Is it costly to be both shariah compliant and socially responsible?

Is it costly to be both shariah compliant and socially responsible?

    Is it costly to be both Shariah-compliant and Socially Responsible? Elias Erragragui, Christophe Revelli PII: DOI: Reference: S1058-...

242KB Sizes 1 Downloads 9 Views

    Is it costly to be both Shariah-compliant and Socially Responsible? Elias Erragragui, Christophe Revelli PII: DOI: Reference:

S1058-3300(15)30058-6 doi: 10.1016/j.rfe.2016.08.003 REVFIN 385

To appear in:

Review of Financial Economics

Received date: Revised date: Accepted date:

30 June 2015 27 June 2016 23 August 2016

Please cite this article as: Erragragui, E. & Revelli, C., Is it costly to be both Shariah-compliant and Socially Responsible?, Review of Financial Economics (2016), doi: 10.1016/j.rfe.2016.08.003

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Title : Is it costly to be both Shariah-compliant and Socially Responsible?

RI

PT

Corresponding author: Elias Erragragui Assistant Professor Kedge Business School Rue Antoine Bourdelle Domaine de Luminy BP 921 13 288 Marseille Cedex 9 – France [email protected] Tél. +33 (0) 491 827 324 – Fax. +33 (0) 491 827 983

NU

SC

Author 2 : Christophe Revelli Professor of Finance Kedge Business School Rue Antoine Bourdelle Domaine de Luminy BP 921 13 288 Marseille Cedex 9 – France [email protected]

MA

Abstract :

AC CE P

TE

D

Positive ethics associated with socially responsible investments (SRI) is challenging the limits of Islamic investments’ conservative approach to promote corporate social responsibility. In this study, we test the integration of social performance measures (companies the most virtuous or high-rated in terms of Environmental Social and Governance (ESG) issues) in Islamic portfolios using KLD social ratings. We seek to determine the financial price of complying both to Islamic investment and SRI principles. To do so, we measure the financial performance of self-composed Islamic portfolios with varying ESG scores. The results indicate no adverse effects on returns due to the application of ESG screens on shariah-compliant stocks during the 2007-2011 periods while reporting substantially higher performance for the portfolios with good records in governance, products, diversity and environment issues. On the opposite, a negative performance is associated with an SRI strategy of disengagement from shariah-compliant stocks with community and human rights controversies. Our performance measures are controlled for market sensitivity, investment style, momentum factor and sector exposure.

KEYWORDS: Islamic investment; Environmental, Social and Governance (ESG) performance; Investment decision; Portfolio choice.

JEL: A13, G11, G15, M14

ACCEPTED MANUSCRIPT Is it costly to be both Shariah-compliant and Socially Responsible?

PT

ABSTRACT

D

MA

NU

SC

RI

Positive ethics associated with socially responsible investments (SRI) is challenging the limits of Islamic investments’ conservative approach to promote corporate social responsibility. In this study, we test the integration of social performance measures (companies the most virtuous or high-rated in terms of Environmental Social and Governance (ESG) issues) in Islamic portfolios using KLD social ratings. We seek to determine the financial price of complying both to Islamic investment and SRI principles. To do so, we measure the financial performance of self-composed Islamic portfolios with varying ESG scores. The results indicate no adverse effects on returns due to the application of ESG screens on shariah-compliant stocks during the 2007-2011 periods while reporting substantially higher performance for the portfolios with good records in governance, products, diversity and environment issues. On the opposite, a negative performance is associated with an SRI strategy of disengagement from shariah-compliant stocks with community and human rights controversies. Our performance measures are controlled for market sensitivity, investment style, momentum factor and sector exposure.

AC CE P

TE

KEYWORDS: Islamic investment; Environmental, Social and Governance (ESG) performance; Investment decision; Portfolio choice.

JEL: A13, G11, G15, M14

ACCEPTED MANUSCRIPT 1. Introduction

AC CE P

TE

D

MA

NU

SC

RI

PT

In 2013, a Human Rights Watch report1 revealed the systematic “ethnic cleansing” of the Rohingya minority and the extensive human rights violations perpetrated by Burma’s oppressive military regime over the course of more than a decade. Active oil companies in the region were found to be highly involved in those controversial human rights violations. Unocal (a subsidiary of Chevron group) was named in a civil lawsuit in the U.S. federal court. In response, ethical indexes providers decided to exclude Chevron from their universe. Yet, while Islamic investors argue their willingness to conform to Islamic ethics, controversial firms such as Chevron are often prominent within Islamic indexes2.Yet a recent survey supported by the Global Reporting Initiative reveals that a large majority of professional investors consider the use of extra-financial data “very relevant to their investment decision making”3. Today, media diffusion and increasing awareness of firms’ social controversies challenge the current conservative practice of shariah4 screening. Unlike conventional type investments, socially responsible investments (SRI) apply a set of screening methods to exclude or include stocks based on Environmental, Social and corporate Governance (ESG) criteria, often engaging in local communities and shareholder activism to foster corporate strategies with such aims (2008). Islamic investment refers to an investment practice that conforms with shariah whereby guidelines and principles govern all aspects of human activity including investment practices such as portfolio allocation, trading activity and dividend distribution (Girard & Hassan, 2008). Current Islamic investment’s screening exclude activities or instruments considered “illicit” (Miglietta & Forte, 2007). Namely, fund managers undertake industry and financial screening5 to ensure the final portfolio’s compliance with Islamic legal prescriptions. Islamic asset management sector has come a long way since the first Islamic fund appeared several decades ago. This niche market reached $62 billion in 2014 according to Reuters6, still minimal in comparison with SRI assets, which in socially screened portfolios rose to $10.4 trillion at the start of 2012 for the European and U.S markets7, the two core SRI markets. Proponents of Islamic finance argue that the ethical principles emphasized by Islamic finance can help reducing the disconnection between the financial sector and the “real economy” due to its equity-based nature (Warde, 2010; Choudhury, 2010). They highlight that Islamic finance can reconcile modern finance with the original “spirit of capitalism” by virtue of its market-driven allocation of resources. According to Warde (2010, p.169) “Islamic finance could bring about more efficient mobilization of savings, more equitable and just distribution of resources, more responsible and profitable lending, as well as less volatile business cycles and more stable banking systems”. Beside this, some point out that in order to be in line with its ethical principles, Islamic finance needs not only to promote economic development but a sustainable economic development that would preserve the rights of future generations on earth resources (Al-Damkhi, 2008). These objectives that are commonly upheld by SRI promoters triggered our interest in investigating the convergence of their respective practices. The general question that motivated this research is: should Islamic investors include environmental, social and governance (ESG) criteria in their investment decisions? 1

available at http://www.hrw.org/sites/default/files/reports/burma0413webwcover_0.pdf. For e.g., the Dow Jones Islamic Market and MSCI Islamic index series. 3 available at http://www.accountingforsustainability.org/wp-content/uploads/2012/07/What-investors-andanalysts-said-The-value-of-extra-financial-disclosure.pdf 4 Shariah is usually referred to as Islamic law but also embodies the ethical, moral and legal principles governing all aspects of Muslim life. 5 The first stage or industry screening considers the company's main activities while the second stage or quantitative screening refers to debt leverage, liquidity ratio and interest exposure. 6 estimation extracted from Reuters Global Islamic Asset Management Report published in 2014. 7 Figure extracted from the USSIF and European SRI Study 2012 reports. 2

ACCEPTED MANUSCRIPT

NU

SC

RI

PT

Although Islamic finance and SRI appear to trigger the same expectations among their proponents in terms of being more ethical than conventional finance, they also face the criticism of being unable to meet these expectations, as suggested by the “form-over-substance” debate among Islamic finance scholars and the “green-washing” issue raised by SRI critics (Hayat, 2013). Islamic finance and SRI share similarities in their objectives claiming for the promotion of social welfare while emphasizing on ethics. Several studies have highlighted the compatibility of Islamic ethics with classical business ethics theory (Beekun & Badawi, 2005; Brammer, Williams, & Zinkin, 2007; Dusuki, 2008; Dusuki & Abdullah, 2007; Rice, 1999; Williams & Zinkin, 2010). Its proponents agree however that a strategy focusing solely on excluding “illicit” activities is insufficient to comply with all the positive ethics and social prescriptions extracted from Islamic sources. A survey conducted among Islamic finance practitioners notes that 98.8% of respondents believe that promoting social responsibility in financial transactions would reconcile Islamic financial institutions with their ethical aspiration (Sairally, 2007). The author of the survey concluded that Islamic investment could further “learn from the more proactive engagement practices of SRI funds whereby they encourage companies to be more responsive to society’s expectations”.

TE

D

MA

Despite these calls for reconciliation, no serious attempts have been made to combine these two ethical investing styles (Hayat, 2013). Our research intends to fill this gap by examining the effect of positive SRI strategy (integrating the companies the most virtuous in terms of ESG scores) applied to an Islamic stock universe. We design an experimental portfolio construction approach in which we use various ESG screenings to specify whether socially responsible Islamic portfolios differ in their performance characteristics from their counterparts. Due to potential bias linked to sector exposure, we test the robustness of the results using an industry-adjusted performance model.

