Content uploaded by Bader Alsubaiei
Author content
All content in this area was uploaded by Bader Alsubaiei on May 29, 2021
Content may be subject to copyright.
10
The Scientific Journal of King Faisal University
Humanities and Management Sciences
00966544949552, balsubaiei@kfu.edu.sa
Corresponding Author: Badr Jawid Alsubaiei
Do Stock Market Risk Factors Explain Mutual Fund Returns? Evidence from Saudi
Arabia
Bader Jawid Alsubaiei
Department of Finance, Business School, King Faisal University, Al Ahsa, Saudi A rabia
ASSIGNED TO AN ISSUE
01/03/2022
PUBLISHED ONLINE
27/05/2021
ACCEPTED
27/05/2021
RECEIVED
24/03/2021
LINK
https://doi.org/10.37575/h/mng/210044
ISSUE
1
VOLUME
23
YEAR
2022
NO. OF PAGES
7
NO. OF WORDS
7726
ABSTRACT
This paper examines the risk factors of the Saudi Arabian equity market using an extensive data set. The study demonstrates which risk factors explain mutual
fund returns in the largest mutual fund market in the Middle East, a fast-growing economy and a major player in the oil market. This paper also assesses the global
and emerging market risk factors. This study analyzes 256 equity funds that operated in Saudi Arabia from January 2006 to July 2017. Time series regression
models (e.g., the CAPM, the Fa ma and French three-factor model and the Carhart four-factor model) are used. In addition, modified vers ions of the asset pricing
models were applied by adding stock market volatility and oil market volatility. The results indicate that the single-factor model, representing the market portfolio,
captures most of the mutual funds’ excess returns. Size, value and momentum factors do not enhance the explanatory power of mutual fund returns significantly.
The emerging market risk factors capture a small portion of the return variations where most effects were explained by the market risk factor. In explaining these
results, we emphasize the important implications for investors, academics and regulators to better understand the risk factor s that drive fund returns in a fast-
growing emerging market.
KEYWORDS
Equity funds, market risk, market volatility, oil volatility, performance, asset pricing
CITATION
Alsubaiei, B.J. (2022). Do stock market risk factors explain mutual fund returns? Evidence from Saudi Arabia.
The Scientific Journal of King Faisal University: Humanities and
Management Sciences
, 23(1), 10–6. DOI: 10.37575/h/mng/210044
1. Introduction
The mutual fund industry has attracted the interest of academics and
financial market participants due to the rapid growth of total assets
under management over recent decades. The finance literature has
provided a large body of work on mutual fund performance from
different aspects (e.g., Cuthbertson
et al
., 2016; Wang
et al
., 2018;
Wulfmeyer, 2016). The existing literature provides a sufficient
analysis of developed markets, whereas emerging markets have
received relatively little attention despite their accelerated growth,
dynamic change and rapid influence on the world economy. Market
risk is one of the major determinants of asset returns. Thus, this paper
aims to identify what risk factors capture fund returns in Saudi Arabia
by applying multiple asset pricing models: the capital asset pricing
model (CAPM), the Fama and French three-factor model and the
Carhart four-factor model. The motivation of this study is driven by
the significant role that the mutual fund sector plays in the Saudi
Arabian financial market due to the management skills provided by
professionals. Therefore, providing evidence on what risk factors
should be included to estimate abnormal returns in Saudi Arabia
enables asset management companies and investors to better
evaluate the return performance.
We investigate the mutual fund industry in Saudi Arabia for several
reasons. First, Saudi Arabia is an influential market in the world
economy as the largest oil exporter worldwide, which highlights its
economic role in supplying oil to the leading global economies.
Second, its financial market is the largest in the GCC (Gulf
Cooperation Council) region and the Middle East. Third, this work is
going to be the first extensive empirical work that provides evidence
for the risk factors that explain the return movements of the mutual
fund sector in Saudi Arabia. Seminal studies have explored mutual
fund performance in terms of both the time-series and cross-sectional
dimensions (e.g., Carhart, 1997; Fama and French, 1993; Grinblatt
and Titman, 1992; Jensen, 1968; Sharpe, 1966). The capital asset
pricing model, three-factor model and four-factor model have been
tested widely using global samples. We provide evidence on whether
these models capture the common return movements. Furthermore,
we examine whether the stock market volatility or oil market
volatility increases the explanatory power of the time-series
regressions. Finally, we have applied the global and emerging market
factors as proxies to represent the Saudi Arabian stock market risk
factors.
This paper contributes to the literature in three main ways. First, our
sample includes all existing funds in the investigated period, which
provides an extensive research background of the mutual fund
market in Saudi Arabia. To the best of our knowledge, previous
studies that investigate the asset pricing model in the context of Saudi
Arabian mutual fund return do not include all operating funds in the
market. Second, Cheng
et al
. (2010) indicate that the Saudi Arabian
stock market is primarily segmented from international markets and
can be affected by regional and global factors, so this paper
contributes to the finance literature by providing evidence of an
emerging market. Finally, the findings shed light on the behavior of
mutual fund return movements and whether risk returns that explain
the movements of return variations in developed markets are
applicable in emerging markets.
The approach that is used to test asset pricing models is the average
returns following Fama and French (1993). We use the time-series
regressions of the monthly excess return of funds on the market
portfolio and mimicking portfolios for size, value and momentum.
Time-series regressions provide clear evidence of the sensitivity to
risk factors such as the R-squared values to indicate whether risk
factors explain common variation in equity fund returns. The findings
suggest that the single-factor model (market portfolio) captures most
of the return variations, estimated by value-weighted average return.
Second, the results of applying the three-factor and four-factor
models are in line with the previous model, which indicates that the
market index explains most of the common variations of funds’
excess returns.
11
Alsubaiei, B.J. (2022). Do stock market risk factors explain mutual fund r eturns? Evidence from Saudi Arabia.
