Evaluating the behavior of stock market returns has been a challenging and
complex task to investors and researchers, since there is a conflict over the
predictability of stock indices performance (Senol and Ozturan,
2008; Gunasekaran and Ramaswami, 2011). The willingness
of investors for determining the linkages between market risk and expected return
has been one of the most crucial factors in developing the Capital Asset Pricing
Model (CAPM). Sharpe (1964) and Lintner
(1965) are among the first who extended asset pricing theory by using the
CAPM. Since, then evidences show that the CAPM is generally applicable in major
stock markets (Wong and Tan, 1991). This study employs
CAPM owing to its power in examining the relationship between risk and return
of any risky asset. Estimating beta is very important since it is applicable
for pricing stocks, determining the cost of capitals and explaining returns
(Bruner et al., 2008; Hasan
et al., 2011). The condition for the Islamic stocks is different
from its conventional counterparts, due to the prohibition of charging interest
on deposit in Islamic transactions.
Islamic stocks are traded based on the Islamic principles rules (Shariah) which
prohibits dealing in interest (Riba). Moreover, sell and purchase of any firms
stocks which produce or deal liquor, pork, gambling, pornography and any other
unlawful things (Haram) are absolutely forbidden in accordance with the Islamic
transaction (Muamulat). Additionally, Islamic financial instrument follows the
Profit/Loss Share (PLS) contract while the conventional finance is based on
debt and return (Echchabi and Olaniyi, 2012). That is
why the hypothesis that Islamic stocks bear higher risk than the conventional
stocks is still remained.
Over the last decades, a huge number of investors look for environmental-social
issues and ethical funds in their financial investment. Socially Responsible
Investing (SRI), therefore, has become increasingly defined as a mean to promote
environmentally sustainable development (Ferruz et al.,
2007). According to the 2007 Report on socially responsible investing trend
size of these funds has grown from USD 1.18 trillion to USD 2.16 trillion between
1997 and 2000 alone. Since its climbing from USD 2.29 trillion in 2005, SRI
enjoyed a growth rate of 18% until 2007. Investment in the Islamic finance can
be recognized as a SRI because both consider the investors financial return
and the impact of investment to the society. Islamic finance is based on keeping
in view certain social goals intended for the benefit of society (Venardos,
The Kuala Lumpur Composite Index (KLCI) was introduced in 1986 as a capitalization-weighted
stock index. In addition, the Kuala Lumpur Shariah Index (KLSI) was launched
in 1999 that represents an average price of Islamic compliant stocks, while
the KLCI indicates the average price of 100 stocks comprising both Islamic and
non-Islamic compliant stocks (Albaity and Ahmad, 2008).
In 2006, Bursa Malaysia with the collaboration of FTSE group launched FTSE Bursa
Malaysia index. Two FTSE Islamic indices, FTSE Bursa Malaysia Hijrah Shariah
index (tradable) and FTSE Bursa Malaysia EMAS Shariah index (benchmark) are
launched in 2007 due to the growing interest in Shariah-compliant portfolio.
According to Securities Commission Malaysia, the updated list features a total
of 847 securities which are classified by the Shariah Advisory Council (SAC)
as Shariah-compliant securities.
||Percentage of Syariah-compliant securities in total securities,
Source: The Syariah advisory council of the securities commission
These counters constitute 88% of the listed securities on Bursa Malaysia in
2010.1 Figure 1 indicates the
upward trend of Islamic-based investment in Malaysia since 2000.
While Malaysia as one of the leading country in terms of Islamic financial market plays a vital role in enhancing the Islamic investments in the Muslim world, further studies and researches are certainly the requisites of this enhancement via variety of methods and samples. This study evaluated the performance of Islamic stocks using the CAPM to find any authenticity as reported by earlier researches. To encourage the investors using the CAPM in managing return-risk nexus is the main purpose of the study. Particularly, the study investigated the return and risk characteristics of Shariah-compliant indices of Malaysian stock market to find out whether or not the Islamic indices are as risky as the conventional portfolios. It also analyzed the impact of 2008 global financial crisis on the performance, risk-return relation and the volatility in return for the above-mentioned indices.
