ESG Reporting and Analysts’ Recommendations in GCC: The Moderation Role of Royal Family Directors

Abstract

This study examines whether financial analysts consider or incorporate the environmental, social and governance disclosures (thereafter ESG) in their recommendations. We then test whether royal family directors affect this relation. Using a dataset from six Gulf Cooperation Council (GCC) countries, we find evidence that analysts’ recommendations are influenced by ESG information. Further, we find the political connection negatively moderates the relationship between sell-side analysts’ recommendations and ESG. This suggests that financial analysts may assess the ESG disclosure in those firms with the political connection of royalty, in GCC countries, as superficial compliance rather than a genuine commitment. Our results are robust when subjected to endogeneity tests.

1. Introduction

Recent anecdotal evidence on the link between society, environment and business continues to highlight the extent to which analysts pay heed to environmental, social and governance (ESG) engagements by firms. This is not only in developed markets but also in emerging markets, such as those in the Gulf Cooperation Council (GCC) countries. The purpose of this study is to examine whether financial analysts react to and consider ESG disclosures in their recommendations in GCC countries. We then test whether the social and cultural players such as royal family directors that are politically connected, a prominent feature of GCC boards, affect this relationship (Halawi et al. 2008; Al-Hadi et al. 2016).

Financial analysts1 are important in the international business context, where information asymmetry between management and investors is high (Fung et al. 2016). The increasing interest by international investors in GCC equity markets, which provide tax haven opportunities and a strong return in capital markets (Bley and Saad 2012), signifies the important role played by financial analysts who gather and process information about companies and distribute this information through their stock recommendations. As knowledgeable experts who conduct research and provide intelligence on firms they follow, their stock recommendations help investors to make investment decisions through interpreting complex information and converting it to simple buy, hold and sell recommendations (Jegadeesh et al. 2004; Brown et al. 2015). They also play significant roles in mitigating agency problems and information asymmetry and have the ability to assess ESG reporting. In addition, the increasing numbers of fund managers and investors who allocate their investments to those firms with better ESG have encouraged financial analysts to consider ESG when issuing investment recommendations. Moreover, many big investment institutions have sustainability indices which are essential for such investors. We assume and theorize that the financial analysts with such expertise will serve investors by incorporating ESG into their valuation models and recommendations.

The value of the analysts’ stock recommendations comes from at least two sources. First, analysts are skilled in analyzing and synthesizing both private and public information from management and other sources to investors (Newton 2019). Second, they can gather a wide range of information unavailable to the investors, integrate the diverse information, and professionally assess the prospects of firm future cash flows (Ivković and Jegadeesh 2004). The summary judgment recommending “buy/hold/sell” is the investment opinion that analysts disseminate to investors regarding whether a given stock is worth buying or selling. In essence, the recommendation captures forward-looking information that helps investors gauge future cash flows and firm value (Luo et al. 2010).

Ioannou and Serafeim (2015) argue that the corporate social responsibility (CSR) activities taken by a firm may affect analysts’ recommendations through the following channels. First, CSR enhances value by improving a firm’s long-term financial performance. Changes in financial performance, therefore, may have direct impacts on analysts’ recommendations. Second, the substantial amount of funds invested by socially responsible and environmentally conscious investors in CSR-friendly firms might positively affect the stock prices of those firms, thus also affecting analysts’ recommendations. For instance, Dhaliwal et al. (2011) find that companies with higher CSR disclosures are more likely to attract institutional investors and analysts’ coverage.

The GCC is a political and economic union of six Arabic monarchical countries. It was established in 1981 and includes Saudi Arabia, Qatar, Oman, Kuwait, Bahrain and UAE. GCC countries have many similarities. They share the same religion, culture and political systems. They are monarchies and they have high oil and gas reservations, which makes them among the richest countries in the world. In recent decades, an influx of foreign capital and international investment institutions has been observed. The GCC region provides a high return on investments and it is considered a tax haven.

Further, GCC countries are characterized by royal families that are involved in different aspects: social life, business, state ruler ministers in many key ministries, owners, and directors on company boards (Al-Hadi et al. 2016; Kamrava et al. 2016). We theorize that the huge overlap of those royal families, in many power positions in general and as members of boards of directors in particular, may have a significant influence on the decisions process of these companies, especially with regard to ESG. The equity analysts may perceive this appearance pessimistically or optimistically. Thus, it will be interesting to study the impact of the royal family directors, especially in a unique characteristic area that cannot be found in any other area.

In this study, we introduce seven emerging-frontier markets from the GCC countries as a new unique setting and context to re-examine this association. In particular, this study investigates whether in developing countries the participants that monitor and channel the flow of ESG information in capital markets perceive and assess this as important information informing earnings expectations and valuing securities.

While a growing body of research examines the ESG issues in both developed and developing countries, the social-political environment in developing countries has been ignored. Our study responds to the call for more research looking into the contextual motivations and challenges specifically faced by developing countries (Islam and Deegan 2008; Belal et al. 2013; Tilt 2016). Our study also responds to the calls of the studies of Ioannou and Serafeim (2015); Eccles et al. (2011); and Yu (2011) for more research into ESG and analysts’ recommendations.

We focus on GCC countries because foreign investors have been attracted to GCC capital markets, especially after the economic reforms (Balli et al. 2011; Al-Hadi et al. 2015). For example, the listed companies in GCC capital markets have increased from 399 to 4682 between 2000 and 2017. Furthermore, many foreign investors’ entrance barriers have been removed. For instance, the “ease of doing business index” has increased in Saudi Arabia, from 15% in 2008 to 96% in 20152; in Oman, from 60% in 2008 to 77% in 2014; and in Bahrain, from 18% in 2008 to 66% in 2015. Therefore, it is reasonable to expect that investors will look for information intermediaries in the global capital markets such as analysts to help them assess corporate performance.