AC CE P

From a practical point of view, by combining two different forms of ethical investing practices this research intend to identify diversification benefits for investors from both families, seeking to include new asset classes in their portfolio composition. Furthermore, investigating the financial merit of SRI initiatives may provide arguments for the promotion of transparency in social accountability for Islamic finance regulatory institutions. In the next section, we present the theoretical background from which our hypothesis and the empirical analysis framework derive. In the third section, we describe the methodological settings of the study. In the fourth section, we report the results and test their robustness. In the last section, we present the main findings of our research and offer conclusions.

2. Literature review This section reviews the literature on SRI and Islamic investment performance as compared to unrestricted traditional investment. We provide a general view of the impact of Islamic and SRI screenings on the risk, return and portfolio investment style characteristics.

2.1. Does Islamic investment’s performance suffer from screening constraints? The resilience of the Islamic financial market during the last “subprime” crisis has triggered the interest of many researchers. Previous findings suggest that sector screening imposed by Islamic funds would not seem to negatively affect performance (Ahmed, 2001; Wilson, 2001). Similar findings where provided by Hakim and Rashidian (2004), Elfakhani and Hassan (2005), Abderrezak (2008)

ACCEPTED MANUSCRIPT

D

MA

NU

SC

RI

PT

and Hayat and Kraeussl (2011) who report no statistically significant difference between the risk adjusted performance of Islamic mutual equity funds or indexes and their conventional benchmarks. Hussein and Omran (2005) suggest that Islamic indices may provide better returns than conventional indices in times of financial distress. The authors point out that the exclusion of high-profile firms such as Enron and WorldCom, due to their excessive indebtedness, enabled Dow Jones Islamic Market Index (DJIMI) to outperform its conventional counterparts after both firms collapsed in respectively 2001 and 2002. Significant international studies also confirm the outperformance observed during the last subprime crisis by Islamic developed market indices (Jawadi, Jawadi, & Louhichi, 2014; Walkshaeusl & Lobe, 2012) and funds (Hoepner, Rammal, & Rezec, 2011), confirming prior conclusions in relation to the Asian financial crisis in 1997 (Abdullah, Hassan, & Mohamad, 2007). Walkshaeusl and Lobe (2012) note that the outperformance was largely attributable to the exclusion of financial stocks from shariah-screened portfolios, predicting that such superior performance shouldn’t persist over time. Abdullah et al. (2007) and Merdad et al. (2010) confirm this insight as they highlight the underperformance of Islamic funds during buoyant periods. Hoepner et al. (2011) assert that Islamic equity funds “exhibit a hedging function”, as their investment universe is limited to low leveraged stocks suggesting the hedging benefit that conventional investors can derive from this “exotic” investment style. All these studies provide an avenue to investigate how the outperformance recorded by Islamic investments during times of financial distress can be sustained during stable periods through the inclusion of long-term risk mitigating factors.

TE

2.2. Why add SRI screening to Islamic portfolios?

AC CE P

From a social impact perspective, one of the potential merits of incorporating positive ESG screening into Islamic investments is in counteracting its adverse environmental effect. Indeed, Islamic portfolios are found to be much more exposed to industrial and fossil energy firms (Forte and Miglietta, 2007) which are more frequently involved in environmental issues. This overexposure to polluting activities, in contradiction with Islamic environmental ethics (Rice, 2006), triggers the criticisms of pro-environmental activists (Al-Damkhi, 2008). The screening methodology adopted by Islamic investors is in theory subject to a continuous process of revision and improvement in accordance with the contemporary opinions of modern legal scholars. A study conducted by Derigs & Marzban (2008) helped to examine the different screening approaches adopted by Islamic indexes’ providers. They touch on some points that are crucial to the development of the Islamic fund management industry as they revealed the inconsistencies of shariah compliance criteria. As such, one firm could be considered shariah compliant for one index and non-shariah compliant for another. Arguably, the authors conclude that “mathematical formalism may not be able to fully account for the subtitle and subjective interpretation of the Islamic sources” and that “the effect of bundles of such formal constraints may be too complex to be anticipated on every possible asset universe”. This insight provides justification for the need to develop the current shariahcompliance approach toward a more practical inclusion of social objectives. From a financial theory viewpoint, despite Modern Portfolio Theory’s classical argument suggesting an under diversification cost for screened portfolios (Markowitz, 1952), many SRI studies provided contradictive conclusions. They refute the traditional argument suggesting that a reduction in the stock universe imposes an additional set of constraints to the optimization problem faced by returnmaximizing investors. In fact, some researchers have found that a simple trading strategy relying on SRI stock selection may offer superior portfolio performance (Derwall, Guenster, Bauer, & Koedijk, 2005) even after including various levels of transaction costs (Brzeszcsynski & McIntosh, 2013;

ACCEPTED MANUSCRIPT

PT

Kempf & Osthoff, 2007). Barnett and Salomon (2006) further show that SRI funds’ performance increases with intensified screening. Capelle-Blancard and Monjon (2012) note that greater strategy distinction is associated with better financial performance and that positive screenings promoting best ESG practices are less harmful than negative screenings that exclude entire sectors. The study conducted by Statman and Glushkov (2009) suggest that investors benefit from a return advantage of tilts towards stocks with high ESG scores but note that this advantage is offset by the return disadvantage deriving from the exclusion of stocks from banned sectors.

NU

SC

RI

From a risk perspective, a diversification benefit can also be expected from the application of an SRI strategy within a shariah-compliant stock universe. Forte and Miglietta (2007) show that Islamic and SRI investments have different characteristics in terms of asset allocations, econometric profiles and sector exposure. Empirical evidences show that Islamic funds are more tilted towards growth and small-cap stocks while SRI funds are more value and large-cap oriented 8 (Bauer, Koedijk, & Otten, 2005). Integrating SRI criteria to an Islamic portfolio or inversely can therefore provide diversification benefits to funds’ managers by reducing portfolios’ idiosyncratic risk due to risk profile differences between Islamic and SRI stocks.

D

MA

All these observations pave the way for additional studies addressing the still open debate on Islamic and SRI portfolios’ performance. Our experimental innovative portfolios’ study seeks to contribute to the debate by investigating the performance of a large panel of shariah-compliant SRI portfolios with various ESG profile.

AC CE P

TE

Finally, and from a management viewpoint, the adoption of ESG standards by Islamic funds managers should contribute to a reduction of firms’ cost of capital on the account of the incentives to disclose their CSR policies (Dhaliwal et al. 2011). CSR disclosure policies push firms’ managers to favor equity financing over debt leading to a “shariah-compliance virtuous circle” in firms being less subject to exclusion from Islamic funds due to their lower leverage. Moreover such inclusion of ESG criteria by Islamic funds would attract conventional institutional investors as they represent the major drivers of SRI markets (Ussif, 2014). As expressed by Hayat (2013), attracting institutional investors represents one of the crucial stage of development to be achieved by Islamic asset management industry in order to sustain its growth. To do so, Islamic funds are encourage to converge toward SRI practices by, for instance, adhering to UN PRI principles.

3. Data and methods 3.1. Dataset Our dataset combines ESG ratings and financial data on 238 firms listed in U.S. stock markets as of 2007 and up to 20119 . MSCI Corporation provided us with the constituents of the MSCI U.S. Islamic index from its inception in 2007 which forms our shariah-compliant stock universe10. As KLD is the property of MSCI, we assumed the MSCI Islamic index series coverage would match the KLD 8

See Abderrezak (2008), Girard and Hassan (2008) and Walkshaeusl and Lobe (2012) for Islamic investment and Bauer et al. (2005) for SRI. 9 We imported the ESG ratings from Kinder, Lynderberg and Domini (KLD) Research and Analytics STATS (Statistical Tool for the Analysis of Trends in Social and Environmental Performance) and retrieved financial data from Datastream. Our KLD dataset was only available up to 2011. 10 The methodology used by MSCI to screen Shariah-compliant stocks is available at http://www.msci.com/eqb/methodology/meth_docs/MSCI_May11_IslamicMethod.pdf

ACCEPTED MANUSCRIPT universe, to ensure broader stock coverage. Of the 270 firms listed in the MSCI U.S. Islamic index in 2007, we retained the 238 firms that received ESG ratings from KLD.

RI

PT

KLD provides annual snapshots of the ESG performance of U.S companies, evaluated on multiple criteria that constitute two broad categories: qualitative and exclusionary. The qualitative criteria are used for the positive and best-in-class strategies. The exclusionary screenings eliminate companies involved in controversial business areas11. The seven qualitative screenings are: community, governance, diversity, employee relations, environment, human rights and products.

MA

NU

SC

For each domain, KLD lists several criteria and considers both strengths and concerns. In each domain, each constituent criterion receives a score of either 0 or 1 and the rating lists: strengths, concerns, both or neither. Concerns indicators measure the severity of the controversies the firm faces (e.g. fines/sanctions for causing environmental damage, toxic chemical emissions, poor employee union relations, abuse of employee labour rights, etc.). Strengths indicators measure positive ESG engagement or initiatives. It can be related to products/services, management, policies or operations (e.g. air emission mitigation, use of recycled materials, establishment of pro-minority or local community involvement policies (e.g., indigenous peoples), development of employee benefits, etc.).

