205 1.366 1.555 1.735 1.767 1.996 2.115

Asian Journal of Business and Accounting 15(2), 2022

Standard
9.253 9.262 9.273 9.357 9.421 9.502 9.545 9.603 9.893 10.362 10.654 10.975 11.686 12.469 13.171 13.293 14.127 14.545
deviation (σp)
Return/risk 0.602 0.612 0.671 0.754 0.793 0.830 0.846 0.864 0.923 0.961 0.974 0.983 0.994 0.998 0.999 0.999 0.999 0.999
Sharpe ratio 0.333 0.342 0.402 0.487 0.528 0.567 0.585 0.604 0.670 0.720 0.739 0.756 0.781 0.798 0.809 0.811 0.822 0.827

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.

98 Asian Journal of Business and Accounting 15(2), 2022

 Table 3: Alpha, beta, unsystematic risk and ERB of 38 individual stocks

Unsystematic risk Excess return to

Stocks Alpha (α) Beta (β)
(σ(e)) beta (ERB)
ACES 0.0010 0.0050 0.0006 0.1466
ADRO 0.0003 0.0210 0.0008 0.0042
AKRA 0.0004 0.1146 0.0005 0.0014
ANTM 0.0000 0.0391 0.0007 -0.0055
ASII 0.0002 -0.0391 0.0004 0.0023
BBCA 0.0008 0.0252 0.0002 0.0238
BBNI 0.0005 0.0749 0.0004 0.0042
BBRI 0.0008 0.0949 0.0004 0.0058
BBTN 0.0005 0.0845 0.0005 0.0037
BMRI 0.0006 0.0570 0.0004 0.0065
BSDE 0.0004 0.0593 0.0006 0.0032
CPIN 0.0010 0.0020 0.0008 0.3819
CTRA 0.0007 0.0645 0.0008 0.0079
ERAA 0.0008 0.3275 0.0012 0.0019
EXCL 0.0002 0.0008 0.0008 -0.0768
GGRM 0.0000 0.5046 0.0004 -0.0002
ICBP 0.0005 0.4652 0.0005 0.0009
INCO 0.0003 0.5495 0.0009 0.0005
INDF 0.0003 0.5963 0.0003 0.0004
INKP 0.0013 0.5581 0.0010 0.0022
INTP 0.0001 0.8258 0.0005 0.0001
ITMG -0.0004 0.3980 0.0007 -0.0014
JPFA 0.0006 0.6384 0.0008 0.0009
JSMR 0.0002 0.4752 0.0003 0.0001
KLBF 0.0005 0.5754 0.0004 0.0007
MEDC 0.0006 0.4025 0.0010 0.0012
MNCN 0.0003 0.8311 0.0008 0.0004
PGAS 0.0000 0.6132 0.0007 -0.0001
PTBA 0.0001 0.4816 0.0007 -0.0001
PTPP 0.0008 0.8098 0.0007 0.0009
PWON 0.0007 0.7888 0.0007 0.0009
SMGR 0.0001 0.7712 0.0005 0.0001
SMRA 0.0004 0.9131 0.0007 0.0004
TBIG 0.0007 0.3296 0.0005 0.0015
TLKM 0.0005 0.5168 0.0003 0.0009
UNTR 0.0000 0.5866 0.0005 -0.0001
UNVR 0.0004 0.6422 0.0003 0.0006
WIKA 0.0100 0.6309 0.1848 0.0158

Asian Journal of Business and Accounting 15(2), 2022 99

 The next step is calculating Ci of the remaining 30 stocks using
Equation 18. Table 4 shows ERB, Ci and C* for the remaining 30
stocks that fulfil the criteria. C* is the highest value of individual Ci.
From the table, the value of C* is 0.00007, the value of Ci of PTPP,
PWON and TLKM. A stock will be selected into optimal portfolio
when its ERB value is higher than C*. Based on its ERB value, all 30
stocks meet that requirement, so those 30 stocks will be included in
optimal portfolio.