The Scientific Journal of King Faisal University: Humanities and Management Sciences
, 23(1), 10–6. DOI:
10.37575/h/mng/210044
Third, we include the stock market volatility of the local stock market
in the time-series regressions to determine whether it enhances the
explanatory power of the model. The findings provide further
evidence that local stock market return captures about 35% of the
mutual funds’ excess return variations by itself and adds
approximately 4% if it is included with the market portfolio. Finally,
the oil market volatility is included in our tests due to the vital role the
oil market has in the Saudi Arabian economy. The results suggest that
oil market return has a negative impact on the fund performance,
which is in line with the literature that suggests a negative
relationship between the stock market and the oil market (e.g.,
Alsubaiei
et al
., 2020; Diaz
et al
., 2016; Kang
et al
., 2015). However,
the oil volatility index does not capture a significant amount of the
funds’ return variations. As a result, the findings in this paper suggest
that the market portfolio and stock market volatility capture
approximately 90% of the mutual funds’ excess returns in Saudi
Arabia, indicating that these two risk factors should be included to
estimate the funds’ abnormal return.
Our findings have important implications for (i) investors to improve
their understanding of the risk factors that should be considered
when evaluating their funds’ return performance, which would make
them more informed about funds’ management performance to
better allocate their assets (e.g., individuals would have more
information on what drives the mutual funds’ returns and whether
the result is produced by fund management or the market), and (ii)
academics and financial market authorities to enhance their
understanding of the behavior of funds’ returns in a major emerging
market and to provide new evidence on the extant risk-adjusted
performance model (the four-factor model) that has proved its
significance in developed markets.
The remainder of this paper is organized as follows. Section 2 briefly
discusses the related literature. Section 3 presents the inputs of the
regressions. Section 4 describes the data and model specifications.
Section 5 discusses the main findings. Section 6 examines the
robustness checks. Section 7 provides concluding statements.
2. Literature Review
There is rapid growth in the finance literature to evaluate mutual fund
performance. However, a clear agreement is yet to emerge on funds’
abnormal returns because researchers are still trying to develop the
asset pricing model (Fama and French, 2015). The seminal work of
Jensen (1968) investigated the relationship between the returns of
mutual funds with similar risk by applying the single-factor model,
which includes the market excess return. Then, Grinblatt and Titman
(1989) examine funds’ abnormal returns by applying Jensen’s single-
index measure with four sets of benchmarks, which shows the
significant role the market portfolio plays in explaining mutual fund
returns. Fama and French (1993) developed the three-factor model,
which adds the size and value risk factors to the return evaluation.
Recent studies have examined the mutual funds’ performance by
applying the three-factor model. With a sample including five
European mutual fund markets, Otten and Bams (2002) determine
the explanatory power of the risk factors as ranging from 76% to 97%.
The Carhart (1997) four-factor model has been applied widely in the
finance literature to evaluate mutual funds’ abnormal returns, which
adds the momentum anomaly to the assets pricing model. Ferreira
et
al
. (2012) examine the determinants of the mutual fund performance
of 26 countries using the Carhart four-factor model, and the results
show that the Carhart model captures 74% of the return for the
Taiwanese market (minimum) and 94% of the return for the Thai
market (maximum), while the total explanatory power is 87%.
Emerging markets are different from developed markets in terms of
market influencers and whether or not they are segmented or
integrated with the world economy, and such markets have received
less scholarly attention than developed markets. Białkowski and
Otten (2011) examine the performance of the Polish mutual fund
sector as an emerging market by applying the four-factor model. Their
results suggest that the Carhart model explains approximately 92%
of the domestic equity and 68% of international equity, and the
sample is free of survivorship bias. Huij and Post (2011) examine the
performance of emerging equity mutual funds in the US and use the
single-factor and four-factor models. Their results suggest that the
single-factor model captures 88% to 97% of the variations, whereas
the four-factor model explains between 90% and 97% of the return
movements. This indicates that applying the Carhart model does not
increase the explanatory power significantly.
The recent trend in the literature focuses on the importance of
volatility as a measure of risk due to its role in affecting investors’
behavior and the market direction. Busse (1999) suggests that mutual
fund performance is associated with market volatility, and a recent
paper by Wang
et al
. (2018) provides evidence of investors’ reaction
to the volatility level of the market. Jordan and Riley (2015)
investigate the relationship between funds’ performance with market
volatility. They applied the four-factor model, and their findings
suggest that portfolio volatility is a predicator of fund abnormal
return. The explanatory power of the risk factors varies from 75% to
96%, and low (high) market volatility is associated with a positive
(negative) abnormal return. Hu
et al
. (2014) study the effect of
diversification on returns and find that well-diversified funds are less
affected in high market volatility conditions.
The existing work that has covered the Saudi Arabian stock market is
very limited. For example, Salameh (2020) investigates the
application of asset pricing models in the Saudi exchange where his
data includes only 44 observations over less than three years. The
findings suggest that the Fama and French model is the best model to
be applied. Aldaarmi
et al
. (2015) apply the multiple asset pricing
model to the Saudi Arabian stock market. Their empirical analysis
covers only 60 monthly periods, and they find that the three-factor
model has the best explanatory power in explaining return variations.
As a result, this paper investigates the application of different asset
pricing models, including other risk factors, to ascertain the best
model that can explain the mutual fund return variations.
3. Inputs to Time-Series Regressions
(Factors Definition)
3.1. Fund Return and Risk Portfolios:
First, we test the single-factor model following the CAPM developed
by Sharpe (1964). The capital asset pricing model indicates whether
the market portfolio captures the common excess return variation.