Recently, several studies have been conducted to find the risk-return relationship
using CAPM model (Rahman et al., 2006a, b;
Rhaiem et al., 2007; Chang
et al., 2011). Despite the growing of Islamic stocks, the literature
on Islamic market is still thin compared to the conventional stocks. Some studies
analyzed the performance of Islamic funds and compared the performance with
the conventional funds. The other group examined the Islamic stocks with some
Islamic indices as proxy and compared with the non-Islamic indices. Some of
these studies are reviewed as follows.
Ismail and Shakrani (2003) examined the relationship
between return and beta for Islamic unit trusts in Malaysia with both unconditional
CAPM and conditional CAPM2. They found a flat unconditional
and insignificant relationship between beta and risk premium, while with the
conditional CAPM analysis, there is a significant positive relationship between
beta and returns. Their results also indicated that investors in Islamic unit
trusts are risk averse because of their willingness to invest in the indices
that have a lower level of risk.
Abdullah et al. (2007) observed the differences
in terms of performance between Islamic and conventional mutual funds in Malaysia
through employing 51 conventional and 14 Islamic funds, respectively. Their
findings suggested that the Islamic funds have a better performance compared
to their conventional counterparts in the bear market, while the latter performs
much better than the former during bullish economic trend. They also concluded
both kinds of funds have a performance below than the market portfolio.
On contradict, Hayat and Kraeussl (2011) found Islamic
equity funds underperform both their Islamic and conventional benchmarks and
this underperformance has increased during the 2008 financial crisis. They concluded
that this underperformance might lead to increase some specific risks in the
Islamic equity funds, while the conventional investments typically are free
from these types of risks.
Hakim and Rashidian (2004) investigated the linkages
between risk and return on Dow Jones Islamic Market (DJIM) and Wilshire 5000
which represent Islamic index and non-Islamic index, respectively. They found
that DJIM has unsystematic characteristics reflected in the risk-return tradeoffs,
while the Wilshire 5000 is out of this characteristic.
Hussein (2004) utilized the CAPM to estimate if there
is a significant different on returns earned between the investors who purchase
equities in the FTSE Global Islamic index and their counterparts who invest
in the FTSE All World index. He also compared the performance of the mentioned
indices with a socially responsible index (FTSE4Good index). The sample period
is divided into two sub-periods, namely bull period and bear period. Hussein
(2004) suggested the performance of Islamic index is as good as its conventional
counterpart. Moreover, the Islamic index performs better during the economic
growth than when the economy suffers a bearish pattern. Hussein
(2004) also pointed out that the ethical investment outperforms the unscreened
portfolio in the bull market period.
Reviewing studies of the Islamic indices in the Malaysian stock market, Sadeghi
(2008) investigated the impact of the launch of Shariah-compliant index
on the financial performance and liquidities of securities involved in this
index via an event study methodology.3 The findings
of the authors study reveal that despite finding negative abnormal returns
from immediately prior and after the event day, the market has been a positive
reaction against the launch of the index. In line with Sadeghi
(2008) suggestion regarding to further investigations, we use the two newly-launched
indices, namely the FTSE Bursa Malaysia Hijrah Shariah Index (FBMHS) and FTSE
Bursa Malaysia EMAS Shariah Index (FBMSHA) as tradable index and a benchmark
Ahmad and Ibrahim (2002) compared the performance of
Kuala Lumpur Syariah Index (KLSI) and Kuala Lumpur Composite Index (KLCI) for
the period of 1999 to 2002. They found the risk-adjusted return of the KLSI
is lower than the KLCI; however the latter bears higher risk than the former.
Similarly, Albaity and Ahmad (2008) compared the performance
of the KLSI and the KLCI for the period of April 1999 to December 2005. They
provided evidence that the KLSI is marginally underperformed KLCI. It might
be due to the existence of positive linkages between size and return in developing
countries. Their result is consistent with Hakim and Rashidian
(2004) which stated that investor who choose Islamic securities experienced
as worse as those who invest in the non-Islamic securities. The concentration
of the two above-mentioned studies is on the linkages between the screened investment
(KLSI) and non-screened investment (KLCI). Nevertheless, the present study evaluated
the performance of the Islamic indices through involving the conventional indices
in the models and investigating risk-return relation.