Table 4: ERB, Ci and ERB to C* of 30 stocks

Stocks ERB Ci ERB to C* Remarks

ACES 0.14662 0.00000 ERB > C* Portfolio candidate
ADRO 0.00423 0.00000 ERB > C* Portfolio candidate
AKRA 0.00136 0.00000 ERB > C* Portfolio candidate
BBCA 0.02383 0.00001 ERB > C* Portfolio candidate
BBNI 0.00419 0.00001 ERB > C* Portfolio candidate
BBRI 0.00582 0.00001 ERB > C* Portfolio candidate
BBTN 0.00368 0.00000 ERB > C* Portfolio candidate
BMRI 0.00646 0.00001 ERB > C* Portfolio candidate
BSDE 0.00320 0.00000 ERB > C* Portfolio candidate
CPIN 0.38185 0.00000 ERB > C* Portfolio candidate
CTRA 0.00787 0.00000 ERB > C* Portfolio candidate
ERAA 0.00194 0.00002 ERB > C* Portfolio candidate
ICBP 0.00091 0.00004 ERB > C* Portfolio candidate
INCO 0.00047 0.00001 ERB > C* Portfolio candidate
INDF 0.00035 0.00003 ERB > C* Portfolio candidate
INKP 0.00217 0.00006 ERB > C* Portfolio candidate
INTP 0.00011 0.00001 ERB > C* Portfolio candidate
JPFA 0.00088 0.00004 ERB > C* Portfolio candidate
JSMR 0.00013 0.00001 ERB > C* Portfolio candidate
KLBF 0.00068 0.00005 ERB > C* Portfolio candidate
MEDC 0.00122 0.00002 ERB > C* Portfolio candidate
MNCN 0.00036 0.00003 ERB > C* Portfolio candidate
PTPP 0.00093 0.00007 ERB > C* Portfolio candidate
PWON 0.00087 0.00007 ERB > C* Portfolio candidate
SMGR 0.00007 0.00001 ERB > C* Portfolio candidate
SMRA 0.00043 0.00004 ERB > C* Portfolio candidate
TBIG 0.00153 0.00003 ERB > C* Portfolio candidate
TLKM 0.00086 0.00007 ERB > C* Portfolio candidate
UNVR 0.00056 0.00006 ERB > C* Portfolio candidate
WIKA 0.01582 0.00000 ERB > C* Portfolio candidate
Maximum value (C*) 0.00007

100 Asian Journal of Business and Accounting 15(2), 2022

 Sharpe’s single-index model not only identifies stocks that
construct the portfolio, but also recommend the proportion of
funds to be invested for each stock to reduce unsystematic risk and
construct a highly diversified portfolio. Therefore, after finding stocks
in an optimal portfolio, the weight of those stocks can be found by
calculating the z-value by using Equation 20. The proportion of
z-value of each individual stock compared to total z-value is the
weight of the stock in the portfolio. Table 5 presents z-value and
weight of each stock in the portfolio. Five stocks with the highest
proportion in the single-index model portfolio are BBCA (12.860%),
TLKM (6.641%), BBRI (6.590%), INKP (5.488%) and ACES (5.325%).

Table 5: Z-value and weight of 30 individual stocks in portfolio

Stocks z-value Weight (%)

ACES 1.142 5.325
ADRO 0.107 0.501
AKRA 0.276 1.287
BBCA 2.758 12.860
BBNI 0.807 3.763
BBRI 1.413 6.590
BBTN 0.563 2.624
BMRI 0.966 4.504
BSDE 0.313 1.458
CPIN 0.899 4.190
CTRA 0.600 2.797
ERAA 0.526 2.455
ICBP 0.856 3.993
INCO 0.244 1.138
INDF 0.505 2.357
INKP 1.177 5.488
INTP 0.066 0.308
JPFA 0.615 2.866
JSMR 0.073 0.342
KLBF 0.965 4.501
MEDC 0.472 2.200
MNCN 0.298 1.391
PTPP 1.000 4.662
PWON 0.943 4.399
SMGR 0.001 0.003
SMRA 0.448 2.091
TBIG 0.951 4.437
TLKM 1.424 6.641
UNVR 0.983 4.582
WIKA 0.054 0.251
Total 21.445 100.00

Asian Journal of Business and Accounting 15(2), 2022 101

 The last step is calculating expected return and standard
deviation of portfolio that consists of 30 stocks whose weight has
been determined. Before that, α, β and unsystematic risk for the
portfolio have to be calculated. The α of the portfolio is the weighted
average of individual stock α, and so too for β and unsystematic risk.
Table 6 provides us the α, β and unsystematic risk of the portfolio.