This paper contributes to the prior literature as follows. First, by answering two key questions: (1) Do the financial analysts incorporate and consider ESG in their recommendations? (2) What is the key role of royal family directors in the above-mentioned relationship? Second, we enrich the scant literature on non-financial information disclosure and its implications on the stock market players overall and in developing countries (i.e., the GCC markets). Most of the securities in the GCC capital markets rarely receive analyst coverage and analyst recommendations (Al-Ajmi and Kim 2012; Al-Hadi et al. 2015). This may create a shortage in information disclosure, which, in turn, leads to higher uncertainty regarding firm-specific information and information asymmetry. It is crucial to study whether such securities obtain benefits in terms of favorable recommendations from their ESG disclosures. There are very limited studies that investigate the impact of ESG disclosures on analysts’ behavior, thus, we provide new evidence from the GCC emerging markets.

We provide new evidence that the relationship between ESG and analysts’ recommendations is positive, and the existence of the royal family on the board negatively moderates the relationship between ESG and analysts’ recommendations. These results demonstrate that the politically connected firms of the royal family exacerbate the information asymmetry in the firms. These findings assume that the practical implication of emerging and GCC markets must consider political issues when making decisions.

The rest of this paper is organized as follows. Section 2 and Section 3 presents the background and GCC setting, literature review, and hypotheses development. In Section 4 and Section 5, we discuss the research methodology, and Section 6 highlights the key findings and results discussion. We present the conclusion of our study in Section 7.

2. Literature Review

2.1. Social Responsibility Policy, Environment, and Regulation in the GCC

Corporate social responsibility (CSR)3 regulation in GCC countries is considered in its infancy stage and still voluntary. There is no clear CSR regulation or policies in GCC governments and stock markets. There are, however, some initiatives released in some GCC countries. For instance, in 2008, the Saudi Arabian General Investment Authority (SAGIA) launched the Responsible Competitiveness Index (SARC) to rank more socially responsible firms. However, this is a voluntary adoption. Oman is the first country in the GCC that required firms to adopt a corporate governance code in 2003 (Al-Hadi et al. 2016), and the Omani Capital Market Authority (CMA) in several releases urges all joint stock companies and investment funds to adopt the Oman Social Responsibility Initiative, launched by the CMA Oman on November 2, 20094. “The company shall seek to exercise its role as good citizen and to mitigate any adverse impact of its activities on the national economy, community or environment at large”. The new code requires a CSR charter or code, a CSR strategy, and an annual report on CSR activities.

The Kingdom of Bahrain, in order to diversify its economy from being reliant only on oil, has promoted itself as an international banking center and encourages foreign investments. Many multi-national corporations, specifically banking corporations including banks and corporations, are encouraged to practice CSR and to adhere strictly to the regulations. The Kingdom of Bahrain believes that in order for CSR to be effective, it must be controlled, regulated, and standardized by the government. This is because, by nature, a corporation will always look after its own interests. Thus, the government established the Bahraini Association for Social Responsibility in 2011.

In July 2008, Qatar launched its 2030 national vision. The 2030 vision rests on four pillars: human development, social development, economic development, and environmental development, which are the crux of corporate social responsibility. Emir Tamim, Amir of Qatar has demanded hard work to help accomplish the CSR goals and vision and to advance the nation’s development5. To achieve this vision, the Ministry of Economy and Commerce (MEC) has launched a Qatari CSR index that takes into account similar international experiments based on relevant United Nations (UN)standards. In the UAE, there are also trends to make CSR disclosures by listed companies compulsory.

2.2. Overview of ESG Disclosure Theories

Numerous theories explain voluntary disclosure of ESG information (or sustainability reporting) including stakeholder theory, impression management theory, institutional theory, discretionary disclosure theory, and legitimacy theory. For, example, legitimacy theory provides some explanation of why firms adopt CSR. It asserts companies’ behavior in implementing and developing voluntary social and environmental disclosure of information in order to fulfill their social contract. This enables the recognition of their objectives and survival in a jumpy and turbulent environment. Suchman (1995) considers that “legitimacy is a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions”. Legitimacy theory suggests that firms legitimate themselves through various actions, including communication with relevant stakeholders. Sell-side analysts are of key importance for shareholders who really pay attention to firms’ annual reports and interpret the financial and non-financial information. This information is seen to be essential input in their research reports.

According to Schlenker (1980), impression management is the means by which to influence individuals. The success of influencing others through impression management depends on the audience's positive or negative perception (Gardner and Martinko 1998). We assume that the board members from the royal families are more likely to manage impression. Financial analysts are one of the main audiences in the stock market. Part of this impression may be through influencing social responsibility activities, which may be met negatively or positively by analysts.

3. Hypotheses Development

The financial analysts employed by brokerage firms, so-called “sell-side analysts," are commonly considered to be experts in investment analysis and security valuation. A large body of literature documents the significant role of security analysts as information intermediaries in capital markets (Bradshaw 2004; Healy and Palepu 2001; Ioannou and Serafeim 2015; Luo et al. 2015) and in influencing stock prices (Givoly and Lakonishok 1984; Stickel 1995; Givoly and Lakonishok 1984; Stickel 1995). As mentioned earlier, sell-side analysts issue recommendations on securities, which are typically phrased as buy, sell, or hold (Ioannou and Serafeim 2015).