3.2. ESG performance measurement

AC CE P

TE

D

Barnett and Salomon (2011) observe that most Corporate Social Performance (CSP) studies based on KLD data use aggregate measures. Many researchers pointed out the limits of aggregate measures (see for e.g. Brammer, Brooks, & Pavelin, 2006). Stanwick and Stanwick (1998) show that CSP is a multidimensional concept, therefore arguing that studies on the relation between CSP and financial performance should take various dimensions into consideration. Similarly, Sharfman (1996) notes that the simple addition of positive and subtraction of negative ratings across screenings using KLD data fails to give an accurate picture of ESG performance. As suggested by Waddock and Graves (1997), investors allocate different levels of importance to different components of ESG performance. In their study, Galema et al. (2008) argue as well that composite ESG performance measures may confound existing relationships between its specific sub-components and returns. For example, positive news about a firm’s recycling policies may positively relate to expected returns, whereas news on philanthropic activities, perceived as an unproductive cost, relate negatively. These conclusions motivated our choice to rely on a disaggregated ESG scoring approach similar to Dravenstott and Chieffe (2011) (i.e. we use a separate score for each ESG sub-components). Our final portfolio composition approach differs slightly from Dravenstott and Chieffe’s study in that we differentiate portfolios according to ESG performance’s sub-components, dimensions and intensity.

3.3. Portfolios composition We adopt a self-composed portfolio approach to circumvent the methodological biases affecting traditional fund studies (Orlitzky, Schmidt, & Rynes, 2003). Indeed, some authors point out the lack of transparency and uniformity in the rating process of SRI mutual funds so that ratings tend to differ according to local regulations, investor preferences and fund manager abilities. It is therefore difficult to neutralize the specific social premium effect attributed to each ESG indicator (Kurtz & Dibartolomeo, 2011). 11

Namely: alcohol, gambling, tobacco, firearms, military and nuclear power.

ACCEPTED MANUSCRIPT

NU

SC

RI

PT

The portfolio composition thus relies on a simple, passive, value-weighted approach which controls for manager stock-picking ability and enables measuring the effects attributed exclusively to ESG strategic allocations. We isolate each of the seven ESG domains and separate out the positive dimension (strengths) from the negative one (concerns). Next, we classify the score according to three rankings: 0, 1 or greater than 1. For each ESG sub-component, and depending on the type of dimension, 0 indicates either “no ESG engagement” or “no involvement in ESG controversies”; 1 indicates “partial ESG engagement” or “partial ESG controversies involvement”; and a score greater than 1 indicates “significant ESG engagement” or “significant ESG controversies involvement”. Portfolios formed of firms with no ESG engagement or no involvement in ESG controversies are labelled “neutral portfolios”; portfolios formed of firms with partial ESG engagement (resp. partial involvement in ESG controversies) are labelled “partially socially responsible” (resp. “partially irresponsible”) and portfolios formed of firms with significant ESG engagement (resp. significant involvement in ESG controversies) are labelled “significantly socially responsible” (resp. significantly irresponsible”. The final panel consists of 34 portfolios12 to which we add a set of five portfolios constructed on the basis of aggregated scores for comparative purpose. Table 1 describes the portfolio panel structure.

MA

Insert table 1

TE

D

The portfolios are non-mutually exclusive which implies that a stock can be part of several portfolios. In order to identify the possibility that two portfolios have large portion of stocks in common we report the coefficient of cross section correlation between the different ESG domains’ ratings. Insert table 2

AC CE P

Table 2 reports the correlation coefficients between the different ESG strengths and concerns across 227 constituents of the MSCI ISLAMIC index by December 2010. The table reveals significant level of correlation among both strengths and concerns, with however more significant results for strengths. The highest correlation coefficient is 0.63 (at 1% level) between environmental and governance strengths. We may expect that some portfolios’ composition display a certain level of similarity. In terms of timing, at the end of year t – 1, KLD reports the stock ratings. We suppose that the investors rebalance their portfolios once a year based on these ratings. Accordingly, we form the portfolios at the beginning of year t and maintain the portfolios until the end of year t. New ratings are then published and we restructure the portfolios for year t + 1. We thus obtain a time series of monthly returns from 2008 to 201113.

3.4. Financial performance measurement To compare the portfolios’ performances we use four alternative measures of investment performance. These are Smith and Tito ratio, ܶ௜ ; Sharpe information ratio, ܵ௜ ; excess standard deviation adjusted return, eSDAR; and Carhart four-factor model. Smith and Tito ratio (1969) consists of portfolio’s excess return (Jensen’s alpha) adjusted for systematic risk:

12

Due to the limited number of stocks with scores above 1 in our panel we obtained a smaller portfolio panel. The period limitation is due to the historical inception of the MSCI Islamic index that was launched by the end 2007 13

ACCEPTED MANUSCRIPT ܶ௜ =

ఈ೔

(1)

ఉ೔

We estimate Jensen’s alpha as: ܴ௜,௧ – ܴ௙,௧ = ߙ௜ + ߚ௜ (ܴ௠,௧ − ܴ௙,௧ )+ ߝ௜,௧

PT

(2)

RI

where ܴ௜,௧ is the return on portfolio i constructed as previously explained; ܴ௠,௧ is the return in month t on a value-weighted market proxy and ܴ௙,௧ is the return in month t of a one-month Treasury bill extracted from Kenneth French’s data library14.

ഥ ஽ ఙವ

NU

ܵ௜ =

SC

We also use the information ratio, ܵ௜ , suggested by Sharpe (1994). We estimate the information ratio as: (3)

MA

ഥ is the average value of the monthly differences in return between portfolio i and its where ‫ܦ‬ benchmark ܴ௜,௧ – ܴ௠,௧ and ߪ஽ is the standard deviation of the differential return. Note that unlike the Smith and Tito ratio, the Sharpe performance measure adjusts for total risk rather than just systematic risk.

ܴ݅ – ܴ݂ . ߪ݉ ߪ݅

− ܴ݉

AC CE P

eSDAR = ܴ௙ +

TE

D

The third measure of portfolio’s performance is eSDAR (Statman, 2000). The eSDAR of a portfolio represents the excess return of the portfolio over the return of the benchmark portfolio. We measure the eSDAR as: (4)

where ܴ௙ is the monthly return on one-month Treasury bills, ܴ௜ is the monthly return on portfolio i, ܴ௠ is the monthly return on the benchmark portfolio, ߪ௜ is the standard deviation of portfolio i’s return, and ߪ௠ is the standard deviation of return on the benchmark portfolio. We then use the four-factor model based on Fama and French (1993) factors and completed by Carhart (1997) factor measured by the following equation: ܴ௜,௧ – ܴ௙,௧ = ߙ௜ + ߚ௜ (ܴ௠,௧ − ܴ௙,௧ ) + ‫ݏ‬௜ ܵ‫ܤܯ‬௧ + ℎ௜ ‫ܮܯܪ‬௧ + ݉௜ ‫ܯܱܯ‬௧ + ߝ௜,௧

(5)

Where ܵ‫ܤܯ‬௧ is the difference in monthly returns between small and large-cap portfolios; ‫ܮܯܪ‬௧ is the difference in returns between value and growth portfolios; and ‫ܯܱܯ‬௧ is the monthly return on a portfolio long on past one-year winners and short on past one-year losers. The momentum factor captures the risk due to the momentum found in stock returns (Jegadeesh & Titman, 1993). Controlling for investment style is particularly important in light of mounting evidence that returns on style portfolios account for a considerable portion of SRI portfolio performance (Bauer, et al., 2005). In addition, similar to previous works (see for e.g. Derwall, et al., 2005; Galema, et al., 2008) we test returns on a difference portfolio to measure relative performance of socially responsible portfolios compared to their non-socially responsible counterparts. To construct the difference portfolios, we first subtract the returns of portfolios engaged in ESG strengths (panels C+E) from portfolios not-engaged

14

Available at http://mba.tuck.dartmouth.edu/pages/faculty/ken.french.

ACCEPTED MANUSCRIPT in ESG strengths (panel A) and thereafter returns of not-controversial portfolios (panel B) from controversial portfolios (panel D+F). The corresponding equation is as follows: ܴ௜,௧,௣ – ܴ௜,௧,௡ = ߙ௜ + ߚ௜ (ܴ௠,௧ − ܴ௙,௧ ) + ‫ݏ‬௜ ܵ‫ܤܯ‬௧ + ℎ௜ ‫ܮܯܪ‬௧ + ݉௜ ‫ܯܱܯ‬௧ + ߝ௜,௧

(6)

SC

RI

PT

where ܴ௜,௧,௣ represents the returns of socially responsible portfolios and ܴ௜,௧,௡ is the return on their accompanying non-socially responsible portfolios. The independent variables are similar to those in equation (5) except that ߙ௜ represents the differential excess performance between socially responsible and non-socially responsible portfolios.