The excess return is developed based on the value-weighted average
of funds. Jensen (1968) applies the CAPM to measure portfolio
performance and estimates funds’ abnormal returns:
Where is the value-weighted fund returns, and where the return is
winsorized at the top and bottom 1% to avoid extreme negative
return in some periods. is the risk-free rate of return that is the US
three-month T-bill return; the risk-free rate that is applied is the US T-
bill due to the integration between the US and Saudi Arabia
economies. According to the US Treasury Department, Saudi Arabia
was the 10th largest foreign holder of US government bonds by the
middle of 2018. Furthermore, the exchange rate between the local
currency and US dollars has been fixed since 1981, demonstrating the
12
Alsubaiei, B.J. (2022). Do stock market risk factors explain mutual fund r eturns? Evidence from Saudi Arabia.
The Scientific Journal of King Faisal University: Humanities and Management Sciences
, 23(1), 10–6. DOI:
10.37575/h/mng/210044
significant relationship between these markets (Aleisa and Dibooĝlu,
2002). is the return on market index, the Saudi Tadawul All Share
Index (TASI).
There is theoretical and empirical evidence that shows the single-
factor model ignores other important risk factors that can capture
common return variations, such as macroeconomic financial market
factors. Therefore, Fama and French (1993) extended the asset
pricing model by including two additional risk factors, which are size
(SMB) and book-to-market ratio (HML), in addition to the market
proxy. As a result, the risk-adjusted fund performance is calculated as
follows:
Where (eq.4) is the return of the equal-weighted average on the
three small stock portfolios minus the average of the returns on the
three big stock portfolios;
HML
(eq.5) is the equal-weighted average
of the returns for the two top market value portfolios minus the
average of the returns for the two low market value portfolios.
Based on market value, firms that have a market value above the
median are classified as big. Also, based on book-to-market value,
firms within the top 30th percentile are classified as high, firms within
the middle 30th percentile are classified as medium and firms within
the bottom 30th percentile are classified as low. To construct the
Fama risk factors (SMB and HML), six value-weighted portfolios are
developed: (i) SL, which is the return of small firms in terms of market
value and low firms in terms of book-to-market value, (ii) SM , which
is the return of small firms in terms of market value and medium firms
in terms of book-to-market value, (iii) SH, which is the return of small
firms in terms of market value and high firms in terms of book-to-
market value, (iv) BL, which is the return of big firms in terms of
market value and low firms in terms of book-to-market value, (v) BM,
which is the return of big firms in terms of market value and medium
firms in terms of book-to-market value and (vi) BH, which is the
return of big firms in terms of market value and high firms in terms of
book-to-market value.
Then, Carhart (1997) adds the momentum anomaly (MOM) (eq.6) in
addition to the three factors to enhance the pricing error and to
capture cross-sectional variation in the returns. Carhart’s four-factor
model has become one of the most applied models in the mutual
fund literature to estimate mutual fund abnormal returns (Ferreira
et
al
., 2012; Otten and Bams, 2002). As a result, the fund performance
calculation is the following:
Where MOM (eq.6) is the average return on the portfolios of the
highest 30% return minus the average return on the portfolios with
the lowest 30% return.
The MOM portfolio represents the momentum by estimating stock
market firms’ top winners and losers. Firms within the top 30th
percentile are classified as winners and firms within the bottom 30th
percentile are classified as losers. We develop four value-weighted
return portfolios: (i) SW, which is the return of small firms in terms of
market value and firms that perform well, (ii) BW, which is the return
of big firms in terms of market value and firms that perform well, (iii)
SL, which is the return of small firms in terms of market value and
firms that perform poorly and (iv) BL, which is the return of big firms
in terms of market value and firms that perform poorly.
1
According to the S&P Dow Jones Indices website, “The index comprises the principal physical
commodities that are traded in active, liquid futures markets. In addition to numerous
Table 1 reports the summary statistics for the Saudi Arabian market,
global market and emerging market risk factor. Table 1.A indicates
that there are low correlations between the Saudi Arabian market
factors where the average return of the market index is negative.
Table 1.B and Table 1.C present the summary statistics of the global
and emerging four-factor model that is provided from Kenneth R.
French’s website. The global portfolio includes the developed
markets of 23 countries, whereas the emerging portfolio represents
27 developing countries, including Saudi Arabia. The correlations
between the market return and other factors are still low, with the
highest correlation between market excess return and the size factor
being 45%.
Table 1 also shows summary statistics for stock market volatility and
oil market volatility. All provided data are monthly bases for the
period from the beginning of 2006 to July 2017. The Saudi Arabian
stock market (TSAI) has volatility of approximately 5%. The realized
volatility has a negative mean of 9%. Figure 1 illustrates the
fluctuations in the local stock market return and the movements of
stock market volatility. Figure 2 exhibits the line chart of the crude oil
index return and the fluctuations of oil market volatility.
Table 1: Summary statistics of the four-factor model, J anuary 2006 to July 2017
Variable
Obs
Mean
Std. Dev.
Min
Max
SARMRF
139
-0.016
0.085
-0.3
0.177
SASMB
139
-0.01
0.081
-0.455
0.277
SAHML
139
-0.023
0.05
-0.274
0.087
SAMOM
139
0.199
0.087
0.098
0.725
GRMRF
139
0.535
0.045
-0.195
0.115
GSMB
139
-0.015
0.015
-0.035
0.039
GHML
139
-0.001
0.017
-0.046
0.048
GMOM
139
0.278
0.036
-0.244
0.092
ERMRF
139
0.72
6.377
-27.29
17.98
ESMB
139
0.05
1.717
-6.94
4.21
EHML
139
0.397
1.572
-3.06
4.36
EMOM
139
0.584
2.953
-14.92
5.43
Stock Market Volatility
139
0.0598
0.0458
0.0125
0.2612
Oil Market Volatility
139
0.094
0.0457
0.0307
0.297
Mean is the average of our sample, Std. De v. is the sample standard deviation and Min and Max are minimum and m aximum
values, respectively. RMRF is the stock market index return m inus the T-bill return. SMB and HML are the Fama and French
factors (size and book-to-market). MOM is a factor representing return momentum. When prefixed by ‘SA’ these factors
represent the Saudi Arabian market, prefixed by ‘G’ these factors represent the global market, when prefixed by ‘E’ these factors
represent the emerging m arkets. Stock market information is based on the Saudi Arabian market index (TASI), and oil prices
information is from the S&P GSCI crude oil excess return.