Previous literature also investigates several determinants of analyst recommendations. Luo et al. (2015) find international evidence that analysts’ recommendations mediate the relationship between CSR and firms’ stock returns by reducing the strength of the effect between them. They explain that, as skilled industry experts, security analysts are able to obtain access to private information that is not readily accessible to general investors and so are better able to assess the value relevance of a firm’s corporate social performance information. Thus, the analysts’ recommendations act as an informational pathway through which corporate social performance affects corporate financial performance. Yu (2011) finds that analysts tend to issue favourable recommendations for companies with better corporate governance mechanisms in the emerging markets. Dhaliwal et al. (2012) investigate the relationship between CSR and analysts’ forecast accuracy in several countries. They find that the issuance of stand-alone CSR reports is associated with lower analyst forecast errors. This relationship is stronger in more stakeholder-oriented countries. Garrido-Miralles et al. (2016) also find a negative relationship between the earnings forecast error and the issue of sustainability reports in Spain. Dhaliwal et al. (2011) find that companies with superior CSR attract dedicated institutional investor and analyst coverage.

On the other hand, Adhikari (2016) finds that firms with greater analyst coverage have lower corporate social responsibility scores, consistent with the view that spending on CSR is a manifestation of an agency problem. Thus, if it is an agency problem then better monitoring due to greater analyst coverage should force managers to cut back on CSR activities. Garcia-Sanchez et al. (2020) show that firms which adopt sustainable development goals received, immediately in the same or the next year, sell recommendations from analysts, and long CSR strategies have less impact on recommendations. Zhang and Wei (2019) show that analysts’ recommendations are positively and optimistically associated with the firms with less information disclosure. They also find that analysts who have private information have stronger recommendations. Hinze and Sump (2019), in their review, find that many existing literature confirm the positive relationship between CSR performance and optimistic recommendations, and Wang and Jiang (2019) show that analysts’ recommendations mediate the relationship between brand equity and sustainable performance in Chinese listed firms.

The literature review and interviews with financial analysts suggest that analysts increasingly incorporate ESG information in their recommendations. According to Eccles et al. (2011), there is a large and growing market interest in ESG information and policies. Luo et al. (2015) conduct in-depth interviews with sell-side analysts who acknowledge that they heed CSR disclosures and incorporate them into their investment recommendations. Moreover, Fieseler (2011) interview financial analysts, and the results suggest that responsibility issues are increasingly becoming part of mainstream investment analysis.

Thus, we state the following hypothesis:

Hypothesis 1 (H1).

There is a positive relationship between analysts’ recommendations and ESG disclosure.

The Arabian Gulf communities have tribal origins. The tribal elder/Shaikh plays a prominent and important role in the administration of tribal affairs. This role remained prominent as these countries became monarchies and is now administered by royal families. Recently, many royal families have become involved in business (Kamrava et al. 2016). Many GCC-listed companies now have at least one royal family member on their board of directors (Halawi et al. 2008). The effect of royal family directors on ESG disclosure can be seen in two ways. First, a positive effect of royal family directors on the board will enhance social responsibility activities. Alazzani et al. (2019) mention that to enhance the regimes’ legitimacy, the royal families may become more involved in CSR activities as a way to minimize popular frustration with the increased concentration of wealth in the hands of a few. Princes/Shaikhs must actively contribute to support the communities in which they live and show that they care for the members of their community. From a political and reputational aspect, royal family members have motives to adopt legal and ethical business practices and CSR, to show their social support through sympathetic activities (e.g., Eid Al-Thani in Qatar, the King Faisal Foundation and the Alwaleed Philanthropies in Saudi Arabia; the Khalifa Foundation in the UAE; and the Alsobah Foundation in Kuwait). Drawing on servant leadership theory, Alazzani et al. (2019) investigate whether the presence of royal family members on GCC boards of directors influences CSR reporting. They find a positive relationship between the presence of royal family directors and CSR reporting.

On the other hand, a negative impact of the royal family on CSR disclosures is grounded on the argument that royal family directors will maximize their self-interest. For instance, Al-Hadi et al. (2016) find that ruling/royal families enjoy less risk reporting pressures and more government rent-seeking, and more earnings management.

As mentioned earlier in the theories section, impression management is the activity of controlling or regulating information which influences the impression formed by audiences (Schlenker 1980). The success of the impact on others by impression management depends on the perception of the audience. This influence may be positive or negative (Gardner and Martinko 1998). As powerful people, royal family directors in our case might be expected to perform more impression management than others. One of the key audiences in the stock markets is financial analysts. So, we assume that impression management can explain why royal family directors might moderate the relationship between ESG and analysts’ recommendations.

The key question is how financial analysts evaluate such an impact. We want to see if they perceive it as a positive or negative impact. Thus, we state the following hypothesis:

Hypothesis 2 (H2).

The existence of a royal family director will affect the relationship between analysts’ recommendations and ESG disclosure.

4. Research Methodology and Design

4.1. Sample Selection and Data Source

There were 4386 total observations from 2010 to 2016 for companies in the six GCC countries and the seven stock markets. We used the Bloomberg dataset6 for collecting the data. We eliminated 3849 companies’ results for one of four reasons: (a) investment recommendations were not available, (b) ESG data were not available via Bloomberg, (c) accounting data were not available through Bloomberg and (d) companies only were present for one year. The final sample consists of 537 firm-year observations.

Table 1. Sample distribution based on country and year.