Table 6: Alpha, beta and unsystematic risk of portfolio

α portfolio β portfolio σ (e) portfolio

The expected return of the portfolio is calculated using the single-index model regression
The expected return of0.000709
the portfolio is calculated
0.339544 using the 0.001
single-index model regression
equation.
The expected return of the portfolio is calculated using the single-index model regression
equation.
equation. The expected return of the portfolio is calculated using the single-
index model regression equation.
E(Rp ) = a p + b p E(RM )
E(R ) = a p + b p E(RM )
= 0.000709p+ 0.3339544
E
= 0.000709 a p + b p E´(´R(0.031%)
(Rp )+=0.3339544 )
(0.031%)
M
E ( R
= 0.000709
) = 0.00081 = 0.081%
E( Rp ) += 0.3339544
p ´ (0.031%)
0.00081 = 0.081%
E( Rp ) = 0.00081 = 0.081%
The variance of the portfolio is calculated using Equation 23.
The variance of the portfolio is calculated using Equation 23.
The variance
The variance of the
of the portfolio portfoliousing
is calculated is calculated using Equation 23.
Equation 23.
s P2 2= b p 2s2 M2 2 + s 2 2(e p )
s P = 2b p s M + s 2(e p )
= 0.339544
s P2 = b2´ p 2(0.009%)
s M2 + s 2 (e2+p )0.001
= 0.3395442 ´ (0.009%) + 0.001
s P22´= (0.009%)
= 0.339544 0.0010 2 + 0.001
s P = 0.0010
s P2 = 0.0010
The standard deviation is the square root of variance. Therefore, the standard deviation of
The standard
The standard deviation isdeviation is the
the square root square root
of variance. of variance.
Therefore, Therefore,
the standard deviation of
the portfolio
Thethe is:
standard deviation
standard is the square
deviation of root
the of variance.
portfolio is: Therefore, the standard deviation of
the portfolio is:
the portfolio is:
0.0010 = 0.031623 = 3.162%
0.0010 = 0.031623 = 3.162%
4.4 Discussion 0.0010 = 0.031623 = 3.162%
The mean-variance and single-index models result in portfolios that
4.4 Discussion
are optimal in their own way. The mean-variance model is more
4.4 Discussion
The complicated
mean-variance and than the single-index
single-index models result model. Thisthat
in portfolios section presents
are optimal theown
in their
4.4 Discussion
The mean-variance
performance and single-index
comparison models
betweenresult the
in portfolios that are optimal
mean-variance in their own
and single-
way. The mean-variance
The mean-variance model is more
and in
single-index complicated
models than
result in the single-index
portfolios that are model. This
optimal section
way. The index model
mean-variance the IDX.
model Both
is more models
complicated are
thancompared
the single-indexusing the in
model.
their
Sharpe own
This section
presents the
ratio.performance
The comparison
previous section between
shows the mean-variance
that the expected and single-index
return
way. The mean-variance model is more complicated than the single-index model. This section of model
portfolio in
presents the performance comparison between the mean-variance and single-index model in
the IDX.using
presents the
the single-index
Bothperformance
models comparedmodel
are comparison using is 0.081%,
the
between Sharpe withThe
ratio.
the mean-variance
a standard
previous deviation
section
and single-index shows ofthat
model in
the IDX.3.162%.
Both models are compared
Referring to Table using the Sharpe ratio. portfolio
2, mean-variance The previous section
7 has the shows
same that
the expected
the IDX.expected return of portfolio
Both models are compared using the
using the single-index model
Sharpe ratio. is 0.081%,
The previous with a
section In standard
shows that
the expected returnreturn of 0.081%
of portfolio usingwith a standard
the single-index deviation of 9.545%.
model is 0.081%, with a this
standard
deviation
the expectedof 3.162%. Referring
return of portfolioto using
Table the
2, mean-variance portfolio
single-index model 7 has thewith
is 0.081%, sameaexpected
standard
deviation of 3.162%. Referring to Table 2, mean-variance portfolio 7 has the same expected
return of 0.081%
deviation of 3.162%.withReferring
a standardtodeviation of 9.545%. In this
Table 2, mean-variance case, 7with
portfolio hasthethe same
same expected
expected
return of1020.081%
with aAsian
standard deviation
Journal of 9.545%.
of Business In this case,
and Accounting 15(2),with
2022the same expected
return, the standard deviation of the single-index model portfolio is
return of 0.081% with a standard deviation of 9.545%. In this case, with the same expectedlower than the mean-
return, the standard deviation of the single-index model portfolio is lower than the mean-
variance model
return, the portfolio.
standard Tableof7 presents
deviation a comparison
the single-index modelofportfolio
the five ishighest
lowerstocks,
than thestandard
mean-
case, with the same expected return, the standard deviation of the
single-index model portfolio is lower than the mean-variance model
portfolio. Table 7 presents a comparison of the five highest stocks,
standard deviation and Sharpe ratio between the two models with
the same expected return, 0.081%.
BBCA stock has the highest proportion in both models, with more
than 20%. TLKM and ACES are in the top five stocks for both models.
The Sharpe ratio calculated using Equation 24 with a 0.025% risk-free
rate shows that the single-index model portfolio with a 1.77% ratio
dominates the mean-variance model with 0.59%.
The study shows that the single-index model dominates the
mean-variance model in creating the optimal portfolio. This finding
is consistent with the findings of Ozkan and Cakar (2020), who state
that the single-index model performs better in developing markets.
This may happen because developing markets are more volatile than
developed markets (Darby et al., 2019), and because the single-index
model considers all aspects of the economy that avoid portfolio losses
(Chanifah et al., 2020).