3.2. Volatility Measures:
This paper includes the volatility of the stock market and oil market
to investigate whether they explain the common return variation in
the Saudi Arabian mutual funds. We compute the uncertainty based
on the realized volatility (following, for example, Christiansen
et al
.,
2012; French
et al
., 1987; Paye, 2012; Schwert, 1989). Realized
volatility provides estimations closer to normality due to the inclusion
of the total squared daily return to approximate the standard
deviation of the equity or commodity benchmark for the frequency in
the study. The first volatility estimation is based on the local stock
market (i.e., TASI) to explore whether market risk captures some of
the excess return movements because French
et al
. (1987) find a
significant relationship between equity market returns and market
volatility. The second volatility estimation is the oil market because
the literature indicates a significant relationship between the oil
market and the equity market (e.g., Diaz
et al
., 2016; Kang
et al
.,
2015). We use the S&P Dow Jones Index because it provides a reliable
benchmark for crude oil market performance over time.
1
The S&P
GSCI crude oil index excess return is applied to calculate oil market
volatility, which represents a portfolio of crude oil futures contracts’
return by using the world production weighted basis as its weighting
method and return on daily contracts as its calculation base (
S & P
GSCI Crude Oil
, 2017). As a result, we estimate the volatility variable
is as follows:
related and sub-indices calculated on a single-component and multi-currency basis, thematic
baskets such as biofuel and petroleum are available.”
13
Alsubaiei, B.J. (2022). Do stock market risk factors explain mutual fund r eturns? Evidence from Saudi Arabia.
The Scientific Journal of King Faisal University: Humanities and Management Sciences
, 23(1), 10–6. DOI:
10.37575/h/mng/210044
Where R is the daily continuously compounded return in the month t
for the stock market and crude oil market, and represents the
number of trading days in the month.
Figure 1: Saudi Arabian stock market return and volatility by month - Jan. 2006 to July 2017
Stock Market Return
Stock Market Volatility
Figure 2: Oil market return and volatility by month - January 2006 to July 2017
Oil Market Return
Oil Market Volatility
4. Overview of Mutual Fund Data, Model
Specifications and Hypotheses
4.1. Data Description:
This paper focuses on open-end equity mutual funds in the Saudi
Arabian market to determine the optimal asset pricing model. Mutual
fund data are extracted from the Lipper for Investment Management
database. Our sample spans all equity funds that are available from
the beginning of 2006 until mid-2017. There are 256 equity funds
that are operated in the Saudi Arabian mutual fund industry, of which
175 are currently active. There are 121 funds that invest only in the
Saudi Arabian stock market, which counts for approximately 50% of
the total sample. We apply monthly data in this paper following the
literature (e.g., Barber
et al
., 2016; Ferreira
et al
., 2012). Monthly data
are suitable for our analysis due to the following reasons: (i) the
required data for the Saudi Arabian market are poor for less than a
monthly basis and (ii) monthly data can capture higher mutual fund
return movements because it mitigates any bid-ask effect biases in
the daily data (Arouri and Nguyen, 2010). Finally, our sample is free
of survivorship bias because we include all available equity funds that
have existed in our sample period.
Table 2 presents the summary statistics for the mutual fund sample
included in this paper. On average, mutual fund returns experience a
slightly positive monthly return close to zero (0.07%), which is a raw
return before adjusting for the risk-free rate. The risk-adjusted return
(excess return) is −0.7% per month. The fund size has a net asset
value of 0.1%, where the total asset value at the end of our sample is
US$33 million. Moreover, the average fund age is about nine years,
and the oldest fund has been operating for more than 25 years.
Table 2: Summary statistics of mutual fund performance and other variabl es, January 2006 to July
2017
Variable
Obs.
Mean
Std. Dev.
Min
Max
Raw return
19,900
0.0007
0.078
-2.354
0.580
Risk-adj. return
19,900
-0.007
0.062
-0.277
0.143
Size
18,413
-0.012
0.257
-6.806
13.935
Age
256
8.676
6.477
0
25
Obs. is the number of observations for the study period, Mean is the aver age of our sample, Std. Dev. is the sample standard
deviation and Min and Max are minimum and maximum values, respectively. Raw return is the return of fund i in period t
before adjusting for the risk-free rate and risk factors, Risk-adj. Return is the excess return, Size is the log of the total fund asset
of fund i in period t, and Age is the total years since the fund launched.
4.2. Model Specification:
We apply the time-series regressions approach because the main
finding in the regressions is the R-squared, which indicates the
explanatory power of the risk factors included. Times-series models
vary based on the risk factors used in the regressions. This paper tries
to provide evidence on whether mimicking portfolios for risk factors
related to the equity market capture the variations of the mutual fund
returns. The calculations are as follows:
Where is the value-weighted return of fund in month in excess
of the risk-free rate, is the risk-free rate of the return, is the
return on the market index (TASI), is the size risk factor (eq.4),
HML
is the book-to-market ratio risk factor (eq.5),
MOM
is the
momentum anomaly risk factor (eq.6), is the stock market
volatility at a month (eq.7) and is the oil market volatility at a
month (eq.7).
4.3. Hypotheses:
The capital asset pricing model indicates whether the market
portfolio captures the common excess return variations by itself. We
hypothesize the following:
• H1: The market return has significant explanatory power in explaining
the mutual fund return variations.