Hashim (2008) documented that the FTSE Global Islamic
index bears higher risk than the market; however, the realized return is fair
and appropriate, also the risk of the Islamic index is less than the socially
responsible index (FTSE4Good). He confirmed this hypothesis that the risk of
the Islamic index is tolerable, although the level of risk achieved is appropriate
for this level of risk. In other words, he considered a pure risk-oriented approach
to evaluate the issue of risk opposition in the Islamic investment. However,
ours attempt was to detect the behavior of risk-adjusted return using CAPM.
As discussed above literatures, most authors and researchers have made the
comparison studies between the performances of the Islamic investments and non-Islamic
portfolios. None of them have addressed the presence of volatility clustering
in stock return of the Islamic indices. Meanwhile, these studies mainly have
been conducted in the periods when the stock markets are operating under normal
conditions, while investigation on the volatility and return of the asset markets
is probably more interesting when the asset markets are under stress (Abdulnasser
and Roca, 2005). The present study covered these absences by using the GARCH
test and re-estimating all models for the 2007 global financial crisis period
(Angabini and Wasiuzzaman, 2011).
MATERIALS AND METHODS
The sample data consists of three FTSE Bursa Malaysia indices which indicate
the market portfolio such as FTSE Bursa Malaysia KLCI Index (FBMKLCI), FTSE
Bursa Malaysia 100 Index (FBM100) and FTSE Bursa Malaysia EMAS Index4
(FBMEMAS) and two other indices which represent the Islamic portfolio such as
the FTSE Bursa Malaysia Hijrah Shariah Index (FBMHS) and FTSE Bursa Malaysia
EMAS Shariah Index (FBMSHA). Daily closing prices of these indices have been
collected. In addition, we obtained the daily three months Kuala Lumpur Inter-bank
Offer Rate (KLIBOR) representing the risk-free rate in the model. All data are
retrieved from DataStream.
Our entire sample period is from 1 March 2007 to 28 February 2011, it is restricted by the availability of data owing to the launching of both FBMHS and FBMSHA in 2007. Furthermore, in order to capture the impact of the 2008 global financial crisis on the behavior of the indices, a sub-period (from 1 March 2008 to 31 March 2009) has been determined which indicates as the crisis period.
The descriptive statistics for each FTSE index returns for entire sample period
and crisis period are shown in Table 1 and 2,
respectively. As shown in Table 1, the average daily returns
for all indices are positive. FBMHS has the highest mean return while FBMKLCI
shows the lowest mean return. The standard deviation of different indices return
shows that the dispersion of the Islamic indices is larger than the market portfolio.
We note that both Islamic indices have larger standard deviation than the market
indices consistent with the common argument that the Islamic stocks bear higher
risk than the conventional stocks. The skewness implies that all of the data
are negative skewed which indicates that they are non-symmetric.
||Descriptive statistics of FTSE Bursa Malaysia indices return
The large values of kurtosis states that the distribution of return for all
indices has thicker tails than the normal distribution.
On the other hand, Table 2 expresses that the average daily return for all sample indices are negative. Therefore, the 2008 global financial crisis has had a significant effect on both conventional and Islamic stock indices in Malaysian FTSE market. However, the values of standard deviation indicate that the Islamic indices still bear higher risk compared with their conventional counterparts.
Figure 2 represents the daily price movements of our sample indices. As it can be obviously seen, the indices have been moved together. Moreover, the indices have achieved the highest point in January 2008 and then they have experienced a downward trend due to the subprime and global financial crisis at the end of 2008 and start to rebound from the beginning of 2009 due to the stimulus packages by the government.