Table 7: Comparison between single-index model and mean-variance model

Component Single-index Model Mean-variance Model

Five highest stock composition BBCA (12.860%) BBCA (27.56%)
TLKM (6.641%) TBIG (9.49%)
BBRI (6.590%) TLKM (8.93%)
INKP (5.488%) UNVR (7.76%)
ACES (5.325%) ACES (7.24%)
Expected return 0.081% 0.081%
Standard deviation 3.162% 9.545%
Sharpe ratio 1.77% 0.59%

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.

Asian Journal of Business and Accounting 15(2), 2022 103

 With the same expected return, 0.081%, the single-index model
provides a lower standard deviation of 3.162% compared to the
mean-variance model, with 9.545%. This means the single-index
model has a higher Sharpe ratio as a performance evaluation than
the mean-variance model. The optimal portfolio using the single-
index model consists of 30 stocks. Those stocks, sorted by proportion
in portfolio, are BBCA, TLKM, BBRI, INKP, ACES, PTPP, UNVR,
BMRI, KLBF, TBIG, PWON, CPIN, ICBP, BBNI, JPFA, CTRA, BBTN,
ERAA, INDF, MEDC, SMRA, BSDE, MNCN, AKRA, INCO, ADRO,
JSMR, INTP, WIKA and SMGR. BBCA is the stock with the highest
proportion in the portfolio for both single-index and mean-variance
models.
Based on our findings, several implications can be stated. The
findings could help investors arrange optimal portfolios that benefit
them. They can allocate their money to invest in several combination
of stocks that maximise returns with a certain risk. The findings also
expand previous portfolio literature with respect to the application
of the single-index and mean-variance models in the Indonesian
stock market. However, this study does not differ between risk
preferences—i.e., risk averse, risk moderate and risk taker—and
further research should be carried out to analyse optimal portfolios
for each risk preference. Furthermore, further research can use other
indices besides LQ45 and compare the results of daily, weekly or
monthly closing data.

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.

