The Effect of Fund and Family Characteristics on Islamic Mutual Fund Flows: Evidence from Saudi Arabia

Islamic mutual funds (IMF) are growing as a substitutional investment vehicle for investors who want to combine value and financial objectives in their investment. A group of the funds is managed by one Investment Company called the family of funds. This study investigates the flow-performance relationship in IMF, in addition to the extent to which family and fund characteristics contribute to explaining fund flows in Saudi Arabia for the period from 2007 to 2017. The study uses raw returns to calculate the fund performance and use the percentage money flow (FLOW), defined as money flow scaled by the total net asset of the fund. The results show there are positive relationship between past performance and fund flow that mean IMF investors make rational financial decisions by directing fund flows to better performing funds. In addition, the results show family characteristics have an impact on IMF inflow. Ulum Islamiyyah Journal | Vol.30, April 2020 28


Introduction
Islamic finance has rapidly developed throughout the past decade and still continues to expand. Dubai Islamic Bank, (2016) reported that the total value of global Islamic financial services reached US$2.293 trillion at the end of 2016, compared to US$2.143 trillion in 2015. Moreover, the size of the industry potentially grows by at least 10 percent annually.
The report forecast that by 2020, the industry would grow between US$3.151 trillion and US$4.57 trillion. In addition, today Islamic finance comprises many types of financial services, such as funds management (mutual funds), Islamic banking, Islamic insurance (takaful), and Islamic bond (Sukuk) (Iqmal & Kamaruddin, 2019).
The mutual fund is an investment company that collects money from shareholders and invest it in assorted securities, including money market instruments, stocks, and bonds. Moreover, mutual funds offer attractive advantages such as the ability to invest in an equity fund without incurring transaction costs or cost of collecting information, thus mutual funds allow diversification and provide administration in dealing with investments to reduce the workload of individual investors (Ahmad et al. 2017). Mutual funds can influence the real economy via two channels, that is, the primary and secondary markets. Where the fund flows are positively correlated with subsequent economic growth, it leads to the fact that these flows incorporate additional information related to real economic activities that are not perceived by forecasters (Hoepner et al. 2013). Islamic funds are one of the most important types of mutual funds at the moment.
The purpose of Islamic mutual fund (IMF) is to achieve religious and ethical objectives without infringing the traditional needs of diversification, liquidity, and performance (El Khamlichi et al. 2014;Azmi et al. 2018). Moreover, as Sharia law prevents many high-risk activities, Islamic financial services have been much less affected by financial crises relative to their respective benchmarks (Hoepner et al. 2013;Makni et al. 2015;Boo et al. 2016).
Sharia law prohibits mutual funds from riba al nasiah, maysir, gharar and haram products or services and it requires haram purification, riba al nasiah represents the receipt of interest on capital. Hence, IMF cannot invest in conventional bonds, warrants, preferred stock, certificates of deposit and some derivatives. The IMF industry has recently seen a dramatic increase in the amount of wealth held by mutual fund managers. Malaysia World's slamic Finance Marketplace MIFC, (2017) stated that the total global Islamic assets under management (AUM) in 2017 were USD70.8 billion, and the number of Islamic funds was 1,535. This is a significant increase compared to 2008 when there were only 802 funds with total AUM of USD47 billion. The largest type of IMF is equity funds, representing 40 percent of total Islamic funds. They are followed by fixed income at 17 percent, and real estate and private equity at 12 percent.
The motivation for testing the flow-performance relationship of IMFs is that investors of these funds, besides the expected financial gains, have certain non-financial motives that inclining toward religion (Sharia-compliance). This proclivity effects the behavior of investors in the IMF's investment decision. In addition, the idea of investing in IMF is to prohibit speculative activities, and so Islamic mutual funds will have a lower risk compared to those that include speculative activities, and this feature may turn IMF into a hedging tool in the shadow of market declines periods (Merdad et al., 2010;Hoepner et al., 2013). Many studies suggest that investors select funds based on their past performance and characteristics, according to the feedback trading hypothesis, investors react to lagged returns, with positive (negative) returns leading to positive (negative) flows Azmi et al. 2018). Many previous studies examined the flows of IMF at the funds level (such as Renneboog et al. 2006;Marzuki & Worthington, 2017;Azmi et al. 2018). This study contributes by takes into considering fund family variables because the different families follow different strategies that distinguish them from each other. fund family may utilize strategies that depend on the heterogeneity of the investors in terms of investment horizon, such as showing the possibility for investors to switch to different funds from the same family at no cost and increasing the number of fund in the family, adding more options for the investors to select from (Brown & Wu, 2012;Clare et al., 2014). In addition to the characteristics of fund families, including size, age, and the number of funds in the family, therefore influence the flows of funds in these families.
