Analysis of Style Drift: Evidence from Malaysian Sukuk Fund

Transcription

1. Volume : 4 Issues: 18 [March, 2019] pp. 09-20] International Journal of Accounting, Finance and Business (IJAFB) eISSN: 0128-1844 Journal website: www.ijafb.com ANALYSIS OF STYLE DRIFT: EVIDENCE FROM MALAYSIAN SUKUK FUND Wan Rozima Binti Mior Ahmed Shahimi1 Ahmad Harith Ashrofie Bin Hanafi2 Wan Nur Izni Binti Wan Ahmad Kamar3 1 Faculty of Business and Finance, Universiti Tunku Abdul Rahman, Malaysia (Email: rozima@utar.edu.my) Faculty of Business and Finance, Universiti Tunku Abdul Rahman, Malaysia (Email: harith@utar.edu.my) 3 Faculty of Business and Finance, Universiti Tunku Abdul Rahman, Malaysia (Email: nurizni @utar.edu.my) 2 Accepted date: 8 December 2018 Published date: 31 March 2019 To cite this document: Shahimi, W.R.M.A, Hanafi, A.H.A, & Kamar, W.N.I.W.A. (2019). Analysis of style drift: Evidence from Malaysian sukuk fund. International Journal of Accounting, Finance, and Business, 4(18), 09-20. __________________________________________________________________________________________ Abstract: Investors make their decisions on what is known about a fund through all information regarding the strategies and fundamentals of investment fund. Reliable information is essential in order to make effective investment decisions and to construct an optimal portfolio return. However, funds can be incorrectly classified when evaluating them and making allocation decisions. The misclassification or style drift can also have impacts on fund performance. This study tries to identify the class asset on sukuk fund return, and to examine the style drift of sukuk fund from the return-based style analysis. The findings indicate that majority of sukuk funds invest mainly in short term money market and sukuk to generate fixed income for investors. The result also shows that only a few sukuk funds indicate significant drift in its style from the original objectives. Keywords: Investment Style, Asset Allocation, Return-Based Style Analysis, Style Drift, Sukuk Funds ___________________________________________________________________________ Introduction Investors allocate their investments in different asset classes as a vital strategy to set expectations for risk tolerance and potentially to escalate returns. Their investment decisions are based on style or objective of a managed fund (Brown, Harlow & Zhang, 2015; Elton, Gruber, Brown & Goetzmann, 2014; Kim, Shukla & Tomas, 2000). These styles of funds are disclosed in fund prospectus by the fund provider which also consists of investment policy, objective, risks and fees (Bams, Otten & Ramezanifar, 2017). In other words, the style analysis illustrates the investment style of the securities within a portfolio that helps investors to construct diversified portfolios. According to Bernstein (1995), equity investment style is the process of classifying assets into broad classes that influences the way investors make asset class allocation decisions. 9
2. Researchers such as Fama and French (1992, 1993) and Sharpe (1992) had documented several different analyses of equity investment style to identify the asset allocation and performance measurement. One of the seminal works is Sharpe (1992) who developed the return-based style analysis (RBSA hereafter). RBSA measures the portfolio returns and funds optimization with respect to the investment style groupings and market exposures on the basis of past performance. The prime goal is to determine the best mimicking strategy that is in accordance with the investment style of the fund (Annaert & Van Campenhout, 2007; Waring & Siegel, 2003). The main advantages of RBSA compared to Fama and French (1992, 1993) factor model are simplicity, objectivities and better predictor (Brown & Goetzmann, 1997; Gallo & Lockwood, 1997; Lobosco, 1999). Nevertheless, other studies suggested that the fund’s prospectus is not the best material for investors to obtain information about the fund in searching for the right fund to match the riskreturn acceptance level (e.g. Ainsworth, Fong & Gallagher, 2008; Bams et al., 2017; Kim et al., 2000). DiBartolomeo and Witkowski (1997) and Ainsworth et al. (2008) found return patterns are being seriously misclassified and contrasted from the assigned benchmark declared in the prospectus language because the fund managers do not strictly comply with their stated objective. This style inconsistency or style drift can lead to inferior performance (Bams et al., 2017). Style drift is studied whenever a fund deviates from its initial stated objective and investment style, and shifts towards other investment style. Undeniably, this situation can undesirably affect the investors in various ways (Brown & Goetzmann, 1997; Mintzer & Littmann, 2000). Firstly, investors are easily exposed to unsought risks that are not compatible with their own personal investment targets. Secondly, the style drift may also have negative effects on fund performance, thereby exposing the investors to undesirably returns. Brown et al. (2015) documented that fund managers who operate stated fund objectives with an inconsistent style are less likely to have high returns and make asset allocation errors compared to peer funds. Therefore, style drift contributes to a vibrant conflict of interest between managers and investors. In