4. Results and discussion 4.1. Descriptive statistics

MA

NU

Table 3 contains the descriptive statistics of monthly returns for the 39 portfolios composing our panel from January 2008 - December 2011. They are grouped in three sub groups according to their ESG engagement or controversies’ levels. Insert table 3

AC CE P

TE

D

These basic statistics suggest that, even after adjusting for volatility, portfolios with significant level of ESG engagement or controversies’ display more important excess return than the neutral portfolios with no engagement or no controversies. Globally, partially engaged portfolios (panel C) have a lower average Sharpe ratio than non-engaged portfolios (panel A) (0.019 against 0.051, respectively), both being lower than significantly engaged portfolios (panel E, with 0.137). On the opposite, partially controversial portfolios (panel D) have a higher average Sharpe ratio than non-controversial portfolios (panel B) (0.09 against 0.061, respectively), both being lower than significantly controversial portfolios (panel F, with 0.11). The last columns of Table 1 (Skewness and Kurtosis p-values and their related χ² test statistic) indicate no deviation from normal distribution for a large majority (27 out of 39) of the return series. Overall, our primary results suggest that the portfolios’ excess returns adjusted for total risk (standard deviation) are in not differenced in function on the distinction between engagement and controversies but rather on the basis of their scoring intensity (i.e. their engagement or controversies’ level). This result confirms the relevance of the disaggregated scoring approach. Due to its excessively small size (7 stocks by January, 2008) we decide to exclude the portfolio representing firms engaged human rights (Panel C) from our performance analysis. The size of the remaining portfolios ranges from 33 to 230 stocks.

4.2. Performance analysis Table 4 reports the results of Jensen’s single-factor model, Smith and Tito’s ratio, Sharpe’s information ratio and eSDAR. We use the MSCI U.S. Islamic index as value-weighted market proxy. The results show significant positive alphas only for 5 (out of 39) portfolios. Interestingly, these positive alphas concern portfolios that can be identified as “socially responsible” as they refer to portfolios that are: not involved in diversity controversies (α=4.08, p<.1); partially engaged in governance (α=3.38, p<.1) and product (α=6.36, p<.1); and significantly engaged in diversity (α=3.84, p<.1) and environment (α=3.24, p<.05). The results from Smith and Tito’s ratio, Sharpe’s ratio and eSDAR confirm the superior performance of these portfolios from a risk (systematic and total)

ACCEPTED MANUSCRIPT

PT

adjusted perspective. Precisely, the lower systematic risk displayed by the portfolios partially engaged in governance and product, and significantly engaged in diversity and environment accentuates their performance on a risk adjusted basis. This result is not surprising since portfolios with significant scores (panel C to F) have smaller compositions (see Table 3) and are therefore less sensitive to systematic risk.

NU

SC

RI

We can interpret these results with regard the capacity of a firm to “identify, act on and profit from opportunity to improve stakeholder relationships through CSR” (Barnett, 2007), According to Barnett, the extent to which we can expect that CSR initiatives will impact firms’ performance depends on the operational implication that a firm’s stakeholders can derive from them. Using the concept of Stakeholders Influence Capacity (SIC) the author suggest that CSR considerations such as good governance practices or product quality are more likely to improve key stakeholders’ relationship and then reduce firms’ contracting costs (see Wicks et al, 1999). On the other side CSR considerations such as minority treatment or waste management are expected to have less impact on firms’ cash flows since it concerns stakeholders with lower influence capacity.

MA

The last rows of Table 4 present the results of the four-factor regressions for the five portfolios ranked based on firm aggregate ESG performance. None of the reported alpha estimates were significant. This result confirms the limits of a composite social performance measurement approach and confirms the conclusion of Galema et al. (2008) on the opposed effect across ESG domains.

D

Table 5 reports the results of the four-factor regressions.

TE

Insert table 5

AC CE P

The estimates of the four-factor regressions report significant alpha for five portfolios. Two confirm Jensen’s single factor original estimates: the portfolios partially engaged in governance (α=4.91%, p<.05) and significantly engaged in diversity (α=4.41%, p<.1). The remaining significant alphas are reported for three “socially irresponsible” portfolios. Precisely, the portfolios partially involved in community and human rights controversies and significantly involved in governance controversies outperformed their peer index (α =6.04%, α=5.54% and α=7.06% respectively, p<.05). Additionally, Table 5 report a tilt toward big stocks (with an SMB coefficient at -0.30, p<.01) for the portfolio partially involved in human rights controversies. This result suggests a reverse causality whereby big firms displaying good financial performance are more akin to neglect the bad reputation’s effect associated with human rights controversies considering them as marginal. The results of this four-factor model suggest that three of the originally significant alphas are driven by investment style effects. Beyond the effects attributed to ESG domains’ heterogeneity, results suggest that ESG scoring intensity moderate the impact of ESG performance on portfolios’ return. Namely, considering firms that are significantly engaged (involved) in ESG initiatives (controversies) may lead to superior performance as compared to partial engagement (involvement). The significant alphas reported for the portfolios partially engaged in governance strengths and significantly involved in governance concerns seem contradictory. When examining the indicators considered by KLD for governance domain, we find that good governance is measured solely by firms' declared support of ESG public policies and by the quality of their social reporting15. On the opposite, bad governance is identified by the severity of disputes in relation to executive compensations, misconduct such as bribery, tax evasion, insider trading and accounting irregularities. This paradox 15

The quality assessment of reports is based on compliance with the Global Reporting Initiative.

ACCEPTED MANUSCRIPT

PT

and the positive alpha reported for portfolios partially involved in community and human rights concerns suggest that firms with good financial performance may intentionally decide to conceal their ESG concerns with a proactive CSR communication strategy. This “greenwashing” phenomenon was pointed out by many authors such as Parguel et al. (2011) who argue about the need to review current social rating methods to avoid measurement bias.

NU

SC

RI

The style related coefficients reported by the four-factor regressions show the presence of a size effect. The significant coefficients reported for SMB factor across neutral portfolios suggest that small stocks are more represented in portfolios that display neither strengths nor concerns. This result is in line with previous findings showing that small firms are less likely to receive ESG scorings as compared to big ones (Dravenstott & Chieffe, 2011). The significant negative loadings reported for SMB factor for portfolios partially controversial in human rights and significantly controversial in governance and environment (-0.30, p<.1; -0.24, p<.5 and -0.23, p<.1, respectively) show that large (shariahcompliant) stocks are more concerned by human rights, governance and environmental issues.

MA

The differenced portfolio treatment helps reduce the dimensionality of our panel in addition to indicating differences in factor exposure. Thus, we were able to estimate the relative performance of ESG strengths engagement and ESG concerns disengagement. In our perspective both strategies are to be considered “socially responsible”. Insert table 6

TE

D

As shown in Table 5, the return difference between engaged and not engaged portfolios in terms of governance is significantly positive (α=7.06%, p<.1). On the other hand, the results report negative return differences between not-controversial and controversial portfolios in community and human rights issues (α=-6.74% and α=-6.40% respectively, p<.1).

AC CE P

These results suggest that an SRI engagement strategy targeting good governance impact positively the performance of Islamic portfolios while an SRI disengagement strategy targeting community and human rights concerns impairs Islamic portfolios’ returns. This latter finding confirms previous works showing that the presence of socially irresponsible firms in a portfolio enhance its performance (Dravenstott & Chieffe, 2011).

4.3. Robustness test: the industry-adjusted seven-factor model Kurtz and Di Bartolomeo (2011) provide evidence that sector exposure substantially drive SRI portfolio returns. Dravenstott and Chieffe (2011) point out that Energy and Utility companies are more likely to be considered “irresponsible” using an aggregate scoring approach, while Financial and IT companies are more likely to be considered “responsible”. Since Islamic portfolios are more oriented toward certain specific sectors, we deemed it essential to control for the presence of industry effect in our performance models. Accordingly, we test whether regressions’ estimates change after controlling for industry effect. We used a similar approach to that adopted by Jones and Shanken (2005) and previously applied to SRI portfolios by Geczy et al. (2003) and Derwall et al. (2005). This involves the construction of a factor model composed of the four investment style regressors and three industry factors, orthogonal to the primary factor. To derive these regressors, we performed a principalcomponents analysis on the portion of Fama and French's excess industry-sorted portfolio returns that

ACCEPTED MANUSCRIPT cannot be explained by the four-factor model (i.e., the model's intercept and residual series)16. In turn, we retained the first three components, which captured most remaining industry return variations and added these to the four-factor model. The resultant model takes the following form: (7)

PT

ܴ௜,௧ – ܴ௙,௧ = ߙ௜ + ߚ௜ (ܴ௠,௧ − ܴ௙,௧ ) + ‫ݏ‬௜ ܵ‫ܤܯ‬௧ + ℎ௜ ‫ܮܯܪ‬௧ + ݉௜ ‫ܯܱܯ‬௧ + ‫݌‬௜ ‫ܲܫ‬ଵିଷ,௧ + ߝ௜,௧

Insert table 7

RI

where ‫ܲܫ‬ଵିଷ,௧ represents three factors (principal components) that capture industry effects. The results of this industry-adjusted regression provide robust, unbiased estimates (Table 7)17.

NU

SC

The loadings recorded for industry-adjustment factors cannot be interpreted with respect to specific industry exposure but provide evidence of industry effects as shown by coefficients’ significance (i.e. excepted for Panel F portfolio, the results report at least one significant coefficients for the three industry-adjustment factors). More importantly, the estimates confirm the robustness of the original alphas with a slight change in the significance level of the alpha referring to community controversies disengagement strategy.

5. Conclusion

TE

D

MA

From an investment style perspective, the results presented in Table 7 show that a governance engagement strategy tilts Islamic investments toward big stocks while a community and human right concerns disengagement strategy intensifies the original small cap effects produced by Islamic screenings.