This study uses the Sharpe-Lintner CAPM version which is based on some variables such as expected return on stock as the dependent variable and beta of stock, risk-free rate and expected return of market as the independent variables:
where, E(Ri) is the expected return on stock i, Rf is
the risk-free rate, E(Rm) is the expected market return and βi
is the measure of risk and is defined as:
While (Rm-Rf) is defined as the market premium, we have
RS1 and RS2 which represent the return of FBMHS and FBMSHA,
respectively. On the other hand, RM1, RM2 and RM3
are the return of market indices for FBMKLCI, FBM100 and FBMEMAS, respectively.
||Descriptive statistics of FTSE Bursa Malaysia indices return
The adjusted return is used in this analysis instead of the closing index itself. This choice is the transformation of non-stationary to stationary data. The daily returns are calculated as the natural log differences in prices as follow:
where, Rt is the daily return in period t, Pt is the closing price at t and Pt-1 is the closing price at t-1. Then, the daily return is adjusted by risk free rate (rf) as below:
For each of the Islamic indices in our sample, an estimate of beta, can be obtained by running an Ordinary Least Square regression (OLS) using the following model:
where, rS1 and rS2 are the risk premium (excess return)
of FBMHS and FBMSHA, respectively and rM1, rM2 and rM3
are the risk premium of FBMKLCI, FBM100 and FBMEMAS,, respectively. In CAPM,
alpha (α) and beta (β) are two estimated parameters. Alpha is known
as the risk-adjusted return and is used to identify the ranking of indices performance.
|| Price movement of FTSE Bursa Malaysia indices
Alpha measures the extent by which the stock index beats the performance expected
from an index of its same level of risk. Beta as an independent unit is the
systematic risk or the risk level of an investment in relation to the risk level
of the market. In fact a firms beta reflects the sensitivity of the stocks
return to the markets overall return (Madura, 2008).
When beta is more than one, it will indicate a risky investment and more sensitive
to the movement in the market. However, if beta is less than one, it means a
low risk investment and less sensitive to the market movement and lastly when
beta is equal to one, there is an equally risky investment. As usual practice,
some diagnostic tests such as Durbin Watson (DW), Ramsey RESET and heteroskedasticity
white tests will be conducted after the estimation to determine the specification
of the model.
It is well known that the presence of volatility clustering in stock return. Hence, taking time varying volatility and leverage effect into account is important for better estimation. However, most of the earlier literature does not take this into consideration in their studies. To encounter this shortcoming, we employ the Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model for the estimation. This study utilizes the simplest GARCH model, namely GARCH (1, 1) model. The model considers the impact of previous conditional variance on the conditional variance equation as follows:
where, τ1 and τ2 are the estimated coefficient of the lagged value of the squared residual term (ARCH term) and the lagged value of the conditional variance (GARCH term), respectively.
RESULTS AND DISCUSSION
The results of OLS estimation of two Islamic indices against each of the market indices for the entire period have been shown in Table 3 and 4. As can be seen in Table 3 and 4, the alphas for all models are almost zero with positive sign except for the FBMSHA on FBM100 and FBMSHA on FBMEMAS which are negative signs. However, all alpha are not significantly different from zero.
Beta for all models are positive and significant infers that the Islamic stocks are moving at the same direction with the market. Moreover, the beta is less than one except those on FBMKLCI. The beta of less than one is significant for all the models. The beta of less than one inferring that the Islamic index is less risky and less sensitive to the movement of market. However, FBMHS on FBMKLCI bears higher systematic risk than other Islamic indices.
|| Statistical results of FBMHS based on the OLS and AR(1) Model
|Values are statistically significant at ***1, **5 and *10%
levels, Values in the parentheses are t-statistic, DW: Durbin Watson, AR:
Based on the DW statistics as displayed, we found serial correlation in the error term for all models. Results also show that the presence of heteroskedasticity and specification errors in the models. As the original models are not correctly specified with the serial correlation and heteroskedasticity problem, an AR(1) is added in the model and correct the standard error term by the White Heteroskedasticity-Consistent Standard Errors and Covariance. The estimation results for the AR(1) model are reported in Table 3 and 4. Nonetheless, no special difference is observed between AR(1) and OLS models in terms of alpha and beta.