104 Asian Journal of Business and Accounting 15(2), 2022

Bikhchandani, S., & Sharma, S. (2000). Herd behavior in financial
markets. IMF Staff Papers, 47(3), 279-310. http://www.hozir.org/
pars_docs/refs/252/251684/251684.pdf
Bodie, Z., Kane, A., & Marcus, A. J. (2014). Investments (10th edition).
New York: McGraw Hill Education.
Bradford, J., & Miller, T., Jr. (2009). A Brief History of Risk and Return.
Fundamentals of Investments (5th edition). New York: McGraw Hill.
Chanifah, S., Hamdani, & Gunawan, A. (2020). The comparison of
applying single index model and capital asset pricing model by
means achieving optimal portfolio. Agregat: Jurnal Ekonomi dan
Bisnis, 4(1). https://doi.org/10.22236/agregat_vol4/is1pp8-24
Chen, L., & Pan, H. (2013). A dynamic portfolio theory model
based on minimum semi-absolute deviations criterion with an
application in the Chinese stock markets. China Finance Review
International, 3(3), 284-300. https://doi.org/10.1108/CFRI-05-
2012-0052
Chitnis, A. (2010). Performance evaluation of two optimal portfolios
by Sharpe’s ratio. Global Journal of Finance and Management, 2(1),
35-46.
Curtis, G. (2004). Modern portfolio theory and behavioral finance.
The Journal of Wealth Management, 7(2), 16-22. https://doi.
org/10.3905/jwm.2004.434562
Darby, J., Zhang, H., & Zhang, J. (2019). Institutional trading in
volatile markets: Evidence from Chinese stock markets. Pacific-
Basin Finance Journal, 65, 101484. https://doi.org/10.1016/j.
pacfin.2020.101484
Din, S. M. U., Mehmood, S. K., Shahzad, A., Ahmad, I., Davidyants,
A., & Abu-Rumman, A. (2021). The impact of behavioral
biases on herding behavior of investors in Islamic financial
products. Frontiers in Psychology, 11. https://doi.org/10.3389/
fpsyg.2020.600570
Elton, E. J., Gruber, M. J., & Padberg, M. W. (1976). Simple criteria
for optimal portfolio selection. Journal of Finance, 33(1), 296-302.
Fabozzi, F., Gupta, F., & Markowitz, H. (2002). The legacy of modern
portfolio theory. Journal of Investing, 11(3), 7-22. https://doi.
org/10.3905/joi.2002.319510
Fama, E. F., & French, K. R. (2004). The capital asset pricing model:
Theory and evidence. Journal of Economic Prespectives, 18, 25-46.

Asian Journal of Business and Accounting 15(2), 2022 105

Francis, J. C., & Kim, D. (2013). Modern Portfolio Theory: Foundations,
Analysis and New Developments. Hoboken: John Wiley & Sons.
Frankfurter, G. M., Phillips, H. E., & Seagle, J. P. (1976). Performance
of the Sharpe portfolio selection model: A comparison. Journal
of Financial and Quantitative Analysis, 11(2), 195-204. https://doi.
org/10.2307/2979049
Fransiska, M., Sumani, Willy, & Pangestu, S. (2018). Herding behavior
in Indonesian investors. International Research Journal of Business
Studies, 11(2), 129-143. https://doi.org/10.21632/irjbs.11.2.129-143
Garcíaa, F., Bueno, J. A. G., & Oliver, J. (2015). Mean-variance
investment strategy applied in emerging financial markets:
Evidence from the Colombian stock market. Intellectual Economics
9, 22–29. http://doi.org/10.1016/j.intele.2015.09.003
Gatfaoui, H. (2009). Sharpe ratios and their fundamental components:
An empirical study. SSRN. https://ssrn.com/abstract=1486073
Hadiyoso, A., Firdaus, M., & Sasongko, H. (2015). Building an
optimal portfolio on Indonesia Sharia Stock Index (ISSI).
International Journal of Administrative Science and Organization,
22(2), 111-121. https://doi.org/10.20476/jbb.v22i2.5699
Halicki, M., & Uphaus, A. (2014). The efficient frontier and
international portfolio diversification. Zeszyty Naukowe
Uniwersytetu Szczecińskiego nr 803, Finanse, Rynki Finansowe,
Ubezpieczenia” (66), 101-110.
Hamka, Halimah, H. S., Jupri, M., Budiono, R., & Tambi, A. M. (2020).
The influence of financial literacy on interest in investing for the
academic community of Akademi Keuangan & Bisnis Indonesia
Internasional (AKBII), Bandung, Indonesia. International Journal of
Business, Economics and Social Development, 1(1), 1-12. https://doi.
org/10.46336/ijbesd.v1i1.13
Indonesia Central Securities Depository (KSEI) (2020). KSEI
awarded best central custodian and sub registry during its 24th
anniversary. https://www.ksei.co.id/files/uploads/press_
releases/press_file/en-us/158_ksei_awarded_as_best_custodian_
in_southeast_asia_for_the_second_time_20190410114029.pdf
Jayeola, D., Ismail, Z., & Sufahani, S. (2017). Effects of diversification
of assets in optimizing risk of portfolio. Malaysian Journal of
Fundamental and Applied Sciences, 13(4). https://doi.org/10.11113/
mjfas.v0n0.567