AC CE P

Islamic Equity investment is ready for the next step in its development. The current practice of negative screening in Islamic investing is the cornerstone of Islamic equity investments. Now, in order to expand the equity asset class of Islamic investments and bring Islamic investments closer to its social tenets, scholars and practitioners argue for the need to include a positive approach promoting the social responsibility of businesses. Our study intended to investigate the financial cost of converging Islamic investment toward the adoption of positive practices (integrating in the portfolios the best companies in terms of environment, social and governance issues) currently observed in SRI. To do so, we used KLD’s ESG ratings and conducted an experimental analysis using a panel of selfconstructed portfolios formed from an existing shariah-compliant stocks universe. We set out to investigate the combined effect of ESG and Islamic screenings using a set of traditional and more precise performance measures. In addition to the four-factor model developed by Fama and French (1993) and Carhart (1997), we used a robust industry-adjusted seven-factor model. Our results point to the presence of varying and confounding effects attributed to the different type of ESG scoring dimensions, domains and intensities. In particular, results of Jensen’s single factor model provide substantial conclusive evidence on the outperformance of socially responsible Islamic portfolios expressing partial engagement in governance and product, and significant engagement in diversity and environment, as compared to their Islamic peer index during the 2007-2011 periods. The 16

We used seven industry-sorted portfolios: consumer, manufacturing, energy, high-tech, telecommunications, shops, utilities, and others. The exhaustive industry-sorted portfolios list composed by Fama & French was not appropriate for our study due to the limited number of sectors composing our Shariah-compliant universe. 17 The results reported in Table 6 are limited to the eight portfolios that exhibited significant alphas in the original four-factor model. The remaining results can be submitted upon request.

ACCEPTED MANUSCRIPT

PT

results of the four-factor and seven-factor models confirm the persistence of the outperformance of portfolios expressing partial engagement in governance and significant engagement in diversity after controlling for investment style and industry effect. Our study confirms the argument of Galema et al. (2008) stating that a composite ESG score screening approach actually produce compound effects that obliterate SRI portfolios’ outperformance.

SC

RI

The results of the difference portfolio analysis provide additional insights. Precisely, when comparing the relative performance of socially responsible Islamic portfolios as compared to their non-socially responsible Islamic counterparts, we find that outperformance only holds for governance engagement. Precisely, we find that an inclusive strategy promoting good governance among shariah-compliant stocks provides superior return while it may potentially lead to negative performance in the case of an exclusive strategy discriminating firms with human rights and community controversies.

AC CE P

TE

D

MA

NU

Globally, this study provides empirical evidence that incorporating ESG criteria into an Islamic investment process does not impair portfolio performance and provides partial supporting evidence for the good governance premium (Derwall, et al., 2005; Galema, et al., 2008). From a practical perspective, this pioneering study offers Islamic investors an avenue to promote corporate social responsibility without fearing any financial cost. As demonstrated by previous works (Core, Guay, & Rusticus, 2006; Gompers, Ishii, & Metrick, 2003) our study further confirms that Islamic portfolio managers can take advantage of the governance premium in their portfolio choices. Furthermore, and from a market efficiency perspective, as Islamic screening favours smaller companies that are less liquid than companies with higher market capitalization, the adoption of ESG standards by these firms should increase transparency, mitigating information asymmetry and improve their liquidity (Chung, Elder, & Kim, 2010).

ACCEPTED MANUSCRIPT References

PT

Abderrezak, F. (2008). The Performance of Islamic Equity Funds: A Comparison to Conventional, Islamic and Ethical Benchmarks. Failaka, available at: www.failaka.com/downloads/AbderrezakF_Perf_IslamicEquityFunds.pdf (accessed September, 2013).

RI

Abdullah, F., Hassan, T., & Mohamad, S. (2007). Investigation of performance of Malaysian Islamic unit trust funds: Comparison with conventional unit trust funds. Managerial Finance, 33, 142153.

SC

Ahmed, O. B. (2001). Islamic equity funds: The mode of resource mobilization and placement. Jeddah, Islamic Research and Training Institute, Islamic Development Bank.

NU

Al-Damkhi, A. M. (2008). Environmental ethics in Islam: principles, violations, and future perspectives. International Journal of Environmental Studies, 65, 11-31. Barnett, M. L. (2007). Stakeholder influence capacity and the variability of financial returns to corporate social responsibility. Academy of Management Review, 32, 794-816.

MA

Barnett, M. L., & Salomon, R. M. (2011). Does it pay to be really good? addressing the shape of the relationship between social and financial performance. Strategic Management Journal, 33, 1304-1320.

D

Bauer, R., Koedijk, K., & Otten, R. (2005). International Evidence on Ethical Mutual Fund Performance and Investment Style. Journal of Banking and Finance, 29, 1751-1767.

TE

Beekun, R. I., & Badawi, J. A. (2005). Balancing Ethical Responsibility among Multiple Organizational Stakeholders: The Islamic Perspective. Journal of Business Ethics, 60, 131145.

AC CE P

Brammer, S., Brooks, C., & Pavelin, S. (2006). Corporate Social Performance and Stock Returns: UK Evidence from Disaggregate Measures. Financial Management (Blackwell Publishing Limited), 35, 97-116. Brammer, S., Williams, G., & Zinkin, J. (2007). Religion and attitudes to corporate social responsibility in a large cross-country sample. Journal of Business Ethics, 71, 229-243. Brzeszcsynski, J., & McIntosh, G. (2013). Performance of Portfolios Composed of British SRI Stocks. Journal of Business Ethics, 1-28. Carhart, M. M. (1997). On Persistence in Mutual Fund Performance. Journal of Finance, 52, 57-82. Choudhury, M. A. (2010). Islamic Perspective of Socioeconomic Development. Journal of Islamic Economics, Banking and Finance, 6, 37-54. Chung, K. H., Elder, J., & Kim, J.-C. (2010). Corporate Governance and Liquidity. Journal of Financial & Quantitative Analysis, 45, 265-291. Core, J. E., Guay, W. R., & Rusticus, T. O. (2006). Does Weak Governance Cause Weak Stock Returns? An Examination of Firm Operating Performance and Investors' Expectations. Journal of Finance, 61, 655-687. Derigs, U., & Marzban, S. (2008). Review and analysis of current Shariah-compliant equity screening practices. International Journal of Islamic and Middle Eastern Finance and Management, 1, 285–303.

ACCEPTED MANUSCRIPT Derwall, J., Guenster, N., Bauer, R., & Koedijk, K. (2005). The Eco-Efficiency Premium Puzzle. Financial Analysts Journal, 61, 51-63.

PT

Dhaliwal, D. S., Li, O. Z., Tsang, A., & Yang., Y. G. (2011). Voluntary nonfinancial disclosure and the cost of equity capital: The initiation of corporate social responsibility reporting. The Accounting Review, 86, 59-100. Dravenstott, J., & Chieffe, N. (2011). Corporate Social Responsibility: Should I Invest for It or against It? Journal of Investing, 20, 108-117.

SC

RI

Dusuki, A. W. (2008). What Does Islam Say about Corporate Social Responsibility? Review of Islamic Economics, 12, 5-28. Dusuki, A. W., & Abdullah, N. I. (2007). Maqasid al-shariah, maslahah and corporate social responsibility. The American Journal of Islamic Social Sciences, 24, 25-45.

NU

Elfakhani, S., & Hassan, M. K. (2005). Performance of Islamic mutual funds. In Economic Research Forum, 12th Annual Conference (pp. 1-33).

MA

Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33, 3-56. Galema, R., Plantinga, A., & Scholtens, B. (2008). The Stocks at Stake: Return and Risk in Socially Responsible Investment. Journal of Banking and Finance, 32, 2646-2654.

TE

D

Geczy, C., Stambaugh, R. F., & Levin, D. (2003). Investing in Socially Responsible Mutual Funds. In: Wharton School, Université de Pennsylvanie. Girard, E., & Hassan, M. K. (2008). Is There a Cost to Faith-Based Investing: Evidence from FTSE Islamic Indices. Journal of Investing.

AC CE P

Gompers, P., Ishii, J., & Metrick, A. (2003). Corporate Governance and Equity Prices. The Quarterly Journal of Economics, 118, 107-156. Hakim, S., & Rashidian, M. (2004). Risk and Return of Islamic Stock Market Indexes. In International Seminar of Non-Bank Financial Institutions: Islamic Alternatives. Kuala Lumpur, Malaysia. Hayat, R., & Kraeussl, R. (2011). Risk and return characteristics of Islamic equity funds. Emerging Markets Review, 12, 189-203. Hayat, U. (2013). Islamic Finance and Socially Responsible Investing. In CFA Institute Magazine (pp. 8): EMEA VOICE Hoepner, A. G. F., Rammal, H. G., & Rezec, M. (2011). Islamic Mutual Funds' Financial Performance and International Investment Style: Evidence from 20 Countries. European Journal of Finance, 17, 829-850. Hussein, K., & Omran, M. (2005). Ethical Investment Revisited: Evidence from Dow Jones Islamic Indexes. The Journal of Investing, 105-124. Jawadi, F., Jawadi, N., & Louhichi, W. (2014). Conventional and Islamic stock price performance: An empirical investigation. International Economics, 137, 73-87. Jegadeesh, N., & Titman, S. (1993). Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency. Journal of Finance, 48, 65-91. Jones, C. S., & Shanken, J. (2005). Mutual Fund Performance with Learning across Funds. Journal of Financial Economics, 78, 507-552.