The four-factor model includes size, book-to-market ratio and
momentum as risk factors to explain return variations. We
hypothesize the following:
• H2: The four-factor model has significant explanatory power in
explaining the mutual fund return variations.
Stock market volatility and oil market volatility have a significant
relationship with equity market returns, as suggested in the existing
literature. Therefore, we anticipate that these risk factors can improve
the explanatory power of the mutual fund return variations. We
hypothesize the following:
• H3: The stock market volatility and oil market volati lity have
significant explanatory power in exp laining the mutual fund return
variations.
5. Empirical Findings
5.1. Local Risk Factors:
This section provides the initial empirical analysis of the factors that
explain mutual fund returns in Saudi Arabia. This paper uses a time-
series regression approach following Fama and French (1993) for
equity funds. Table 3 presents the time-series regression results for
the four risk factors on equity funds’ excess returns in Saudi Arabia
(eq.8). The first model is based upon the market proxy (TASI)
following the seminal work of Jensen (1968), which applies the
single-factor model to evaluate portfolios’ abnormal returns. The
regression results reveal that market proxy captures about 88% of the
variation of funds’ excess returns by itself. In specifications 2 and 3 of
Table 3, we include the size effect (ME) and value effect (BE/ME) as
proposed by Fama and French (1993) to enhance the explanatory
ratio on the mutual fund returns to generate an accurate estimation
of abnormal returns. Surprisingly, the three-factor model does not
better explain the variation of the fund returns than the single-factor
model because the R-squared does not increase significantly when
14
Alsubaiei, B.J. (2022). Do stock market risk factors explain mutual fund r eturns? Evidence from Saudi Arabia.
The Scientific Journal of King Faisal University: Humanities and Management Sciences
, 23(1), 10–6. DOI:
10.37575/h/mng/210044
we add the Fama and French factors.
Table 3: Times-series regressions of funds’ excess returns on thre e-factor model
VARIABLES
(1)
(2)
(3)
(4)
(5)
(6)
SARMRF
0.727***
0.729***
0.708***
(0.034)
(0.040)
(0.038)
SASMB
0.070
-0.035
-0.041
(0.121)
(0.040)
(0.038)
SAHML
0.398**
0.014
-0.027
(0.153)
(0.053)
(0.056)
SAMOM
-0.351***
-0.066**
(0.075)
(0.032)
Constant
0.001
-0.011*
-0.002
0.001
0.059***
0.012**
(0.002)
(0.006)
(0.006)
(0.002)
(0.012)
(0.005)
Observations
139
139
139
139
139
139
R-squared
0.879
0.007
0.091
0.881
0.216
0.886
This table reports time-series regressions of mutual fund value-weighted excess returns on the Fama and French three risk
factors. Detailed definitions of variables are in Table 1. Constant is the intercept of the model. O bservations are the number of
observations in each model. R-squared is the coefficient of determination. Robust standard errors in parentheses. *** p<0.01,
** p<0.05, * p<0.1.
Models 4–6 provide the results by adding another risk factor, which
is the momentum anomaly. The Carhart (1997) four-factor model
improves the pricing error, which has become the most applied
model to calculate mutual fund performance (Ferreira
et al
., 2012;
Otten and Bams, 2002). Overall, the findings in model 1 suggest that
fund return is significantly related to the local market momentum,
which explains about 21% by itself. However, when we run the time-
series regressions of all risk factors (Carhart four-factor model), the
overall explained variation is approximately 89%, which means there
is a minimal impact compared to the market portfolio (88%).
Recent literature has provided evidence on the importance of
volatility as a measure of risk and how the return movement can be
directed by market uncertainty (e.g., Alkhaldi, 2015; Kang
et al
., 2015;
Liu, 2014; Nazlioglu
et al
., 2015). Therefore, we add the market price
volatility and oil market volatility in our time-series regressions to
investigate whether these uncertainty indexes explain some of the
mutual fund return variation in Saudi Arabia.
Table 4 includes the results of whether the volatility affects the equity
funds’ excess returns. Model 1 includes the market volatility and
shows that it captures a significant portion of the return variations by
itself. The R-squared is about 35%, which means that it is a key risk
factor that should be considered when the mutual fund abnormal
return is estimated. The findings indicate the strong negative
relationship with market volatility, which confirms the conclusions
of existing studies (e.g., French
et al
., 1987; Hammoudeh
et al
., 2009).
We run a robustness check of the relationship between market
volatility and value-weighted excess returns by applying the GARCH
model, and the result is in line with the time-series findings. Models 2
and 3 include the four-factor model and market volatility, and the
results show that market volatility has a significant impact on mutual
fund return. The model explanatory power increased to 90%.
Table 4 : Times-series regressions of funds’ excess returns on stock m arket volatility
VARIABLES
(1)
(2)
(3)
(4)
(5)
(6)
SARMRF
0.668***
0.669***
0.705***
0.669***
(0.034)
(0.034)
(0.038)
(0.035)
SASMB
-0.017
-0.016
-0.031
-0.016
(0.030)
(0.032)
(0.041)
(0.034)
SAHML
-0.018
-0.014
-0.022
-0.014
(0.046)
(0.050)
(0.056)
(0.050)
SAMOM
0.008
-0.053
0.008
(0.033)
(0.033)
(0.033)
Stock Market Volatility
-0.063***
-0.018***
-0.018***
-0.018***
(0.010)
(0.004)
(0.004)
(0.005)
Oil Market Volatility
-0.318*
-0.071
-0.000
(0.177)
(0.060)
(0.063)
Constant
-0.201***
-0.055***
-0.058***
0.019
0.017**
-0.058**
(0.033)
(0.014)
(0.017)
(0.016)
(0.007)
(0.023)
Observations
139
139
139
139
139
139
R-squared
0.349
0.901
0.901
0.048
0.888
0.901
This table reports time-series regressions of mutual fund value-weighted excess returns on the Fama and French three risk
factors. Detailed definitions of variables are in Table 1. Stock Market Volatility is the realized volatility which is the sum of
trading days’ squared return in a month on TASI (Tadawal All Share Index). Oil Market Volatility is the realized volatility which
is the sum of trading days’ squared return in a month on the S&P GSCI crude oil excess return. Constant is the intercept of th e
model. Observations are the number of observations in each model. R-squared is the coefficient of determination. Robust
standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1.