The results of GARCH model are reported in Table 5. These coefficients are statistically significant. Moreover, these results show that the ARCH effect and the GARCH effect are both positive and their sum is between zero and one, as required by the theory of GARCH (1, 1). This suggests that the shocks to the volatility have low persistent effect.
|| Statistical results of FBMSHA based on the OLS and AR(1)
|Values are statistically significant at ***1, **5 and *10%
levels, Figures in the parentheses are t-statistics, AR: Autoregressive,
DW: Durbin Watson
|| Estimation Results of the Islamic Indices Based on GARCH
|Values are statistically significant at ***1, **5 and *10%
levels, Figures in the parentheses are t-statistics
In order to examine the impact of global financial crisis, we re-estimate all models for the crisis period. The results for the crisis period are shown in Table 3 and 4. In contrast with the results of the entire period, the sign of alphas in all models are negative. However, all the alphas are still not significantly different from zero. Moreover, in similar with the entire sample period the beta for all the models is positive and significant, while its value is more than one. The beta is less than one only for FBMSHA on FBM100 in both OLS and AR(1) models. Obviously, in the crisis period, the Islamic indices have become more risky to the market in comparison with the entire sample period. Meanwhile, R-squared statistics in the crisis period are higher than entire sample period confirming the above result. This result weakens the hypothesis that the Islamic indices are less risky during the crisis period.
The result probably implies that the Islamic index promises higher return in the crisis period. The FBMHS on FBMKLCI has the highest beta among the models in the crisis period. On the other hand, The FBMHS on FBMEMAS has the lowest beta among the models which means it experiences lower systematic risk than others in the entire period. In fact, the values of betas indicate that the Islamic indices bear higher systematic risk in the crisis period compared with the entire period.
Furthermore, as can be seen in Table 5 the result of crisis period indicates that some of the ARCH effect and GARCH effect are not positive. However, their sum is still between zero and one with the exception of FBMHS on both FBM100 and FBMEMAS which are negative. A negative sign on the squared error term in the previous time period (ARCH ) or on its conditional variance in the previous time period implies that bad news in the market will increase the volatility more than good news of an equal magnitude.
This study investigated the performance of the Islamic indices in the Malaysia
FTSE stock market. For this purpose, we employed the data of five FTSE Bursa
Malaysia indices in the study as the proxy for Islamic index and market portfolio.
We followed the traditional CAPM which has been developed by Sharpe
(1964) and Lintner (1965) to form six models. Ultimately,
we pointed out that the proposed model is appropriate and applicable in examining
the relationship between risk and return in the Malaysian Islamic stock market.
In the entire period the Islamic indices have earned excess return in comparison
with the expectation from an investment at the same level of risk, while the
excess return have been lost during the global financial crisis. The results
indicate that the Islamic indices have high risk-adjusted returns and low risks,
while after the crisis all the Malaysian Islamic indices have been converted
into the risky and sensitive indices. Hence, following Hashim
(2008) the result of this study approves the hypothesis that the risk of
the Islamic indices is tolerable and high.
Excluding some uncertainty elements from the Islamic investment (e.g., gambling)
is perhaps the cause of this lower degree of risk exposure (Abdullah
et al., 2007). Nevertheless, the low-volatility effect has been captured
using GARCH. In addition, in line with Ismail and Shakrani
(2003), this study states that in the entire sample period the Islamic securities
perform very well in terms of lowness in the risk level which is adorable for
Muslim investors. Despite of this low level of risk, investors yield higher
return compared with the crisis period. Therefore, Shariah-compliant securities
might attract and preoccupy the attention of risk-averse investors. This is
consistent with the Hayat and Kraeussl (2011).
Malaysia FTSE market has the potential to attract more Muslim and non-Muslim
investors. In such scenario, more companies which possess high working capital
(blue chip stocks) would be attracted to the securities.
1List of securities approved by the Syariah Advisory Council of
the Securities Commission, 31 May 2010.
2Conditional CAPM was proposed by Pettengill
et al. (1995). It considers the weak relationship between risk and
return as a potential explanation of the observed positive risk-average portfolio
3He defined the day which the Shariah-compliant index was launched
as the event
4FTSE Bursa Malaysia EMAS index represents all
the ordinary securities which are listed on the main board of the Bursa Malaysia
that qualified for the rules of eligibility, free floating as well as liquidity
(Lean and Tan, 2010, p. 5)