106 Asian Journal of Business and Accounting 15(2), 2022

Jones, C. K. (2017). Modern portfolio theory, digital portfolio theory
and intertemporal portfolio choice. American Journal of Industrial
and Business Management, 7(7), 833-854. https://doi.org/10.4236/
ajibm.2017.77059
Kumar, B. R., & Fernandez, M. (2019). Examination of index
model and Prediction of beta: A case study examination in IT
sector. Accounting and Finance Research, 8, 226-231. https://doi.
org/10.5430/afr.v8n2p226
Kumar, N., Singari, R. M., & Archana, S. (2014). Portfolio
optimization: Indifference curve approach. International Journal of
Advance Research and Innovation, 2(1), 127-133.
Lan, Q., Xiong, Q., He, L., & Ma, C. (2018). Individual investment
decision behaviors based on demographic characteristics: Case
from China. Plos One 13(8): e0201916. https://doi.org/10.1371/
journal.pone.0201916
Lee, Y., Kim, W. C., & Kim, J. H. (2020). Achieving portfolio
diversification for individuals with low financial sustainability.
Sustainability 12(17), 7073. https://doi.org/10.3390/su12177073
Lekovic, M. (2018). Investment diversification as a strategy for
reducing investment risk. Economic Horizons, 20(2), 169-184.
https://doi.org/10.5937/ekonhor1802173L
Letho, L., Chelwa, G., & Alhassan, A. L. (2022). Cryptocurrencies and
portfolio diversification in an emerging market. China Finance
Review International, 12(1), 20-50. https://doi.org/10.1108/CFRI-
06-2021-0123
Mahmud, I. (2019). Optimal portfolio construction: Application
of Sharpe's single-index model on Dhaka Stock Exchange.
Jurnal Ilmiah Bidang Akuntasi dan Manajemen, 16(1). http://doi.
org/10.31106/jema.v16i1.1736
Malini, H. (2019). Efficient market hypothesis and market anomalies
of LQ 45 Index in Indonesia Stock Exchange. Sriwijaya International
Journal and Dynamic Economics and Business, 3(2), 107-121. https://
doi.org/10.29259/sijdeb.v3i2.107-121
Mangram, M. E. (2013). A simplified perspective of the Markowitz
portfolio theory. Global Journal of Business Research, 7, 59-70.
Markowitz, H. (1952). Portfolio selection. Journal of Finance, 7(1), 77-91.
Masry, M., & Menshawy, H. E. (2017). The impact of unsystematic
risk on stock returns in an emerging capital market (ECM)