A group of the funds is managed by one asset management company (AMC) called the family of funds, these fund families act as financial intermediaries offering a variety of mutual funds under a common brand name and via common marketing and distribution channels. Like any other typical goods market, interactions between investors (on the demand side) and the mutual fund families (on the supply side) determine the equilibrium, resulting in an equilibrium level of diversification as well. In addition to the differing characteristics of these families, and therefore, this difference influence the investor's behavior of the fund (Bhattacharya et al. 2013). Investing in many funds within a family of funds may present for many advantages, a fund family may provide for investors "one-stop" shopping, by providing more investment opportunities (Bhojraj et al. 2011;Bani Atta & Marzuki, 2020). Inflows would not only depend on the rank of a fund or family within the market, but also on the families and funds characteristics such as a number of the fund in the family, family size, and family age, this evaluation is important as some investors only consider one fund family when invest. The studies on family characteristics focus on conventional funds, so this study on of first studies focus on IMF (Nanda et al. 2004;Bani Atta & Marzuki, 2019a;, Adrianto et al. 2019. Family size is calculated as all equity funds under management by a company. The larger fund families can share expenses among a large number of funds, also can use the same economic data and experts to explain data across funds, in addition to large fund families can benefit from economies of scale from trading commissions and lending fees (Chen et al. 2004). These options act as externalities for all funds belonging to the same family, affecting the purpose level of flows the family needs to achieve and the number of funds it wants. In addition, the number of funds within the same family can increase the investors' selection options and thus increase the diversity that leads to increased flows. As there some studies indicate the IMF has the ability to achieve profit and preserve positive returns through the bear market (Abdullah et al. 2007;Abderrezak, 2008). Therefore, IMF considered an alternative in portfolio selection for investors, especially during bear markets, then can consider IMFs a good hedging investment, especially against market downturns that could The Saudi market is relatively new as the main TASI index was launched in 2003. It is intensively engaged in activities aiming to increase its financial liberalization, trying to enhance the market's efficiency and the corporate governance within, but still suffers the limitations existing in emerging markets in general as the unavailability of free, complete information, transaction costs, and illiquidity of some of its stocks. Conventional and Islamic financial systems operate side by side in KSA. However, because most of the population of Saudi Arabia is Muslims, ideally concerned with Islamic rules and principles, and expected to avoid dealing in interest-bearing transactions, more conventional banks prefer to deal in operations that are Sariah-compliant, aiming higher profits and bigger market share. The size of the mutual fund industry in Saudi Arabia ranks first among the GCC countries, followed closely by Kuwait and UAE. There are currently around 240 mutual funds in the Saudi capital market. The investments are spread across local and international stocks, bonds, trade finance and money markets www.tadawul.com.sa.
The objective of this study is to investigate the flow-performance relationship in IMF, in addition to the extent to which family and fund characteristics contribute to explaining fund flows in Saudi Arabia for the period from 2007 to 2017. The studying fund flows is very important to understanding the behavior of investors in Islamic mutual funds by knowing how the investors in IMF will respond to past performance and other fund characteristics, fund family variables. The significance of studying the flow-performance relationship lies in three aspects. The first is for fund families to determine, via fund flows, the asset under management (AUM) of fund management companies. Second, for the fund managers, by the convex flow-performance relationship that may encourage fund manager risk-taking to increase the likelihood that they are winners. Finally, the way flows respond to past performance also matters as it has implications for fund persistence. This is because the flow-performance relationship will determine the degree to which fund size is affected by a past performance which conditions how a fund performs in the future. This paper will address the following three questions:

Q1. Do Islamic mutual fund (IMF) investors consider past performance in their fund
selection process?