Country	2010	2011	2012	2013	2014	2015	2016	Total	%
Saudi Arabia	11	23	26	26	27	28	28	169	0.31
United Arab Emirates	11	21	22	24	24	25	13	140	0.26
Qatar	4	10	11	13	17	18	12	85	0.16
Kuwait	7	9	10	10	11	11	4	62	0.12
Oman	5	7	8	9	9	9	9	56	0.1
Bahrain	1	3	5	5	5	5	1	25	0.05
Total	39	73	82	87	93	96	67	537	100

Panel A shows that Saudi Arabia represents about 31% of the total observations (169), followed by the United Arab Emirates with 140 observations (26%) from the two Dubai and Abu Dhabi Stock Markets. The other four countries, Qatar, Bahrain, Oman, and Kuwait, constituted 43% of the sample.

4.2. Empirical Models and Variable Definitions

The following baseline ordinary least squares (OLS) and regression models are used to examine the effect of ESG on analysts’ stock recommendations:

REC = β₀ + β₁ESG + β₂ROA + β₃LEVERAGE + β₄RI + β₅BIDASK + β₆GOV_OWN + β₇GDP + β₈EXANALYST
+ β₉BOARD + β₁₀SIZE + β₁₁TOBIN + Year-effect + Industry-effect + Country-effect + ε

(1)

The moderation effects of the royal family model are presented as;

REC = β₀ + β₁ESG + β₂RFP + β₃ESG * RFP + β₄ROA + β₅LEVERAGE + β₆RI + β₇BIDASK + β₈GOV_OWN + β₉GDP +
β₁₀EXANALYST + β₁₁BOARD + β₁₂SIZE + β₁₃TOBIN + Year-effect + Industry-effect + Country-effect + ε

(2)

The following is an explanation for these variables:

Variable	Definition
REC	Indicates the analysts’ opinions on the stock performance. The rating is calculated by converting each of the analysts’ recommendations into a number from 1–5 and taking the average. Originally, this measure is reported with 5 = strong buy, 4 = buy, 3 = hold, 2 = underperform, and 1 = sell.
ESG	Proprietary Bloomberg scores based on the extent of a company's Environmental, Social, and Governance (ESG) disclosure.
RFP	Percentage of royal family directors on the board.
ESG*RFP	The interaction of ESG with RFP.
ROA	Return on Assets.
LEVERAGE	Total debt divided by total shareholders’ equity.
RI	The total excess return is defined as the return for the company above the market return. It is calculated as ri,t+ 1 − rm,t+1, where i is the firm, and m is the market.
	The Bloomberg ticker: TOT_RETURN_INDEX_GROSS_DVDS.
GOV_OWN	Percentage of government ownership.
GDP	Gross domestic product growth rate.
EXANALYST	Log of (the number of directors on the board with experience in financial analysis plus 1)7.
BOARD	The number of directors on the board.
BIDASK	A bid-ask spread is the amount or percentage by which the ask price (THE LOWEST) exceeds the bid price (the highest) for security in the stock market and reflects the liquidity of the stocks. It is also a reflection of the supply and demand for a particular security.
SIZE	Log of total market value of a company's shares.
TOBIN	Ratio of the market value of a firm to the replacement cost of the firm’s assets.
SOCIAL	Bloomberg score based on the extent of a company’s Social disclosure.
ENV	Bloomberg score based on the extent of a company’s Environmental disclosure.
GOV	Bloomberg score which measures the quality of corporate governance.

We also included year, country, and industry effects. Nearly one-third of sample firms are involved in environmentally sensitive industries. Moreover, we used White’s (1982) robustness test to obtain unbiased standard errors of OLS coefficients under heteroscedasticity to ensure that our estimator has the lowest variance among all unbiased estimators.

5. Discussion of the Variables Used

5.1. Dependent Variable: Analysts’ Recommendations

Bloomberg rates the analysts’ recommendations by converting each of the analysts’ recommendations into a number from 1–5 and taking the average. A score of 3 means that analysts believe that the stock should be held, less than 3 means it should be sold, and greater than 3 means it should be bought. The company receives many recommendations during the year from many employees of a brokerage or fund management house who study companies and make buy-and-sell recommendations on stocks of these companies. Bloomberg calculates the average of these investment recommendations (REC) during the year. We use this average as a measurement of the analysts’ recommendations.

5.2. Independent Variable

5.2.1. ESG

Recently, rating firms (e.g., Bloomberg, KLD, Thomson Reuter’s ASSET4) have played an important role in assessing areas ranging from sustainability to corporate (governance?). Managers, investors, and scientists are increasingly relying on these ratings for strategic decisions, investing trillions of dollars in capital, and studying CSR, guided by the implicit assumption that the ratings are valid. ESG data have the potential to provide crucial market transparency and a unique lens through which to assess the future company and investment performance. It is crucial to acknowledge that there is a divergence between different ESG ratings and organizations. Moreover, every organization can rank a particular company differently (Chatterji et al. 2016). In this study, we use Bloomberg ESG rating. Bloomberg LP focuses on a future in which environmental and social issues will have increasingly critical implications for firms and investors. Increasing demand for sustainability analytics has become an essential part of any investment decision. Bloomberg began collecting ESG data in 2008.

The Bloomberg ESG Disclosure Score is based on Global Reporting Initiative (GRI’s) guidelines and covers a total of 247 possible criteria across environmental, social and governance dimensions (Eccles et al. 2011). This disclosure score out of 100 is based on whether actual information is revealed (Mueller 2014) for each of the environmental, social and governance categories (Wang and Sarkis 2017). This is not an assessment of a firm’s strengths or concerns, as in the case of Kinder, Lydenberg, Domini Research and Analytics (KLD) (Hillman and Keim 2001). These Bloomberg ESG Disclosure Scores are not precise performance metrics. They specify the degree to which a firm reports ESG information (Eccles et al. 2011).