Asian Journal of Business and Accounting 15(2), 2022 107

 country: An empirical study. International Journal of Financial
Research, 9(1), 189-202. https://doi.org/10.5430/ijfr.v9n1p189
Nyokangi, C. O. (2016). Relative Performance of the Single Index Versus
Mean Variance Optimization in Equity Portfolio Construction in Kenya
[Unpublished Master’s dissertation]. Strathmore University,
Nairobi, Kenya.
Omisore, I., Yusuf, M., & Christoper, N. (2012). The modern portfolio
theory as an investment decision tool. Journal of Accounting and
Taxation, 4(2), March 2012, 19-28. https://doi.org/10.5897/
JAT11.036
Otoritas Jasa Keuangan (2019). Booklet Survei Nasional Literasi dan
Inklusi Keuangan Tahun 2022. https://www.ojk.go.id/id/berita-
dan-kegiatan/info-terkini/Pages/Booklet-Survei-Nasional-
Literasi-dan-Inklusi-Keuangan-Tahun-2022.aspx
Ozkan, O., & Cakar, R. (2020). Comparison of mean-variance and
single index optimization methods in developed and developing
markets. Journal of the Faculty of Economics and Administrative
Sciences, 10(1), 63-79. https://doi.org/10.18074/ckuiibfd.441098
Partono, T., Widiyanto, Yulianto, A., & Vidayanto, H. (2017).
The analysis of optimal portfolio forming with single index
model on Indonesian most trusted companies. International
Research Journal of Finance and Economics, 163. http://www.
internationalresearchjournaloffinanceandeconomics.com/ISSUES/
IRJFE_163_04.pdf
Putra, I. K. A. A. S., & Dana, I. M. (2020). Study of optimal portfolio
performance comparison: Single index model and Markowitz
model on LQ45 stocks in Indonesia stock exchange. American
Journal of Humanities and Social Sciences Research (AJHSSR), 4(12),
237-244. https://www.ajhssr.com/wp-content/uploads/2020/12/
ZE20412237244.pdf
Rachmat, D., & Nugroho, A. B. (2013). Portfolio determination and
Markowitz efficient frontier in five Indonesian industrial sectors.
Journal of Business and Management, 2(1), 116-131. https://journal.
sbm.itb.ac.id/index.php/jbm/article/viewFile/550/419
Reilly, F. K., & Brown, K. C. (2003). Investment Analysis and Portofolio
Management (7th edition). Hoboken: Wiley.

108 Asian Journal of Business and Accounting 15(2), 2022

Rigamonti, A. (2020). Mean-variance optimization is a good choice,
but for other reasons than you might think. Risks, 8(1). https://
doi.org/10.3390/risks8010029
Rodríguez, Y. E., Gómez, J. M., & Contreras, J. (2021). Diversified
behavioral portfolio as an alternative to modern portfolio theory.
The North American Journal of Economics and Finance, 58, 101508.
https://doi.org/10.1016/j.najef.2021.101508
Rui, K. X., Rasiah, D., Yen, Y. Y., Ramasamy, S., & Pillay, S. D.
(2018). An analysis of the relationship between risk and expected
return in Malaysia stock market: Test of the CAPM. International
Journal of Engineering & Technology, 3(21), 161-170. https://doi.
org/10.14419/ijet.v7i3.21.17154
Sharpe, W. F. (1963). A simplified model for portfolio analysis.
Management Science, 9(2), 277-293.
Singh, S., & Gautam, J. (2014). The single index model and the
construction of optimal portfolio: A case of banks listed on NSE
India. Risk Governance and Control: Financial Market and Institutions,
4(2), 110-115. https://doi.org/10.22495/rgcv4i2c1art3
Wagdi, O., & Tarek, Y. (2019). The impact of financial risk on
systematic risks: International evidence. Journal of Applied Finance
& Banking, 9(6), 203-216.
Yuwono, T., & Ramdhani, D. (2017). Comparison analysis of
portfolio using Markowitz model and single index model: Case
in Jakarta Islamic Index. Jounral of Multidisciplinary Academics, 1(1).
https://www.kemalapublisher.com/index.php/JoMA/article/
view/278/265
Zakamulin, V. (2008). Portfolio performance evaluation with
generalized Sharpe ratios: Beyond the mean and variance. Journal
of Banking & Finance 33(7), 1242-1254. https://doi.org/10.1016/j.
jbankfin.2009.01.005

Asian Journal of Business and Accounting 15(2), 2022 109