Model 4 indicates whether the oil market volatility explains some of
the return variations. Oil market volatility is included due to the
important role Saudi Arabia plays in the oil sector as well as the
dependence of the local market on oil. The findings show that oil
market volatility, which is estimated by realized volatility, explains
approximately 5% of the mutual fund market variations. This result
can be attributed to the fact that stock market volatility is directly
related to the oil market, and the movement is already reflected in the
local stock market; this finding is in line with existing work , which
finds the spill-over from the oil market to the stock market in Saudi
Arabia and in other markets. However, model 5 shows that oil market
volatility has a smaller contribution in capturing the variation in
equity fund returns, which suggests that the inclusion of oil market
volatility when estimating mutual fund abnormal returns does not
provide a significant influence.
To provide better estimates, we ran further tests that included other
risk factors. Approximately 50% of the mutual funds under
investigation invest in international markets. As a result, we include
the global risk factors in our regressions to determine whether some
of the return variation can be explained. Table 5 presents multiple
regressions for local risk factors and global risk factors on equity
funds raw returns in Saudi Arabia. Model 1 shows that the market
portfolio captures a very small part of the equity market variations
where the R-squared is about 4%. The findings show that adding the
global market portfolio to the regression does not increase the
percentage of explained variations. Model 5 includes four risk factors
for the local market and global market, and the findings indicate that
there is no added benefit from the global risk factors because the R-
square (89%) is similar to Table 4 (without global risk factors).
Table 5: Times-series regressions of funds’ excess returns on global r isk factors
VARIABLES
(1)
(2)
(3)
(4)
(5)
GRMRF
0.267*
0.259**
0.253*
0.229*
-0.012
(0.137)
(0.131)
(0.133)
(0.136)
(0.059)
GSMB
0.261
0.272
0.264
0.155
(0.389)
(0.383)
(0.386)
(0.123)
GHML
0.075
-0.004
0.026
(0.281)
(0.303)
(0.131)
GMOM
-0.109
-0.046
(0.155)
(0.073)
SARMRF
0.706***
(0.037)
SASMB
-0.042
(0.039)
SAHML
-0.029
(0.055)
SAMOM
-0.067**
(0.033)
Constant
-0.013**
-0.013**
-0.013**
-0.012**
0.013**
(0.006)
(0.006)
(0.006)
(0.006)
(0.005)
Observations
139
139
139
139
139
R-squared
0.034
0.037
0.037
0.040
0.888
This table reports time-series regressions of mutual fund value-weighted excess returns on the Fama and French three risk
factors. Global factors are obtained from the Kenneth R. French website. Constant is the intercept of the model. Observations
are the number of observations in each model. R-squared is the coefficient of determination. Robust standard errors in
parentheses. *** p<0.01, ** p<0.05, * p<0.1.
Finally, our results suggest that the market portfolio proxy has the
largest power to explain the mutual fund returns in Saudi Arabia. This
finding is in line with Fama and French (1993), who find that market
return explains between 80% and 90% of the stock market variations.
Stock market volatility also has a significant role in capturing the
return variation. Combining both risk factors (market portfolio and
market volatility) explains approximately 90% of the mutual fund
excess returns (market risk-adjusted return). Therefore, these two risk
factors should be included to estimate fund abnormal returns (alpha)
to identify funds that outperform their benchmarks.
5.2 Emerging Market Risk Factors:
Fama and French developed a proxy to represent the risk factors of
emerging markets that allows practitioners in those countries to have
ready estimations for their local risk markets. Therefore, time-series
regressions are run by applying the emerging markets risk factors to
represent the Saudi Arabian stock market. The reason for applying
these regressions is to diagnose whether our main findings persist
and to address whether emerging market risk factors capture most of
the equity mutual funds’ returns variations in Saudi Arabia. If the risk
factors do explain a large share of the return movement, we provide
evidence to apply them in future studies rather than estimate the risk
factors for the local market individually.
Model 1 in Table 6 tests the single-factor model, which is the market
15
Alsubaiei, B.J. (2022). Do stock market risk factors explain mutual fund r eturns? Evidence from Saudi Arabia.
The Scientific Journal of King Faisal University: Humanities and Management Sciences
, 23(1), 10–6. DOI:
10.37575/h/mng/210044
portfolio on the mutual funds’ excess returns. The regression results
reveal that the market portfolio of Saudi Arabian stocks explains
about 25% of the variations, which is significantly less than the local
estimations of the market return, which captures about 88%.
Specifications 2 and 3 include the size effect and value effect,
respectively, as proposed by Fama and French (1993), whereas model
4 includes the momentum effect as proposed by Carhart (1997). The
results show that estimating the abnormal return from the four-factor
model does not provide a larger explanation of the value-weighted
fund returns. As a result, the single-factor model captures most of the
market return movements when we use the emerging risk factor,
which confirms our previous findings. Model 5 shows evidence that
stock market volatility does explain a significant part of the return
movements. The direction of the relationship is negative, as
suggested in the literature.