Marquis et al. (2011) assert that the ESG database provided by Bloomberg is the most comprehensive methodology to evaluate and assess firms’ ESG activities and outcomes. Many previous studies have used ESG data from Bloomberg to measure ESG disclosure, among them, Dorfleitner et al. (2015), Fatemi et al. (2017) and Halbritter and Dorfleitner (2015). Thus, this study followed such papers and used the Bloomberg ESG data. We think that it provides sufficient information to examine the relationships between ESG disclosure, investment recommendations, and politically connected firms.

5.2.2. Royalty Political Connection

Following Al-Hadi et al. (2016), we measure the existence of royal family directors using the percentage of royal family directors on the board of directors.

5.2.3. Control Variable

At the firm level, many control variables have been included in the model to control for any potential confounding effects. We follow Ioannou and Serafeim (2015) and Luo et al. (2015) by incorporating these variables in the model. Financial performance (ROA) is one of the key factors in analysts’ decisions. We also control for LEVERAGE because we think it is consistent with the idea that analysts mitigate information asymmetry, and so firms with less analyst coverage may have higher debt ratios since they are unable to issue equity regularly. Total returns index (RI) is also one of the capital market performance factors that influences analysts’ decisions (see variable definition). We also control for the GOV_OWN, as it is likely that government ownership in companies affects many inputs related to investment recommendations. It is also expected that countries with high GDP may put some effort into increasing disclosures and better organizing and monitoring of capital market intermediaries. The directors with experience in financial analysis (EXANALYST) might also affect the type of information disclosed by the firm.

Some previous studies found that board size (BOARD) affects financial analysts’ forecast accuracy (e.g., Byard et al. 2006). Liquidity of shares might also be another factor that influences analysts’ recommendations Following Cumming et al. (2011) and Drake et al. (2010), this study uses bid-ask spreads as a proxy for liquidity. Ioannou and Serafeim (2010) use BIDASK as a control variable, and they argue that larger spreads characterize more opaque companies. Analysts might find those companies harder to understand and thus be less optimistic about them. Controlling for firm size (SIZE) is motivated by Ioannou and Serafeim (2010), who argued that “the analysts might issue more optimistic recommendations for large firms since trading in these firms generates more trading commissions and these firms are more likely to generate investment banking business”. Tobin’s q (TOBIN), the ratio of the market value of a company's assets, has been used widely in the literature as a firm value proxy. We conjecture that a firm with a low Tobin’s q might be more likely to be undervalued than a firm with a high Tobin’s q. This will thus affect analysts’ recommendations. We also control for year effects, industry effects and country effects.

6. Findings and Result Discussion

6.1. Descriptive Statistics

The descriptive statistics of the variables are exhibited in

Table 2. Descriptive statistics.

Variable	n	Mean	S.D.	Min	0.25	Med	0.75	Max
REC	537	3.7	1.03	1	3	4	4	5
ESG	537	12.29	10	2.07	6.61	10.33	10.97	47.11
RFP	537	10.44	14.57	0	0	0	17.65	57.14
ROA	537	4.35	5.35	−16.55	1.55	2.26	6.37	29.83
LEVERAGE	537	92.3	136.92	0	25.08	60.11	113.83	1692.94
RI	537	133.96	342.07	0.12	2.78	18.92	77.97	2452.89
BIDASK	537	1.33	1.6	0.07	0.48	1.01	1.33	18.18
GOV_OWN	537	0.12	0.23	0	0	0	0.1	0.84
GDP	537	4.43	2.96	−2.37	2.7	3.98	5.41	19.59
EXANALYST	537	0.33	0.56	0	0	0	1	2
BOARD	537	8.99	1.77	5	8	9	10	18
SIZE	537	8.9	1.82	0	7.84	9.24	10.17	12.72
TOBIN	537	1.25	0.44	0.51	1.02	1.09	1.33	3.82

All variables are defined in Empirical models and variable definitions section.

. The average ESG disclosure score is 12.3, which is quite low and consistent with the wide range of studies conducted in the Middle East and GCC countries. In reality, in this region, the disclosure of non-financial information is still in its infancy. The average REC is 3.7, which implies that most companies gain favourable recommendations by analysts either to hold or buy. A hold recommendation generally expects the security to perform at a market rate or the same pace as comparable securities. A buy recommendation is given by analysts for a security that is expected to outperform the average market return of comparable stocks in the same sector or industry. Thus, the analysts believe that most securities in GCC stock markets are performing well. In reality, this issue needs more in-depth investigation of why most of the stocks gain favourable recommendations. The percentage of royal family directors (RFP) is about 10.4%. On average, the sample companies have 92% LEVERAGE. This result implies that most of the companies finance their operations from debt. The ROA of about 4.4% implies that most of the firms are performing quite well. The average of BIDASK is 1.3. A bid-ask spread is the amount or percentage by which the ask price exceeds the bid price for a security in the stock market and reflects the liquidity of the stocks. It is also a reflection of the supply and demand for a particular security. The spread percentage is normal where it is assumed that the ask price is higher than the bid prices. The average BOARD size is around nine members. On average, at least one of the directors has experience in financial analysis (EXANALYST). The average government ownership in our sample is 12% (GOV_OWN). The average of GDP is 4.4%. The average of TOBIN is 1.25.

6.2. Univariate Analysis

Table 3. T-test between politically connected versus non-politically connected firms.