Q2. Do IMF investors consider fund characteristics other than performance when
undertaking an investment decision?

Q3. Do IMF investors consider family characteristics other than performance when undertaking an investment decision?
This study use data on Saudi Arabia IMF, to explore the above questions over the period from 2007 to 2017. The remains of the paper are organized as the following. Section 2 reviews the previous literature, Section 3 describes the data and models are used to evaluate the fund flows and test the fund and family characteristics, then section 4 provides the empirical results and discussion. Finally, section 5 will provide the conclusions.

Literature Review
According to behavioral finance theory, the psychological factors have an effect on the decision making of financial market-making, religion and social factor are one of these factors (Baber et al. 1999). IMFs investors, besides the expected financial gains, have certain non-financial motives that inclining toward religion (Sharia-compliance). Then, this could affect the decision of IMF investors when want to select their funds to direct their money.
One of the more important strands of research in the mutual fund literature is the study of fund flows. Most of this investigates the relationship between cash flows into mutual funds and past performance and other fund characteristics, while this study will contribute by add the family characteristics. Prior studies mainly concentrate on the USA, Australia, the UK and other developed capital markets (Chevalier & Ellison, 1997;Ippolito, 1992;Sirri & Tufano, 1998;Sawicki, 2001). These studies found that the flow-performance relationship of the investigated funds was positive and asymmetric. Funds that have performed well experienced higher money inflow, while funds that perform poorly experienced less money outflow. The implication from this finding is that investors allocate money disproportionately among poor and good performing funds; they would purchase funds with the best past performance but are reluctant to let go of poor performing ones. Whereas there is some evidence relating to the behavior of socially responsible funds (SRF) or ethical funds (Benson & Humphrey, 2008;Renneboog & Ter Horst, 2011;El Ghoul & Karoui, 2017), studies dedicated to IMF flows are regrettably very limited Azmi et al, 2018). The most evidence is that past performance (both raw and risk-adjusted returns) is a significant determinant of money flows into and out of mutual funds.
The beginning evidence indicates a positive relationship between past performance and fund flows, in addition to a positive relation between recent good (worse) performing funds and net money inflows (outflows). That means that investors use past performance to adjust their beliefs on managers' ability to generate superior returns. In addition, this indicates that investors consider recent past performance by rewarding better-performing funds with additional money and removing or not directing money to poorer-performing funds (Chevalier & Ellison, 1997;Del & Tkac, 2008;Sawicki, 2001). Whilst other studies found evidence of a convex or asymmetric flow-performance relationship. This suggests that better-performing funds enjoy huge money inflows, however, funds with poor performance do not suffer large money outflows (Ippolito, 1992;Sirri & Tufano, 1998). Recently, Leung & Kwong (2018) investigated the flow-performance relationship of mutual funds investing in emerging market economies (EMEs) from January 2000 to December 2016. These funds displayed a convex flow-performance relationship, suggesting that past performance is a significant factor driving fund inflow when the fund return is positive, but its influence vanishes when the return is negative. In addition to performance there are other factors may also influence fund flows, investors may consider non-financial attributes in making fund allocation decisions. Such as fund expenses, advertising, fund visibility, age, and size (see, Sirri & Tufano, 1998;Jain and Wu, 2000;Nanda et al. 2004).
On SRF case, the evidence about socially responsible funds has not risen to the level of evidence about traditional funds. Bollen, (2007) was the first study to consider the behavior of SRI investors in the USA. He compares the behavior of SRI fund investors and the behavior of CMF investors, this difference due to the investors of SRI buys mutual funds for both financial and social objectives. The results found that while SRI investors are more sensitive to past good performance, they are less sensitive to past poor performance compared to unscreened investors. He suggested conventional investors may have more options to switch to other funds compared to SRI investors. In addition, SRI investors consume non-financial attributes that mitigate the withdrawal of funds associated with negative performance. Consequently, SRI investors are more "loyal", at least in their reactions to poor fund performance. Renneboog et al., (2011) examine SRI and conventional funds from 21 countries, they found that the sensitivity of SRI funds' inflow-past good performance relationship was lower than their conventional counterparts. Adversely, flows into environmentally screened funds were more sensitive to lagged positive performance than conventional flows. These findings suggest that the flow-return relationship of SRI funds for past positive returns are country-and culture-specific, and that the valuation of ethical and social issues differ from an investor to another. Investors base their decision not only from the financial aspect but also non-financial factors, as suggested by the weak sensitivity of SRI funds' money flows to past negative performance. In contrast, A study of One such variable is fund age; older funds have existed longer, and may have created much awareness among investors due to their longevity. Depending on realized past performance, this awareness may be bad or good for the funds . Some studies argued that younger funds attract more fund flows due to the high marketing costs that they expend (Barber et al. 2005). Still, larger funds are expected to capture more money flow (Gruber, 1996),as they are able to bear higher marketing costs, thus gaining attention from the media. to better performing funds. The results also indicate there is a negative relationship between fund flows and fund risk may be due to IMF performing better during the bearish market since the risk is high. Finally, the results indicate there is a negative relationship between fund flows and fund size, family size, number of funds in the family, this may be because when family or fund become larger and included more funds can't preserve the same high growth rate.
This paper seeks to contribute by examining the family variables, the strategies of fund families are different to attract investors and improve their performance. Moreover, the family characteristics different, then this differential lead to different flows between different funds in these families.