Table 6: Times-series regressions of funds’ excess returns on emer ging risk factors
VARIABLES
(1)
(2)
(3)
(4)
(5)
(6)
ERMRF
0.005***
0.005***
0.006***
0.006***
0.003***
0.005***
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)
ESMB
0.002
0.002
0.002
-0.002
0.002
(0.003)
(0.003)
(0.003)
(0.003)
(0.004)
EHML
-0.004
-0.005
-0.001
-0.005
(0.004)
(0.004)
(0.003)
(0.005)
EMOM
-0.002
-0.004**
-0.002
(0.002)
(0.002)
(0.002)
Stock Market Volatility
-0.058***
(0.010)
Constant
-0.015***
-0.015***
-0.014***
-0.013***
-0.187***
-0.013**
(0.005)
(0.005)
(0.005)
(0.005)
(0.031)
(0.006)
Observations
139
139
139
139
139
139
R-squared
0.252
0.253
0.261
0.265
0.518
0.168
This table reports time-series regressions of mutual fund value-weighted excess returns on the Fama and French three risk
factors. Detailed definitions of variables are in Table 1. Emerging factors are obtained from Kenneth R. French’s website. St ock
Market Volatility is the realized volatility which is the sum of trading days’ squared return in a month on TASI (Tadawal All Share
Index). Constant is the intercept of the model. Observations are the number of observations in each model. R-squared is the
coefficient of determination. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1.
To validate the results, we reran a times-series regression after
removing all the funds that do not invest only in the Saudi Arabian
market. Specification 6 confirms the previous findings on whether the
emerging markets’ risk factors serve as proxies, which leads us to the
following conclusions. First, funds that invest in other markets
besides the Saudi Arabian market have a minimal share of the mutual
fund industry in Saudi Arabia, and we find similar results when the
international funds are excluded. Second, the emerging market risk
factors capture a small size of the return variations compared to the
local market risk factors, which suggests estimating the index of the
investigated local market. Finally, the only factor that shows a major
relation with the funds’ excess returns is the market portfolio, which
confirms our main findings.
6. Robustness Checks
This section conducts several robustness checks of our main
analysis.
2
We first exclude funds that do not invest strictly in the Saudi
Arabian stock market, which removes roughly 50% of the main
sample. The regression results indicate that the market portfolio
captures approximately 90% of the variation of funds’ excess returns
by itself, which confirms our previous findings where the single-
factor model explains most mutual fund return variations even after
including other risk factors.
Moreover, we use excess returns on five portfolios formed by size as
our dependent variables in the time-series regressions. The funds are
ranked in five size quintiles based on their size in millions. Then, we
estimate the value-weighted excess monthly return for funds at the
same level of size from January 2006 to July 2017. The findings
demonstrate that market proxy is the key factor that has a persistent,
significant relation with all funds’ portfolios. Second, it provides
evidence of a strong positive relationship, which indicates the
dominant impact of the stock market on mutual fund returns. Finally,
the explanatory power of market factors increases with the large
funds as it indicates that R-squared increases with larger portfolios
(87%).
7. Conclusion
The main purpose of this study is to identify the risk factors that
capture the common return variations in the Saudi Arabian mutual
funds market. Our study contributes to the existing literature in
different aspects. First, we provide evidence on the asset pricing
models in a major emerging market where the result can be applied
in similar markets. Second, the findings indicate a significant
explanatory power of the local market volatility, which suggests
adding the volatility of the stock market to the model. Third, the most
important risk factor that captures the highest percentage of the
return variations is the market portfolio, which is the local market
excess return. This key result is in line with the finance literature,
which shows the important role of the market return in explaining the
return movements. Finally, the emerging market risk factors can be
used as a proxy to represent the Saudi Arabian market because they
explain a significant amount of the returns. All tests are robust to
different model specifications and generate consistent outcomes.
A notable implication of our results is that it can be used in any work
that has expected returns, such as evaluating abnormal performance
(skills) and selecting portfolios. Consequently, asset managers should
take advantage by applying the best model to estimate performance
and anticipate the fluctuations caused by risk factors. Asset
management firms and mutual fund companies should also include
stock market volatility in their estimations to measure mutual fund
performance. Finally, investors and financial market regulators will
be able to judge managers’ skills by estimating the performance of
managed portfolios to know whether they can beat the market or
generate abnormal returns greater than passive funds.
Biography
Bader Jawid Alsubaiei
Department of Finance, Business School, King Faisal University, Al Ahsa, Saudi
Arabia, 00966544949552,
balsubaiei@kfu.edu.sa
Dr. Alsubaiei is an assistant professor, the CEO of the Endowment and
Investment Fund at King Faisal University and the Director of
Investment and Recourses Development. He obtained his PhD in
finance from Loughborough University in the UK and his master’s
degree with distinction from Brandeis University in the US. He
worked as a part-time lecturer in finance at Loughborough University.
His research interests focus on financial markets, asset pricing and
investment portfolios. He has published many papers in leading
journals and conferences (e.g., the European Journal of Finance).
References
Aldaarmi, A., Abbod, M. and Salam eh, H. )2015(. Implement Fama and
French and c apital asset pricing models in Saudi Arabia s tock
market.
The Journal of Applied Business Research
, 31(3), 953–68.
Aleisa, E.A. and Dibooĝlu, S. (2002). Sources of real exchange rate
movements in Saudi Arabia.
Journal of Economics and Finance
,
26(1), 101–10.
Alkhaldi, B.A. (2015). The Saudi ca pital market: The crash of 2006 and
lessons to be learned.
International Journal of Business, Economics
and Law
, 8(4), 135–46.
Alsubaiei, B.J., Ca lice, G. and Vivian, A. (2020). How does mutual fund flow
respond to o il market volatility?
European Journal of Finance
,
n/a(n/a) 1–25.
16
Alsubaiei, B.J. (2022). Do stock market risk factors explain mutual fund r eturns? Evidence from Saudi Arabia.