	Non-Politically Connected	Politically Connected	Difference	t-stat
	(n = 301)	(n = 236)
REC	3.58	3.84	−0.26	−2.898 ***
ESG	11.59	13.18	−1.59	−1.836 *
SOCIAL	6.02	9.248	−3.23	−2.56 **
ENV	4.1	5.73	−1.63	−1.80 *
ROA	3.51	5.43	−1.92	−4.194 ***
LEVERAGE	93.63	90.59	3.04	0.255
RI	116.95	155.66	−38.71	−1.303
GOV_OWN	0.09	0.15	−0.06	−3.187 ***
GDP	4.23	4.69	−0.46	−1.779 *
EXANALYST	0.34	0.33	0.012	0.259
BOARD	8.98	9.02	−0.04	−0.262
BIDASK	1.31	1.35	−0.042	−0.299
SIZE	8.8	9.02	−0.225	−1.421
TOBIN	1.21	1.31	−0.106	−2.801 ***

* p < 0.10; ** p < 0.05; *** p < 0.01 indicate that the estimated coefficients are statistically significant at the 10 percent, 5 percent, and 1 percent levels, respectively. All variables are defined in Empirical models and variable definitions section.

shows the t-test for all dependent and independent variables partitioned by royally politically connected firms and unconnected firms. Our comments on the significant results are set below. We find that the mean REC for politically connected firms is 3.84, which is statistically larger than the mean REC for non-connected companies of 3.58 (two tailed, p < 0.001). The results suggest that politically connected firms gain more optimistic recommendations than non-politically connected firms. Based on our knowledge, this is the first study which contributes to the politically connected firm literature by showing that sell-side analysts tend to issue favourable recommendations to firms with political connection, particularly, those firms with royal family directors. Importantly, we also find that ESG for royal politically connected firms is higher than for non-connected firms (mean of 13.18 compared to mean 11.6), which is significant at two tailed (p < 0.1). Similarly, politically connected firms also have statistically higher SOCIAL and ENV scores than non-politically connected firms. Further, the means of ROA, GOV_OWN, and TOBIN were also found to be higher for politically connected companies, 5.4%, 15%, and 1.31%, compared to non-connected companies, 3.5%, 9%, and 1.21%, (two tailed p < 0.001), respectively. These results imply that politically connected companies perform well and attract higher government investment than their counterparts.

We conduct further analysis to see which countries disclose more by comparing the countries' average ESG. In

Table 4. Comparison between countries in terms of ESG.

Total	Country	mean_ESG
62	Kuwait	8.343037
85	Qatar	10.23224
25	Bahrain	10.73654
169	Saudi Arabia	10.90623
140	United Arab Emirates	15.2897
56	Oman	17.13645

, we find the average of ESG disclosure in Oman is a score of 17, which is the highest followed by United Arab Emirates. Kuwait has the lowest disclosure with a score of 8.4.

Table 5. Pearson correlation matrix.

		1	2	3	4	5	6	7	8	9	10	11	12	13	14
1	REC	1
2	ESG	0.11 *	1
3	RFP	0.06	0.11 **	1
4	ROA	0.2 ***	−0.05	0.05	−0.08 *
5	LEVERAGE	−0.02	0.05	0.03	0.08	1
6	RI	−0.05	−0.17 ***	0.22 ***	−0.05 *	−0.07 *	1
7	GOV_OWN	0.08	−0.01	−0.04	−0.06	0.16 ***	−0.13 ***	1
8	GDP	0.14 ***	−0.01	0.07	0.06	0.04	−0.16 ***	0.02	1
9	EXANALYST	0.17 ***	0.09 **	−0.13 ***	−0.1 **	0.17 ***	−0.17 ***	0.58 ***	0.04	1
10	BOARD	−0.07	0	−0.03	0.04	−0.11 ***	−0.2 ***	−0.01	0	0.07	1
11	BIDASK	0.02	0.06	0.01	−0.05	0.03	0.1 **	−0.07	0.01	0.05	−0.14 ***	1
12	SIZE	0.12 ***	−0.02	0.04	0.12 ***	0	−0.25 ***	0.06	0.12 ***	0.01	0.15 ***	−0.35 ***	1
13	TOBIN	0	−0.02	0	−0.05	−0.16 ***	0.01	0.17 ***	0.01	0.19 ***	0	−0.12 ***	0.19 ***	1
14	IND	0.1 **	0.01	−0.2 ***	−0.17 ***	0.17 ***	−0.08 *	0.3 ***	0.02	0.33 ***	−0.1 **	0	−0.07 *	0.17 ***	1

* p < 0.10; ** p < 0.05; *** p< 0.01 indicate that the estimated coefficients are statistically significant at the 10 percent, 5 percent, and 1 percent levels, respectively. All variables are defined in Empirical models and variable definitions section, except for IND, which takes a value of 1 if firm is in environmentally sensitive industries, and 0 otherwise.

shows the Pearson Correlation Matrix for the main variables included in the regressions. The REC is positively associated with ESG. This implies that the analysts in our sample were favourable towards ESG disclosure. We also find a positive and significant correlation between ESG and RFP. We also find that there is a positive and significant correlation between REC and ROA, GDP, EXANALYST, SIZE, and environmentally sensitive industries (IND).

6.3. Multivariate Analyses

In this section, we present the findings of estimation results after controlling for several variables. Our dependent variable is investment recommendations (REC), and the independent variables of interest are environmental, social and governance (ESG) and the royal family directors (RFP).

Table 6. Baseline regression for the association between ESG and REC.