ACCEPTED MANUSCRIPT Kempf, A., & Osthoff, P. (2007). The Effect of Socially Responsible Investing on Portfolio Performance. European Financial Management, 13, 908-922. Kurtz, L., & Dibartolomeo, D. A. N. (2011). The Long-Term Performance of a Social Investment Universe. Journal of Investing, 20, 95-102.

PT

Markowitz, H. (1952). Portfolio Selection. The Journal of Finance, 7, 77-91.

RI

Merdad, H., Hassan, M. K., & Alhenawi, Y. (2010). Islamic Versus Conventional Mutual Funds Performance in Saudi Arabia: A Case Study. Journal of King Abdulaziz University: Islamic Economics, 23.

SC

Miglietta, F., & Forte, G. (2007). Islamic mutual funds as a faith-based funds in a socially responsible context. In. Milan: Bocconi University.

NU

Orlitzky, M., Schmidt, F. L., & Rynes, S. L. (2003). Corporate Social and Financial Performance: A Meta-analysis. Organization Studies (01708406), 24, 403-441.

MA

Renneboog, L., Ter Horst, J., & Zhang, C. (2008). Socially responsible investments: Institutional aspects, performance, and investor behavior. Journal of Banking & Finance, 32, 1723-1742. Rice, G. (1999). Islamic ethics and the implications for business. Journal of Business Ethics, 18, 345358.

D

Rice, G. (2006). Pro-environmental Behavior in Egypt: Is there a Role for Islamic Environmental Ethics? Journal of Business Ethics, 65, 373-390.

TE

Sairally, S. (2007). Evaluating the 'social responsibility' of Islamic finance: learning from the experiences of socially responsible investment funds. ISLAMIC ECONOMICS AND FINANCE, 279.

AC CE P

Sharfman, M. (1996). The Construct Validity of the Kinder, Lydenberg & Domini Social Performance Ratings Data. Journal of Business Ethics, 15, 287-296. Sharpe, W. F. (1994). The sharpe ratio. The journal of portfolio management, 21, 49-58. Smith, K. V., & Tito, D. A. (1969). Risk-return measures of ex post portfolio performance. Journal of financial and quantitative analysis, 449-471. Stanwick, P. A., & Stanwick, S. D. (1998). The Relationship Between Corporate Social Performance, and Organizational Size, Financial Performance, and Environmental Performance: An Empirical Examination. Journal of Business Ethics, 17, 195-204. Statman, M. (2000). Socially Responsible Mutual Funds. Financial Analysts Journal, 56, 30. Statman, M., & Glushkov, D. (2009). The wages of social responsibility. Financial Analysts Journal, 33-46. Ussif. (2014). Report on US Sustainable, Responsible, and Impact Investing Trends 2014, USSIF. Waddock, S. A., & Graves, S. B. (1997). The corporate social performance. Strategic Management Journal, 8, 303-319. Walkshaeusl, C., & Lobe, S. (2012). Islamic Investing. Review of Financial Economics, 21, 53-62. Warde, I. (2010). Islamic finance in the global economy: Edinburgh University Press. Wicks, A. C., Berman, S. L. & Jones, T. M. (1999). The structure of optimal trust: Moral and strategic implications. Academy of Management Review, 24, 99-116.

ACCEPTED MANUSCRIPT Williams, G., & Zinkin, J. (2010). Islam and CSR: A Study of the Compatibility Between the Tenets of Islam and the UN Global Compact. Journal of Business Ethics, 91, 519-533.

AC CE P

TE

D

MA

NU

SC

RI

PT

Wilson, R. (2001). The economics of mutual funds: An Islamic approach. University of Durham, 1-25.

ACCEPTED MANUSCRIPT Tables

Table 1 – Islamic SRI Portfolios' Panel Structure* ESG Controversies (Concerns)

Partial (score=1)

Significant (score>1)

Community





×

Governance





×

Diversity







Employee





Environment





Human Rights





Product





No (score=0)

Partial (score=1)

RI

No (score=0)

PT

ESG Engagement (Strengths)

Significant (score>1)



×











×

















×





×

×





×

MA

NU

SC



AC CE P

TE

D

* The × signifies that no portfolio was constructed for this category, due to the absence of scores. The missing portfolios are due to the absence of firms with the identified scoring (for e.g. we find no firms scoring above 1 for Community Strengths criteria).

Table 2 - Cross section correlations between the different ESG domain ratings

Governance

Positive ratings or Strengths

Community

Diversity

Employee

Environment

Human Rights

Governance

1

Community

0.56***

1

Diversity

0.53***

0.56***

1

Employee

0.43***

0.37***

0.50***

1

Environment

0.63***

0.46***

0.51***

0.38***

1

Human Rights

0.22***

0.25***

0.31***

0.27***

0.18***

1

Product

0.32***

0.35***

0.32***

0.31***

0.23***

0.18***

Product

1

Negative ratings or Concerns Governance

Community

Diversity

Employee

Environment

Human

Product

ACCEPTED MANUSCRIPT Rights 1

Community

-0.03

1

Diversity

0.03

-0.05

1

Employee

0.04

0.15**

-0.05

1

Environment

-0.06

0.56***

-0.07

0.26***

Human Rights

0.06

0.16**

-0.16***

0.19***

0.30***

1

0.17***

0.20***

-0.1

0.15**

0.03

0.15**

RI

1

SC

Product

PT

Governance

1

NU

Note: The test was conducted over the ESG ratings of 227 firms part of MSCI ISLAMIC index composition by December 2010. *10% significance; **5% significance; ***1% significance

Table 3 – Descriptive Statistics of self-composed Islamic SRI portfolios, Jan. 2008 - Dec. 2011 Std. Dev.

Sharpe Ratio

Min.

Max.

Pr Pr (Skewness) (Kurtosis)

χ²

Nb of firms

MA

Mean Group 1: Neutral portfolios

Panel A : Not engaged in ESG initiatives (strengths=0) 22

Governance

-0.31

23.35

Diversity

1.99

Employee

2.67

Human Rights Product

-16.47

17.58

0.32

0.17

3.03

164

-0.01

-18.95

17.07

0.17

0.19

3.82

175

-19.29

2.68

21.16

0.13

-18.46

15.48

0.1

0.09

5.38*

148

0.34

17.76

0.02

-15.06

11.04

0.08

0.26

4.48

230

-0.67

18.95

-0.04

-15.25

13.68

0.19

0.27

3.09

203

26.22

0.08

-19.48

18.11

0.19

0.54

2.22

81

20.59

0.13

13.95

0.04

0.05

7.14

116

AC CE P

Environment

0.05

D

1.11

TE

Community

Panel B: Not involved in ESG controversies (concerns=0) Community

1.12

18.74

0.06

-15.58

12.02

0.1

0.27

4.14

177

Governance

2.54

23.91

0.1

-21.33

15.52

0.04

0.1

6.22**

70

Diversity

3.53

18.99

0.18

-17.45

10

0.02

0.12

6.93**

173

Employee

0.55

17.78

0.03

-15.28

10.91

0.16

0.29

3.27

78

Environment

0.94

18.88

0.05

-17.7

10.58

0.03

0.09

6.63**

135

Human Rights

-0.43

18.19

-0.02

-14.63

11.74

0.12

0.32

3.65

193

Product

0.67

21.88

0.03

-18.91

14.56

0.07

0.22

4.78*

140

9.25

0.05

0.21

5.22*

84

Group 2: Partially socially responsible or irresponsible portfolios Panel C : Engaged in ESG initiatives (strengths=1) Community

1.6

16.29

0.1

-14.17

ACCEPTED MANUSCRIPT 3.08

15.24

0.2

-12.88

8.12

0.03

0.2

6.12**

59

Diversity

-1.87

25.55

-0.07

-25.8

18.58

0.01

0.01

10.39***

52

Employee

0.06

22.1

0

-17.96

13.62

0.14

0.38

3.1

69

Environment

-3.54

19.9

-0.18

-14.3

16.01

0.4

0.16

2.81

49

Human Rights

-8.92

35.44

-0.26

-57.45

12.48

Product

6.01

17.43

0.34

-15.22

0

0

47.23***

7

10.56

0.04

0.14

5.97**

33

8.89

RI

PT

Governance

0.06

0.45

4.22

60

10.37

0.07

0.37

4.3

138

14.13

0.52

0.31

1.53

64

Panel D: Involved in ESG controversies (concerns=1) 2.64

17.92

0.15

-14.94

Governance

2.64

18.91

0.14

-15.99

Diversity

-0.65

17.54

-0.04

-11.16

Employee

0.68

19.36

0.03

-15.38

13.78

0.15

0.29

3.41

105

Environment

2.11

18.6

0.11

-16.4

11.21

0.04

0.2

5.57*

50

Human Rights

2.59

18.55

0.14

-15.97

10.76

0.08

0.33

4.19

44

Product

1.56

15.38

0.1

-12.16

8.31

0.14

0.52

2.69

97

MA

NU

SC

Community

D

Group 3: Significantly socially responsible or irresponsible portfolios

3.34

Employee

0.76

Environment

15.39

0.21

-11.86

8.66

0.1

0.53

3.28

104

17.46

0.04

-12.27

14.56

0.55

0.17

2.33

52

16.68

0.16

-13.35

10.69

0.21

0.35

2.62

40

AC CE P

Diversity

TE

Panel E: Significantly engaged in ESG initiatives (strengths>1)

2.72

Panel F: Significantly involved in ESG controversies (concerns>1) Governance Employee Environment

4.37

16.19

0.26

-12.6

8.39

0.09

0.45

3.66

38

0.94

17.31

0.05

-15.08

9.06

0.06

0.26

4.77*

54

0.43

18.1

0.02

-13.24

12.33

0.32

0.63

1.29

52

Portfolios based on aggregated scoring Worst ESG score

-2.07

26.42

-0.08

-20.77

20.2

0.29

0.14

3.58

47

Bad ESG score

0.43

20.35

0.02

-13.88

13.56

0.36

0.96

0.89

48

Mid ESG score

0.26

20.34

0.01

-17.08

14.35

0.11

0.18

4.44

47

Good ESG score

1.81

19.07

0.09

-15.45

9.43

0.08

0.39

4.07

47

Best ESG score

2.56

15.57

0.16

-13.62

8.15

0.05

0.24

5.17*

49

Note: This table presents the summary statistics of self-composed Islamic SRI portfolios for 47 monthly observations. Mean return, standard deviation, and Sharpe ratio, defined as the mean return to the standard deviation of the returns, are annualized. Sample period was January 2008–December 2011.We provide Skewness and Kurtosis p-values and their related χ² stat as well as portfolios’ size by Jan. 2008. * 10% significance. ** 5% significance. *** 1% significance.