addition, the most common way to capture these changes in style exposures over time is to measure using RBSA (Waring & Siegel, 2003). From other perspective, the contemporary innovations in Islamic finance have transformed the innovations of its industry especially in the area of sukuk or Islamic fixed income securities (Kamarudin, Kamaluddin, Manan & Ghani, 2014). Sukuk fund is a diversified portfolio which consists majorly of Islamic bond and other proportion of securities under the principles of Shariah as an alternative to the conventional funds. Notably, Malaysia is the major sukuk market in the world and keeps getting the attention from the worldwide investors to observe the potential this type of asset class (Bakar, 2018). Nonetheless, there are still limited researches in this respect despite the fact that the prevalence of sukuk has increased among corporate issuers and investors. The empirical evidences concerning the effect of sukuk fund returns have yet to be analyzed in detail and comprehensive due to the lack of data (Zakaria, Isa & Abidin, 2012). As a result of the drawbacks identified in this section, the limitation of empirical evidence on sukuk funds with regard to the impact of misclassification on performance is worth considering. Thus, this paper aims to examine the sukuk fund from two perspectives; (1) to identify the class asset on sukuk fund return; and (2) to examine the style drift of sukuk fund from the RBSA analysis. 10
3. The remainder of the paper is structured as follows . The following section is about literature reviews as well as theoretical discussion which lead to the hypothesis development. The methodology and empirical results are presented in the subsequent chapter. The paper is finally followed by the discussion and conclusion chapter. Literature Review The study on equity style analysis has long stirred interests of many groups of people including financial advisors, money managers, and academicians to describe the characteristics of an investment portfolio. The style analysis later might reveal how the investment portfolios are being constructed according to one’s tolerance towards risks. There are few approaches used to examine the fund’s investment style; (1) RBSA (Sharpe, 1988, 1992); (2) Portfolioscomposition based approach; (3) Data development analysis (Alexakis & Tsolas, 2011). The RBSA is particularly used to assess the strategies on making asset allocation (Faff, Gharghori, Ip & Nguyen, 2012). When Sharpe (1988, 1992) first introduced the RBSA, Sharpe used the term “effective asset mix” and “attribution analysis” to explain the RBSA. However, in recent years, the term RBSA is often used to describe Sharpe’s method. The term is used to emphasize on the reliance on past returns as opposed to other methods of style or attribution analysis that do not rely on returns. Among the authors who have adopted RBSA approach includes Faff et al. (2012), Das and Rao, (2013), Dor and Jagannathan, (2002), and Domian and Reichenstein (2008). Faff et al. (2012) employed RBSA to assess the asset allocation strategies of managed and superannuation fund managers in Australia. The data observed were monthly returns for 447 managed funds and 453 superannuation funds for the period January 1990 to December 2011. The researchers stated that RBSA approach suits better when analyzing with multi-sector funds, which investments are made across a number of asset classes. The study found that riskier fund classes have greater exposure to riskier benchmarks. Superannuation fund managers appeared to adopt a much more conservative investment strategy compared to managed fund managers. Das and Rao (2013) in their study evaluated the performance of socially responsible funds by examining the fund’s investment style. They adopted RBSA to analyze the performance of 94 U.S. socially responsible mutual funds. The fund style was used as a benchmark in order to separate the performance that was attributed to style and selection. They found that socially responsible funds were underperformed and it was more pronounced and common compared to what had been documented previously. Active management of mutual funds also was identified to be the crucial determinant of their performance in socially responsible investing industry. These results concluded that actively managed socially responsible funds outperformed their passively managed counterparts. Majority of the previous literature documented used data from overseas like the U.S, Australia, and Spain. However, even with the vast previous literatures available adopting the RBSA (e.g. Castellanos & Alonso, 2004; Das & Rao, 2013; Domian & Reichenstein, 2008; Dor & Jagannathan, 2002; Faff et al., 2012), in the context of the Malaysian market, the evidence of previous literature on Malaysian sukuk funds is hardly to be found. Thus, this paper aims to fill in this gap. While the study on investment style has gained attentions of many researchers, it appears relevant to ask whether or not fund managers actually remain true to self-reported fund 11