Data and Sample
The sample in this study comprises 98 IMFs, distributed to 13 families, the study uses the flows of Islamic mutual funds in Saudi Arabia from 2007 to 2017, since the data about Islamic mutual fund (IMFs) are more available after 2007. The data comprise monthly returns for each fund, total asset, inception date for each fund, and market index prices collected from Bloomberg database, which provides information on mutual funds, it also offers data in the form of media and news, in addition to an Islamic finance platform, presenting comprehensive data on IMF and other Islamic instruments, and provides an Islamic window that provides data on a range of Islamic financial institutions. In addition to the rate of return on a 3-month Saudi Arabia Treasury bill collected from (www.Investing.com). The fund family's variables, such as family age and the number of the fund in each family collected from the annual reports of fund family. Since the family and fund age is an annualized, so the researcher converted to a monthly equivalent, to be consistent with the monthly returns of fund and market return.
The study uses a standard procedure for constructing the flow of funds by using the percentage money flow (FLOW), defined as money flow scaled by the total net asset of the fund as follows: Where, , is the total net asset value of fund i at time t, , −1 is the total net asset value of fund i at time − 1, , is the fund i raw return at time t. According to  this study uses raw return as a performance measure, where the previous studies indicate that the individual investors generally make fund selection decisions based on relatively basic measures, such as the historical raw return, in addition, the most appropriate measure of return is raw return as the average investor finds it easiest to calculate and understand (Sirri and Tufano 1998;Del Guercio and Tkac 2008).

The Models
The analysis section employs panel of fund month observations from 2007 to 2017. The pooled panel regression technique to estimate the relationship between fund flow and performance, fund characteristics, and family characteristics as follows: , = 0 + 1 , −1 + 1 , −1 + 2 , −1 + 3 , −1 + 4 , −1 + 5 , −1 + 6 , −1 + 7 , −1 + Where, , is the fund i flow at time t, , −1 is the fund i return at time − 1, , −1 is the fund i risk at time − 1 calculated from fund beta, , −1 is the fund i size at time − 1 calculated from log of fund total assets, , −1 is the fund i age at time − 1 calculated from fund inception date, , −1 is the fund family i age at time − 1 calculated from family inception date, , −1 is the fund family i size at time − 1 calculated from log family total asset, , is the number of fund in the family i at time − 1, , −1 is the fund i flow at time − 1 and is the error term.