ACCEPTED MANUSCRIPT Table 4 – Jensen’s single-factor model, Smith and Tito's ratio, Sharpe’s information ratio and eSDAR t-stat

α

Rm-Rf

t-stat

T

S

sSDAR

1.50

0.24

1.56

0.38

0.03

0.46

Group 1: Neutral portfolios

1.80

0.75

1.20***

15.10

Governance

0.48

0.15

1.28***

19.80

Diversity

2.76

0.64

1.38***

17.70

2.00

0.21

2.06

Employee

3.24

0.92

1.11***

16.30

2.92

0.47

2.93

Environment

3.36

1.20

1.16***

19.20

2.90

0.52

2.88

Human Rights

0.96

0.66

1.00***

137.00

0.96

0.35

0.93

Product

0.00

-0.03

1.05***

32.10

0.00

-0.02

-0.01

1.01***

33.40

1.66

0.28

1.66

1.29***

19.40

2.51

0.34

2.57

NU

1.68

0.63

Governance

3.24

0.87

Diversity

4.08*

1.68

1.04***

20.70

3.92

0.77

3.89

Employee

1.08

0.46

0.95***

29.20

1.14

0.19

1.13

Environment

1.56

0.54

1.02***

19.80

1.53

0.25

1.49

Human Rights

0.12

0.10

1.00***

42.80

0.12

0.04

0.18

0.53

1.21***

34.10

1.09

0.20

1.21

AC CE P

Product

TE

Community

D

MA

Panel B: Not involved in ESG controversies (concerns=0)

RI

Community

SC

PT

Panel A : Not engaged in ESG initiatives (strengths=0)

1.32

Group 2: Partially socially responsible or irresponsible portfolios Panel C : Engaged in ESG initiatives (strengths=1) Community

2.16

1.21

0.90***

36.10

2.40

0.51

2.28

3.48*

1.72

0.83***

21.40

4.19

0.71

4.07

Diversity

-1.08

-0.29

1.37***

12.40

-0.79

-0.12

-0.57

Employee

0.72

0.24

1.20***

38.90

0.60

0.08

0.72

Environment

-3.00

-0.89

1.06***

14.00

-2.83

-0.43

-2.49

Human Rights

-8.76

-0.56

1.02***

13.70

-8.59

-0.29

-3.63

6.36**

2.18

0.87***

10.80

7.31

0.75

6.65

Governance

Product

Panel D: Involved in ESG controversies (concerns=1) Community

3.12

1.14

0.98***

20.10

3.18

0.62

3.18

Governance

3.24

1.02

1.02***

29.40

3.18

0.54

3.07

Diversity

-0.12

-0.03

0.94***

10.70

-0.13

-0.01

-0.07

ACCEPTED MANUSCRIPT 1.32

0.63

1.08***

26.90

1.22

0.32

1.24

Environment

2.64

1.52

1.02***

35.20

2.59

0.57

2.60

Human Rights

3.12

1.00

1.01***

30.40

3.09

0.57

3.06

Product

2.04

0.97

0.84***

24.20

2.43

0.42

2.31

PT

Employee

Group 3: Significantly socially responsible or irresponsible portfolios

RI

Panel E: Significantly engaged in ESG initiatives (strengths>1) 3.84*

1.82

0.84***

22.50

4.57

0.73

4.34

Employee

1.32

0.53

0.95***

12.00

1.39

0.28

1.35

Environment

3.24*

1.66

0.91***

24.70

0.67

3.43

SC

Diversity

3.56

Governance

4.80

1.64

0.86***

18.30

5.58

0.77

5.30

Employee

1.44

0.54

0.94***

22.20

1.53

0.29

1.53

Environment

0.96

0.40

0.99***

26.70

0.97

0.21

1.01

MA

-1.32

-0.28

1.40***

11.80

-0.94

-0.13

-0.65

Bad ESG score

1.08

0.24

1.06***

17.30

1.02

0.13

1.01

Mid ESG score

0.96

0.44

1.12***

30.10

0.86

0.15

0.87

Good ESG score

2.40

0.96

1.03***

16.50

2.33

0.39

2.29

Best ESG score

3.00

0.99

0.75***

10.80

4.00

0.34

3.42

AC CE P

TE

Worst ESG score

D

Portfolios based on aggregated scoring

NU

Panel F: Significantly involved in ESG controversies (concerns>1)

Note: This table reports the results of Jensen’s alpha (α), Smith and Tito's ratio (T), Sharpe’s information ratio (S) and the excess standard-deviation-adjusted return (eSDAR). Alphas are annualized. T-statistics are derived from Newey– West heteroscedasticity and autocorrelation-consistent standard errors. The sample period was from January 2008– December 2011.* 10% significance. ** 5% significance. *** 1% significance.

ACCEPTED MANUSCRIPT

Table 5 – Results of the Four-factor regression model α

t stat

SMB

t stat

HML

t stat

MOM

t stat

Rm-Rf

t stat

Adj. R²

1.17***

15.00

0.93

0.36

0.15

1.30

0.06

0.55

0.01

Governance

-2.02

-0.68

0.30***

3.44

-0.08

-0.68

-0.06

-2.42 1.23***

19.70

0.94

Diversity

-0.12

-0.02

0.40**

1.69

-0.09

-0.59

-0.01

-0.24 1.32***

13.30

0.86

Employee

0.00

-0.01

0.41***

2.73

-0.13

-1.05

-0.06

-1.33 1.04***

16.80

0.92

Environment

0.96

0.33

0.29***

4.05

-0.11

-1.09

-0.05

-2.70 1.12***

18.70

0.94

Human Rights

1.21

0.69

-0.03

-0.89

0.05

0.87

0.00

0.44

0.99***

117.00

0.98

Product

0.36

0.18

-0.04

-0.74

0.02

0.24

0.03

2.30

1.06***

32.40

0.95

MA

0.27

RI

0.96

NU

Community

SC

Panel A : Not engaged in ESG initiatives (strengths=0)

PT

Group 1: Neutral portfolios

Panel B: Not involved in ESG controversies (concerns=0) -1.07

-0.51

0.41***

Governance

0.36

0.10

Diversity

3.29

1.25

Employee

-0.48

-0.18

Environment

-1.31

Product

0.02

0.27

-0.03

-0.68 0.92***

30.90

0.92

0.35***

2.04

-0.07

-0.62

-0.08

-1.80 1.22***

14.30

0.91

0.13

0.88

0.05

0.63

-0.02

-0.79 1.00***

17.00

0.92

0.17

1.31

-0.04

-0.56

-0.05

-0.94 0.91***

20.80

0.89

-0.59

0.37***

4.75

0.01

0.13

-0.08** -3.16 0.93***

16.80

0.94

-1.19

-0.88

0.21***

4.08

0.08

1.32

-0.03

-1.15 0.94***

40.00

0.96

1.09

0.40

0.04

0.64

0.08

0.90

-0.03

-0.87 1.18***

24.00

0.94

TE

AC CE P

Human Rights

3.30

D

Community

Group 2: Partially responsible or irresponsible portfolios Panel C : Engaged in ESG initiatives (strengths=1) Community