4. indicators . This is because the style classifications are only useful if fund managers adhere to self-reported fund indicators (Brown & Goetzmann, 1997). It is crucial that a fund manager’s self-stated investment objective should be able to convey information to the investors at its highest accuracy, particularly about how the portfolio should be managed. However, previous studies had shown there was a substantial number of funds that deviated from their investment style (e.g. diBartolomeo & Wikowski, 1997; Brown & Goetzmann, 1997; Kim et al., 2000; Kim, White & Stone, 2005). Having style drift could lead to an undesired effect on the fund’s performance, risk and other fund attributes which may contribute to the failing of meeting the expectations of investors (Ainsworth et al., 2008). Besides that, there is also an issue in understanding style drift is in terms of how to define and measure a fund’s style (Ainsworth et al., 2008). According to Cooper, Gulen and Rau (2005) changes in the self-stated style of mutual funds in the U.S. had proven to affect fund’s flow with greater fund inflows experienced by funds that associate their name with styles that are more known and popular at that particular point of time. They also stated that fund managers do not always adhere to maintaining a portfolio consistent with their stated philosophy. In regard to this, diBartolomeo & Witkowski (1997) and Brown & Goetzmann (1997) also found that there was a misclassification problem suffered by some funds and most likely to affect its possibility to deliver the expected risk-return as what the investors desired. Moreover, diBartolomeo & Witkowski (1997) documented that around 40% of the analysed funds were misclassified. The funds observed were both classified in a lower-risk category as well as higher-risk category. Funds that are classified in higher-risk category tend to perform with better results than the average ones. Besides that, authors also analysed the factors that may have contributed towards misclassification problem. They found that the possibility of funds misclassification is lower if it belongs to larger management company. However, the risk of having fund misclassification is found to be higher if it involves large funds. This is due to greater complexities and difficulties associated with larger funds in managing the asset to be in line with their investment styles. Similarly, Kim et al. (2000) also found that 46% of the funds are classified based on their respective categories, however 54% were misclassified. They performed the discriminant analysis where the variables tested were used as the explanatory factors to examine fund classification. Brown and Goetzmann (1997) investigated the accuracy of mutual fund classification system and they also attempted to investigate whether the mutual fund classification system is relevant to be used in explaining the differences in future returns of funds. From the study, they concluded that the current classification system is inefficient. As the existence of conflict of interest between the managers and investors is identified, the authors had proposed an empirically determined classification system based on fund returns which was stated to be able to reduce the incentives for managers to deviate from their original investment styles. As a result, they found that the new proposed system has greater capability in explaining the future variability of fund returns. Cao, Illiev and Velthuis (2017) in their study stated that investors are likely to pick small-cap funds with the assumption that the risk exposures will be relatively low. This has led fund managers to become specialists in particular styles to provide information on the investment set and will also explain the risk exposures to investors. This is particularly true if managers adhere to the stated style however in the event of deviation, investors will be exposed to 12
5. unanticipated risks . Besides, it is also worth knowing that if there is a possibility of obtaining gains though in the short-term, this could also lead the managers to deviate from their given objectives due to labour market pressures. This situation presents a conflict of interest between managers and investors. They found that larger and older small-cap funds are more likely to hold mid- and large-cap stocks. It is consistent with funds deviating from their objective over time. Given all the findings documented from previous studies, it is important to take note that the difference in underlying assets characteristics may change overtime, to a certain extent deviation from the original fund’s investment style on a long-term basis is inevitable. This could be potentially true if a fund manager is passively holding the same stocks over time. For example, the stock might be a small-cap stock and over time may grow up and become a largecap stock (Bams, et al., 2017). Style deviation to a certain extent may be accepted however if it is too much, it can be an indicator that fund managers have changed the investment strategy and could be moving away from the fund’s stated objective. Therefore, this situation could affect investors’ level of risk and return. Nonetheless, if the degree of deviation is too large, it might unconsciously shift investor’s investment objectives over the long-run period. Thus, it is crucial for investors to take into account the degree of the style deviation to ensure that they are still in line with their long-term investment objectives. Thus, it is equally crucial also to know the extent of style deviation by quantifying it and to put mutual funds in terms of