The Scientific Journal of King Faisal University: Humanities and Management Sciences
, 23(1), 10–6. DOI:
10.37575/h/mng/210044
Arouri, M. E. H. and Nguyen, D. K. (2010). Oil prices, stock m arkets and
portfolio investment: Evidence from sector analysis in Europe over
the last decade.
Energy Policy
, 38(8), 4528–39.
Barber, B. M., Huang, X. and Odean, T. (2016). Which fa ctors m atter to
investors? Evidence from mutual fund flows.
Review of Financial
Studies
, 29(10), 2600–42.
Białkowski, J. and Otten, R. (2011). Emerging market mutual fund
performance: Evidence for Pola nd.
The North American Journal of
Economics and Finance
, 22(2),118–130.
Busse, J. A. (1999). Volatility timing in mutual funds: Evidence from daily
returns.
The Review of Financial Studies
, 12(5), 1009–1041.
Carhart, M. M. (1997). On persistence in mutual fund performanc e.
The
Journal of Finance
, 52(1), 57–82.
Cheng, A. R., Jahan-Parvar, M.R. and Rothma n, P. (2010). An empirical
investigation of stock ma rket behavior in the Middle East and
North Africa.
Journal of Empirical Finance
, 17(3), 413–27.
Christiansen, C., Schmeling, M . and Sc hrimpf, A. (2012). A comprehensive
look at financial v olatility prediction by economic variables.
Journal of Applied Econometrics
, 6(27), 956–77.
Cuthbertson, K., Nitzsche, D. and O’Sullivan, N. (2016). A re view of
behavioural and management effects in mutual fund performance.
International Review of Financial Analysis
, 44(n/a), 162–76.
Diaz, E.M., Molero, J.C. and Perez de Gracia, F. (2016). Oil price volatility and
stock returns in the G7 economies.
Energy Economics
, 54(n/a),
417–30.
Fama, E.F . and French, K.R. (1993). Common risk factors in the returns on
stocks and bonds.
Journal of Financial Economics
, 33(1), 3–56.
Fama, E.F. and French, K.R. )2015 ( . A five-factor asset pricing model.
Journal
of Financial Economics
, 116(1), 1–22.
Ferreira, M. A., Keswani, A., Miguel, A.F . and Ramos, S.B. (2012). The
determinants of mutual fund performance: A cross-country study.
Review of Finance
, 17(2), 1–43.
French, K.R., Schwert, G.W. and Stambaugh, R.F. (1987). Expected stock
returns and volatility.
Journal of Financial Economics
, 19(1), 3–29.
Grinblatt, M. and Titman, S. (1989). Mutual fund performance: An analysis
of quarterly portfolio holdings.
The Journal of Business
, 62(3),
393–416.
Grinblatt, M. and Titman, S. )1992(. The persistence of mutual fund
performance.
The Journal of Finance
, 47(5), 1977–84.
Hammoudeh, S.M., Yuan, Y. and McAleer, M. (2009). Shock and volatility
spillovers among equity s ectors of the Gulf Arab stock markets.
Quarterly Review of Economics and Finance
, 49(3), 829–42.
Hu, J.L., Chang, T.P. and Chou, R.Y. (2014). Market conditions and the effect
of diversification on mutual fund performance: Should funds be
more conc entrative under crisis?
Journal of Productivity Analysis
,
41(1), 141–51.
Huij, J. and Post, T. (2011). On the performance of emerging market equity
mutual funds.
Emerging Markets Review
, 12(3), 238–49.
Jensen, M. C. (1968). The performance of mutual funds in the period 1945–
1964.
Journal of Finance
, 23(2), 389–416.
Jordan, B. D. and Riley, T. B. (2015). Volatility and mutual fund manager skill.
Journal of Financial Economics
, 118(2), 289–98.
Kang, W., Ratti, R. A. and Yoon, K. H. (2015). The impact of oil price shocks
on the stock market return and volatility re lationship.
Journal of
International Financial Markets, Institutions and Money
, 34(n/a),
41–54.
Kilian, L. (2009). Not all oil price shocks are alike: Disentangling supply
shocks in the crude oil market.
American Economic Review
, 99(3),
1053–69.
Liu, L. (2014). Extreme downside risk spillover from the United States and
Japan to As ia-Pacific stock markets.
International Review of
Financial Analysis
, 33, 39–48.
Nazlioglu, S., Hammoudeh, S. and Gupta, R. (20 15). Volatility transmission
between Islamic and conventional equity markets: Evidence from
causality-in-variance test.
Applied Economics
, 47(46 ), 49 96–
5011.
Otten, R. and Bams, D. (2002). European mutual fund performance.
European Financial Management
, 8(1), 75–101.
Paye, B. S. (2012). “Déjà vol”: Predictive regressions for aggregate stock
market volatility using macroeconomic variables.
Journal of
Financial Economics
, 106(3), 527–46.
S & P Dow Jones Indices. (2017).
S & P GSCI Methodology Report
. Available
at: http://us.spindices.com/indices/commodities/sp-gsci-crude-
oil (accessed on 18/03/2021).
Schwert, W. G. (1989). Why does stock market volatility change over time ?
Journal of Finance
, 44(5), 1115–53.
Salameh, H. (2020). Application of ass et pricing models: Evidence from
Saudi Exchange.
Investment Management and Financial
Innovations
, 17(1), 348–68.
Sharpe, W. (1964). Capital asset prices: A theory of market equilibrium under
conditions of risk.
The Journal of Finance
, 19(3), 425–442.
Sharpe, W. F. (1966). Mutual fund performance.
The Journal of Business
,
39(1), 119–38.
Wang, Y., Watson, J. and Wickramanayake, J. (2018). The global financial
crisis and the m utual fund flow-performance relationship.
The
World Economy
, 41(11), 1–22.
Wulfmeyer, S. (2016). Irrational mutual fund managers: Explaining
differences in their behavior.
Journal of Behavioral Finance
, 17(2),
99–123.