1.94

1.14

0.05

0.44

0.01

0.31

0.01

0.45

0.89***

30.20

0.94

4.91**

2.54

-0.10

-1.00

0.13

1.86

0.05

1.83

0.84***

24.90

0.93

Diversity

-4.24

-1.22

0.38**

2.60

-0.25*

-1.72

-0.04

-1.02 1.34***

12.50

0.90

Employee

-0.48

-0.16

0.22*

1.27

0.01

0.06

0.03

1.01

1.16***

19.40

0.90

Environment

-1.79

-0.53

-0.15

-1.18

-0.07

-0.64

0.06

2.13

1.12***

15.40

0.88

Human Rights

-11.04 -0.61

0.19

0.75

-0.53

-0.86

-0.06

-0.73 1.07***

10.80

0.21

0.19

1.36

0.14

1.31

-0.09*

-1.48 0.78***

9.61

0.80

Governance

Product

5.03

1.68

Panel D: Involved in ESG controversies (concerns=1) Community

6.04**

1.99

-0.33

-4.96

0.11

1.47

0.04

1.55

1.03***

23.80

0.94

Governance

1.81

0.63

0.27***

2.16

0.15*

2.12

-0.01

-0.20 0.94***

20.60

0.92

Diversity

0.12

0.04

0.06

0.52

0.04

0.51

0.09**

2.75

0.95***

12.30

0.90

Employee

1.57

0.84

0.02

0.38

0.15***

1.94

0.02

0.92

1.05***

26.70

0.97

Environment

2.43

1.01

0.06

0.75

0.06

0.86

-0.01

-0.60 1.00***

31.70

0.94

Human Rights

5.54**

1.61

-0.30***

-4.75

0.02

0.25

0.04

1.81

38.40

0.93

2.67

1.15

-0.06

-1.16

0.07

1.10

0.01

0.84***

22.90

0.92

2.15

RI

SC

ACCEPTED MANUSCRIPT

0.03

0.87

0.82***

25.20

0.91

0.40

0.03

1.68

0.96***

11.50

0.92

2.09

-0.02

-0.46 0.87***

19.70

0.93

0.81

0.05

2.22

0.90***

20.50

0.88

Group 3: Significantly responsible or irresponsible portfolios Panel E: Significantly engaged in ESG initiatives (strengths>1) 4.41*

2.06

-0.01

-0.09

0.14*

Employee

1.57

0.54

0.00

-0.01

0.04

Environment

3.29

1.83

0.04

0.55

0.16**

NU

Diversity

PT

Product

1.07***

0.42

Panel F: Significantly involved in ESG controversies (concerns>1) 7.06**

1.98

-0.24**

Employee

2.43

0.90

-0.14

Environment

3.04

1.10

-0.23***

-0.03

-0.50

0.01

0.43

0.98***

20.10

0.91

-4.14

0.06

0.85

0.05

3.42

1.04***

28.20

0.94

1.26

-0.28*

-1.75

0.00

-0.04 1.41***

15.40

0.86

Worst ESG score

-3.43

-0.77

Bad ESG score

3.66

0.77

-0.31**

-2.79

0.03

0.23

0.08

1.66

1.14***

20.10

0.86

Mid ESG score

1.33

0.63

-0.04

-0.34

0.00

0.00

0.02

0.82

1.13***

20.80

0.93

AC CE P

TE

0.24

0.07

-1.64

D

Portfolios based on aggregated scoring

-1.49

MA

Governance

Good ESG score

2.43

1.13

0.06

0.52

0.24***

3.33

-0.03

-0.98 0.96***

19.90

0.91

Best ESG score

1.21

0.43

0.23

1.60

0.10

1.19

-0.08

-1.21 0.67***

8.38

0.75

Note: This table reports the results from the four-factor regressions based on the following equation: R_(i,t) – R_(f,t) = α_i+ β_i (R_(m,t)− R_(f,t)) + s_iSMB_t+ h_iHML_t+ m_iMOM_t+ε_(i,t). The table displays the adjusted R², coefficients, and their respective t-stat for each regression. Portfolios are grouped into three categories. All alphas are annualized. Tstatistics are derived from Newey–West heteroscedasticity and autocorrelation-consistent standard errors (lag 1). The sample period was from January 2008–December 2011.* 10% significance. ** 5% significance. *** 1% significance.

ACCEPTED MANUSCRIPT Table 6 – Four-factor based Differenced Performance α

t stat

SMB

t stat

HML

t stat

MOM

t stat

Rm-Rf

t stat

Adj. R²

0.12

-0.28***

-2.62

0.30

Panel A: ESG initiatives engagement 0.96

0.30

-0.11

-0.60

-0.05

-0.43

0.01

Governance

7.06*

1.86

-0.39***

-2.77

0.22*

1.34

0.11

Diversity

-4.24

-1.32

-0.02

-0.16

-0.16

-1.50

-0.03

Employee

-0.48

-0.12

-0.19

-1.63

0.14

1.20

Environment

-2.73

-0.57

-0.43***

-2.79

0.04

0.27

Human Rights

-12.11

-0.62

0.22

0.77

-0.58

4.66

1.01

0.23

1.23

0.13

-0.09

-0.39***

-4.16

0.53

-0.69

0.02

0.19

-0.04

0.09**

1.49

0.12**

2.03

0.10

SC

RI

2.23

0.11**

3.06

0.00

0.01

0.24

-0.86

-0.06

-0.70

0.08

0.75

-0.07

0.81

-0.12

-1.66

-0.28***

-2.51

0.15

-1.09

-1.04

-0.10

-1.72

0.39

NU

Product

PT

Community

-0.07

0.54

-0.23*

-1.85

-0.07

-1.08

0.28***

3.50

0.29

0.38

0.01

0.06

-0.11**

-2.17

0.05

0.34

0.07

0.15

1.03

-0.19*

-1.98

-0.07

-1.33

-0.14**

-1.93

0.13

-1.17

0.31**

3.15

-0.06

-0.49

-0.07

-2.33

-0.07

-1.46

0.10

-6.40*

-1.75

0.51***

4.71

0.06

0.57

-0.07

-1.50

-0.13*

-2.57

0.29

-1.55

-0.39

0.11

0.90

0.00

0.03

-0.04

-0.65

0.33***

3.99

0.37

Panel B: ESG controversies disengagement -1.76

0.74***

Governance

-1.43

-0.36

0.08

Diversity

3.17

0.60

0.07

Employee

-2.02

-0.76

Environment

-3.66

Human Rights Product

5.04

MA

-6.74*

AC CE P

TE

D

Community

Note: This table displays the results of the multifactor regressions conducted on difference portfolios expressed by the following equation: R_(i,t,p) – R_(i,t,n) = α_i+ β_i (R_(m,t)− R_(f,t)) + s_i SMB_t+ h_i HML_t+ m_i MOM_t+ε_(i,t). The table displays the R², coefficients, and their respective t-stat for each regression. All alphas are annualized. T-statistics are derived from Newey–West heteroscedasticity and autocorrelation-consistent standard errors (lag 1). The sample period was from January 2008–December 2011.* 10% significance. ** 5% significance. *** 1% significance.

ACCEPTED MANUSCRIPT

t stat

SMB

t stat

HML

IP1

t stat

IP2

t stat

IP3

t stat

Adj. R²

SC

α

t stat

MOM

t stat

Rm-Rf

0.13*

1.86

0.04***

2.81

0.85***

33.3

0

0.04

0.30**

2.01

-0.45***

-3.25

0.94

2.18

1.01***

25.2

-0.20**

-2.4

0.02

0.1

0.11

0.62

0.94

2.56

1.05***

28.2

-0.21**

-2.53

-0.35*

-1.83

0.09

0.5

0.93

0.02

1.11

0.84***

29.8

0.12

0.97

0.41***

2.67

-0.50***

-2.75

0.94

0.8

0.04*

1.82

0.90***

22.6

-0.2

-1.57

0.18

0.7

-0.38

-1.47

0.89

2.47

-0.1

-1.56

Panel D: Involved in ESG controversies (concerns=1) 6.04*

1.96

-0.33***

-5.12

0.11

1.55

0.05**

Human Rights

5.54*

1.75

-0.29***

-4.66

0.03

0.34

0.04***

MA

Community

Diversity

4.41**

2.06

-0.02

-0.25

0.13**

2.15

1.99

-0.24**

-1.84

-0.41***

-3.8

0.21

1.31

0.10***

4.25

-0.35***

-4.38

0.22

0.9

0.74***

2.6

-0.57**

-2.47

0.51

Differenced performance Panel A: ESG initiatives engagement Governance

7.06**

1.94

0.07

AC

6.98**

CE

Panel F: Significantly involved in ESG controversies (concerns>1) Governance

PT ED

Group 3: Significantly socially responsible or irresponsible portfolios Panel E: Significantly engaged in ESG initiatives (strengths>1)

NU

Panel C : Engaged in ESG initiatives (strengths=1) 4.91**

t stat

RI

Group 2: Partially socially responsible or irresponsible portfolios

Governance

PT

Table 7 – Robutness test: Industry-Adjusted Seven-Factor Model

Panel B: ESG controversies disengagement Community

-6.74*

-1.80

0.74***

5.65

-0.1

-1.18

-0.08*

-1.67

-0.08

-1.25

0.22

1.54

0.19

0.6

-0.69**

-2.39

0.62

Human Rights

-6.40*

-1.89

0.50***

4.81

0.05

0.53

-0.08**

-2.48

-0.09

-1.66

0.29**

2.51

0.53**

2

-0.40**

-2.05

0.28

Note: The equation used for industry adjustment derives from the classic four-factor model. Additional industry factors are represented by the IP variable. The equation is modified as follows: R(it)−

ACCEPTED MANUSCRIPT

AC

CE

PT ED

MA

NU

SC

RI

PT

RF(t) = αi + βi[RM(t)− RF(t)]+ siSMB(t)+hiHML(t)+ miMOM(t)+θiIP1-3(t)+ ε(t). T-statistics (in italics) derive from Newey-West heteroskedasticity and autocorrelation-consistent standard errors. The sample period was from January 2008 - December 2011. 10% significance; **5% significance; ***1% significance.