Variables	Model 1
ESG	0.008 **
	(0.004)
ROA	0.064 ***
	(0.02)
LEVERAGE	0.018
	(0.001)
RI	0.001 ***
	(0.001)
GOV_OWN	−0.667 ***
	(0.23)
GDP	0.037 **
	(0.02)
EXANALYST	0.299 ***
	(0.09)
BOARD	−0.066 ***
	(0.02)
BIDASK	0.008
	(0.03)
SIZE	0.145 ***
	(0.05)
TOBIN	−0.717 ***
	(0.16)
Constant	3.904 ***
	(0.52)
Industry effects	Yes
Year effect	Yes
Country effects	Yes
Observations	537
R-squared	0.258
Robust standard errors in parentheses

** p < 0.05; *** p < 0.01 indicate that the estimated coefficients are statistically significant at the 10 percent, 5 percent, and 1 percent levels, respectively. All variables are defined in Empirical models and variable definitions section.

shows the coefficients results of our base regression for the association between ESG and REC.

In particular, the coefficient of REC and ESG of 0.008 is positive and statistically significant at (p < 0.05). This result is consistent with the study of Ioannou and Serafeim (2015), which finds a significant positive relationship between CSR disclosure and investment recommendation. The analysts favorably assess ESG disclosure. This result supports hypothesis H1. Prior literature states that more informative disclosure attracts more financial analysts because gathering information becomes less costly. The investors rely on analysts when they make their investment decisions and might use ESG information. It might also be perceived by market analysts as positively contributing towards a company’s long-term profitability. They are much more likely to recommend potential shareholders to purchase the shares of these firms with better ESG. Therefore, Table 6 provides results consistent with our first hypothesis that analysts perceive ESG as a positive factor for a firm’s long-term financial performance which improves the firm’s value. The rest of the control variables have been found to influence the REC, except BIDASK.

6.4. Moderating Effects of Royalty Political Connection

Table 7. The moderating effect of royal family between ESG and REC.

Variables	Model 2
ESG	0.013 ***
	(0.005)
RFP	0.013 **
	(0.006)
ESG * RFP	−0.001 **
	(0.001)
ROA	0.063 ***
	(0.018)
LEVERAGE	0.001
	(0.001)
RI	0.001 ***
	(0.001)
GOV_OWN	−0.682 ***
	(0.221)
GDP	0.037 **
	(0.016)
EXANALYST	0.292 ***
	(0.089)
BOARD	−0.066 ***
	(0.022)
BIDASK	0.003
	(0.035)
SIZE	0.156 ***
	(0.054)
TOBIN	−0.704 ***
	(0.164)
Constant	2.059 ***
	(0.525)
Industry effects	YES
Year effect	YES
Country effects	YES
Observations	537
R-squared	0.23

Robust standard errors in parentheses: ** p < 0.05; *** p < 0.01 indicate that the estimated coefficients are statistically significant at the 10 percent, 5 percent, and 1 percent levels, respectively. All variables are defined in Empirical models and variable definitions section.

provides results of the moderation effects of royal family directors on the relation between ESG and REC. The coefficients of ESG (.013) and RFP (.013) are found to be positively and significantly associated with REC at p < 0.001 and p < 0.05, respectively. Interestingly, the coefficient for ESG is higher in the recommendation model with moderator, compared to the baseline regression. However, the interaction’s coefficient between ESG and RFP is negative and significant at p < 0.05. One interpretation is that the analysts perceive that those firms with both high ESG and high RFP may engage in window dressing or greenwashing for the sake of appearing responsible. Several studies indicate that the presence of royal family in the firm can attract several self-benefits such as lower cost of debt (Al-Hadi et al. 2016), more government contracts, eliminating competitors from entering a market (Bunkanwanicha and Wiwattanakantang 2009), and enhancing the performance of family firms (Xu et al. 2015). Thus, analysts can recommend firms with royal family in different circumstances. However, our result suggests that high ESG by RFP firms does not necessarily lead analysts to issue more favourable recommendation for those firms. Another interpretation stems from an impression management perspective. Those analysts’ experts see the ESG information disclosed by firms with royal family directors negatively, thus they discount it because the information environment is opaque. In other words, they do not really trust such information. As we mentioned earlier, impression management, especially by powerful leaders, may affect audiences negatively.

6.5. Endogeneity Check Using Two Stage Least Square Equation (2SLS)

The issue of endogeneity (or reverse causality) might be particularly problematic when assessing the relation between analysts’ recommendations and ESG disclosures. The sign, magnitude, or statistical significance of these estimates may be biased due to endogeneity (e.g., causality effects, omitted control variables), that is, the ESG and the error term being correlated. This may be because it is correlated with another determinant that is excluded from or not fully controlled for in our regression models, even though the OLS estimation suggests a positive and significant association between ESG and recommendations. To address this concern, we adopt a two-stage instrumental variable (2SLS) approach to re-examine the findings reported in Table 6. This method is suitable only if the instrumental variable(s) is/are correlated with the endogenous regressor (here ESG) but uncorrelated with the error term of the second-stage regression. To select our instrumental variable, we follow León-Ledesma and Thirlwall (2002) and Wintoki et al. (2012) and use a lagged variable (ESG_LG) of our main independent variable (ESG) in the first stage of the Equation (3).

REC = β₀ + β₁ESG_LG + controls variables + e

(3)

We then include the error term (ê) from Equation (3) into our second stage regression Equation (4).

REC = β₀ + β₁ESG + controls variables + δê + e

(4)

Table 8. Endogeneity check using Two Stage Least Square Equation (2SLS).