Empirical Results and Discussion
In the regression model as shown in table 1, the Breusch-Pagan / Cook-Weisberg test for heteroscedasticity show H0: constant variance means there is no heteroscedasticity, and prob>chi2 is 0.101, more than 0.05 suggesting that the null hypothesis cannot be rejected.
Given the potential problems of multicollinearity among the fund attributes variables, a diagnostic check was performed using variance inflation factors (VIFs). As a rule of thumb, a VIF >10 is taken as an indicator of the presence of multicollinearity and the diagnostic results in table 1, show that none of the past performance, family and fund characteristics variables have a value greater than 10   The Correlation Between Imf Flow and Past Performance, Fund and Family Variables Table 3 presents pairwise correlations for IMF flows and independent variables. IMF flow is correlated positively with past performance, fund size, fund age, family size, past flow and the number of funds in the family. That suggest that the fund and family with big size attract more flow due to the more investment alternative provide by this fund and family.

The Summary of Descriptive Analysis
However, IMF flow is correlated negatively with fund risk, and family age. That suggest that the fund with less risk attract more flow due to this fund achieve good return in future.
In addition, old family attract less flow due to these funds are hard to manage.   The results in table 4 show that the coefficient of past performance is significantly positive (0.0602), that indicates that for every 1 percent increase (decrease) in returns the IMF flow will increase (decrease) by 0.06 percent. This means that the IMF investors reward (punish) better (worse) performing funds by moving money into (out) these funds. Finally, there is strong evidence that past fund flows (t-1) effects on future fund flows (t) (0.6164), can see that in the model, since increases of 1 percent in lagged fund flow lead to increase future flow by 0.616 percent, this due to the persistence in flows signals investors reinvest in the funds already owned. This result is consistent with Benson & Humphrey (2008) and . The results also show the fund flow insignificant with fund risk, that due to IMF try to avoid speculative investment and then investors do not care to the risk.

Imf Flow and Family Characteristics
This section focuses on results about the impact of the family characteristics on IMF flow, to explore whether IMF investors use the family characteristics when selection investment.
The studies on family characteristics focus on conventional funds, so this study on of first studies focus on IMF. Table 4 show the number of fund in the family have positive significantly relationship with IMF flow, that suggest the family with large number of funds attract more flows (0.048) this finding is consistent with the results of .
In the model can see when the number of funds increase by 1 percent the family attract a 4 percent higher fund flow, that may be because when family become larger and included more funds have more investment opportunity. In addition, there are inverse relationship between IMF flows and family age, which mean the flows will increase when the family is young. This can be attributed to the fact that young families are increasing advertising to attract more inflows. The results also show the fund flow insignificant with family size, that means the IMF investors don't look at family size when direct their money.

Conclusion
The objective of this study is to explore the IMF investors behavior by investigate the response of IMF investors to past performance and reflected in fund flows. The investors of IMF care about religious and ethical matters, then maybe these investors sacrifice some return performance. The results show there is a positive relationship between fund flows and past performance. That suggest the IMF investors make rational financial decisions by directing fund flows to better performing funds.
This study also aims to examine the relationship between fund characteristics and fund flows. One of the outstanding results, a positive relation between funds flows and past flows.
This due to the persistence in flows signals investors reinvest in the funds already owned. In addition to, the fund size and fund age are determinants of fund flows. Finally, this study seeks to contribute by examine the relationship between fund flow and family characteristics. The invest in many funds within a family of funds may present for many advantages, a fund family may provide for investors "one-stop" shopping, by providing more investment opportunities. The results show the family age and the number of funds in the family positively impact on fund flows.
The significance of studying flow-performance relationship lies in three aspects. The first is for fund families to determine, via fund flows, the AUM of fund management companies.
Second, for the fund managers, by the convex flow-performance relationship that may encourage fund manager risk-taking to increase the likelihood that they are winners. Finally, the way flows respond to past performance also matters as it has implications for fund persistence. This is because the flow-performance relationship will determine the degree to which fund size is affected by a past performance which conditions how a fund performs in the future. Finally, this study can encourage researchers in the realm of a mutual fund to further extend this subject. First, by expanding the study to other countries to include other countries that have a good number of Islamic funds. Second, by adding other characteristics of fund families like star and poor fund within the family, to know which of these characteristics could determine the flows of Islamic funds.