investment style deviation (Bams et al., 2017). Hence, based on the literature in this research, our testable hypothesis is: H0: There is no significant different in style drift and asset allocation from funds’ original objective. Research Methodology Data Collection The sample of this study was selected based on the prospectus of unit trust funds issued by the asset management companies in Malaysia and the Federation of Investment Managers Malaysia (FIMM). There are 18 funds under the category of sukuk or Islamic bond funds available within January 2013 to December 2017 as shown in Table 1. Table 1: List of Selected Sukuk and Islamic Bond Unit Trusts Fund (Fund Objective) Asset Allocation Amanahraya Syariah Trust Minimum 70% will be invested in Sukuk, 30% in (Fixed Income) Islamic Money Market Instruments AMB Dana Arif 70% to 98% will be invested in Sukuk and Shariah(Balanced) compliant money market instruments, 2% to 30% in Shariah-compliant liquid assets AmDynamic Sukuk - Class A 70% to 98% will be invested in sovereign, quasi(Fixed Income) sovereign and corporate Sukuk, at least 2% will be invested in liquid assets AmBon Islam 70% to 100% will be invested in Sukuk and Islamic (Fixed Income) money market securities, 30% in Shariah-compliant liquid assets 13
6. CIMB Islamic Enhanced Sukuk 70 % to 98% in Sukuk, 40% in Unrated Sukuk, 0% to (Fixed Income) 20% in Shariah-compliant equities, up to 10% may be invested in warrants, at least 2% in Shariah-compliant liquid assets CIMB Islamic Sukuk 70% to 98% in Sukuk, 28% in other permissible (Balanced) investments, 40% in Unrated Sukuk, at least 2% in Shariah-compliant liquid assets Dana Al-Fakhim May be invested in short-term debentures, money (Fixed Income) market instruments and placement in short-term deposits Eastspring Invt Dana Wafi Minimum 70% in Sukuk, minimum 1% in Islamic (Fixed Income) Deposits or Islamic liquid assets Hwang AIIMAN Income Plus Minimum 80% in Sukuk, minimum 20% in cash and (Balanced) Islamic money market instruments Kenanga Bon Islam About 50% to 98% in Sukuk, 2% to 50% in cash (Fixed Income) Libra Asnita Bond About 70% will be invested in quoted Shariah(Fixed Income) compliant equities and equity-related securities, minimum 2% in Islamic liquid assets MAAKL As-Saad Nearly 100% in money market (Fixed Income) MIDF Amanah Shariah Money About 90% will be invested in Islamic Deposits, Islamic Market Money Market Instruments, 10% in Islamic Short-Term (Balanced) Debt Instruments Pacific Dana Murni Minimum 95% in Sukuk, 5% in cash and other Shariah(Fixed Income) compliant liquid assets RHB Islamic Bond Minimum 60% in Sukuk, minimum 5% in Shariah(Fixed Income) compliant liquid assets PB Sukuk Minimum 75% to 98% may be invested in Sukuk, (Fixed Income) others in Shariah-compliant liquid assets PMB Sukuk Minimum 70% to 99.5% in Sukuk (Balanced) Public Sukuk Minimum 75% to 98% in Sukuk, others in Shariah(Fixed Income) compliant liquid assets Source: Fundsupermart Malaysia As for asset classes to show the styles of unit trusts such as growth stocks, value stocks, cash, sukuk, and international stocks, data collected from various website are presented in Table 2. These asset classes were selected based on Lau (2008). 14
7. Asset Class Growth Stocks Value Stocks Cash Sukuk International Stocks Table 2: Asset Classes’ Indices Description *MSCI Malaysia Growth Index as quoted in MYR used to represent growth stock. *MSCI Malaysia Value Index as quoted in MYR used to represent Value stock. Islamic fixed deposit profit rate to represent Malaysian Islamic money market instrument Bloomberg Malaysian Sukuk Ex-MYR Index is used as index for this asset class which represents Malaysia fixed income markets. *MSCI World Index as quoted in MYR is used to represent all international stock indexes. Source of Data www.msci.com www.msci.com Bank Negara Website Bloomberg terminal www.msci.com * MSCI indices developed by Morgan Stanley Capital International According to Sharpe (1992), the selected asset classes should follow some criteria since the usefulness of the analyzed results is highly dependable on the asset classes. The criteria are that all asset classes should be (i) mutually exclusive or asset classes should be in one class only; (ii) exhaustive or it represents all assets within the same class, and (iii) the return of the asset classes should have low correlations or different standard deviation if the correlations are high to ensure they represents the specific category. In order to fulfill the objectives of the study, style analysis (Sharpe, 1992) was used. Sharpe (1992) used quadratic programming to conduct RBSA in order to examine the changes in the funds’ asset allocation based on the changes in the returns of major asset classes. According to Sharpe (1992), style analysis has two major constraints: (1) the sum of all coefficients factors is equal to 100% and (2) coefficients of all factors must be positive. Negative coefficients show that there are short positions in asset classes. However, Sharpe (1992) argued that fund managers rarely use the short position strategy as fund managers use buy-and-hold investment strategy. Thus, by prohibiting negative coefficients in the model, it will provide better and more usable results. This analysis is basically based on Sharpe’s (1992) generic factor model which is as follows: ̃2 +