Variables	Model 1
Second stage
ESG_LG	0.010 **
	(0.01)
ROA	0.070 ***
	(0.01)
LEVERAGE	0.001
	(0.01)
RI	0.001 ***
	(0.01)
GOV_OWN	−0.551 **
	(0.27)
GDP	0.009
	(0.03)
EXANALYST	0.214 *
	(0.11)
BOARD	−0.068 **
	(0.03)
BIDASK	−0.007
	(0.03)
SIZE	0.101 *
	(0.06)
TOBIN	−0.696 ***
	(0.15)
Constant	2.569 ***
	(0.55)
Industry effects	Yes
Year effect	Yes
Country effects	Yes
Observations	441
R-squared	0.26
Second Stage:
1—Under identification test	0.001
2—Weak identification test	185.676
Stock–Yogo weak ID test critical values: 10% maximal IV size	16.38
3—Endogeneity test	0.0433

Results of First stage with the post-estimations are available upon request. Robust standard errors in parentheses. * p < 0.10; ** p < 0.05; *** p < 0.01 indicate that the estimated coefficients are statistically significant at the 10 percent, 5 percent, and 1 percent levels, respectively. All variables are defined in Empirical models and variable definitions section.

shows the results of the second stage regression of the 2SLS. We find consistent evidence with our main H1 that ESG is positively and significantly associated with REC. In particular, the coefficient of ESG is 0.10 at p < 0.05. Furthermore, following prior literature (e.g., Al-Hadi et al. 2016), we check the quality of using the instrumental variable. In Table 8, we show a number of post-estimations tests using ESG_LG as our instrumental variable. First, based on the Heckman endogeneity test, we find that our regressor in the first stage is endogenous (p < 0.0433). Second, we run the under identification test (Kleibergen–Paap rk LM statistic) and find that our p-value is significant at p < 0.001. Third, we test for the weak identification test of our instrumental variable. We find that our Kleibergen–Paap rk Wald F statistic value is 185.676, which is larger than the Stock–Yogo weak ID test critical values: 10% maximal instrumental variables (IV) size of 16.38.

6.6. Additional Analyses

In

Table 9. Sub-analyses with individual ESG index; ENV, SOCIAL, and GOV.

Variables	(1)	(2)	(3)	(4)	(5)	(6)
	ENV	ENV * RFP	SOCIAL	SOCIAL * RFP	GOV	GOV * REP
ENV	0.009 ***	0.015 ***
	(0.003)	(0.004)
RFP		0.009 **		0.008 *		0.020 **
		(0.004)		(0.004)		(0.009)
ENV * RFP		−0.001 **
		(0.000)
SOCIAL			0.007 ***	0.010 ***
			(0.002)	(0.003)
SOCIAL * RFP				−0.0001 *
				(0.000)
GOV					−0.003	−0.001
					(0.004)	(0.005)
GOV * RFP						−0.000
						(0.000)
ROA	0.064 ***	0.062 ***	0.065 ***	0.063 ***	0.063 ***	0.063 ***
	(0.017)	(0.017)	(0.018)	(0.018)	(0.017)	(0.017)
LEVERAGE	0.000	0.000	0.000	0.000	0.000	0.000
	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
RI	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***
	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
GOV_OWN	−0.662 ***	−0.659 ***	−0.686 ***	−0.694 ***	−0.737 ***	−0.783 ***
	(0.225)	(0.221)	(0.225)	(0.222)	(0.226)	(0.222)
GDP	0.036 **	0.036 **	0.036 **	0.037 **	0.037 **	0.036 **
	(0.016)	(0.016)	(0.016)	(0.016)	(0.016)	(0.016)
EXANALYST	0.299 ***	0.292 ***	0.305 ***	0.302 ***	0.330 ***	0.337 ***
	(0.090)	(0.089)	(0.090)	(0.089)	(0.093)	(0.092)
BOARD	−0.065 ***	−0.066 ***	−0.066 ***	−0.069 ***	−0.066 ***	−0.068 ***
	(0.022)	(0.022)	(0.022)	(0.022)	(0.022)	(0.022)
BIDASK	0.009	0.004	0.009	0.005	0.008	0.008
	(0.034)	(0.034)	(0.034)	(0.034)	(0.034)	(0.034)
SIZE	0.136 **	0.145 ***	0.139 **	0.145 ***	0.149 ***	0.156 ***
	(0.056)	(0.056)	(0.055)	(0.054)	(0.055)	(0.055)
TOBIN	−0.711 ***	−0.682 ***	−0.721 ***	−0.712 ***	−0.707 ***	−0.705 ***
	(0.164)	(0.164)	(0.165)	(0.165)	(0.162)	(0.164)
Constant	2.241 ***	2.194 ***	2.210 ***	2.201 ***	2.238 ***	2.257 ***
	(0.537)	(0.533)	(0.530)	(0.528)	(0.544)	(0.549)
Industry effects	Yes	Yes	Yes	Yes	Yes	Yes
Year effect	Yes	Yes	Yes	Yes	Yes	Yes
Country effects	Yes	Yes	Yes	Yes	Yes	Yes
Observations	537	537	537	537	537	537
R-squared	0.261	0.269	0.261	0.267	0.255	0.262

Robust standard errors in parentheses. * p < 0.10, ** p < 0.05; *** p < 0.01 indicate that the estimated coefficients are statistically significant at the 10 percent, 5 percent, and 1 percent levels, respectively. All variables are defined in Empirical models and variable definitions section, except for SOCIAL, ENV and GOV: SOCIAL = Bloomberg score based on the extent of a company's Social disclosure. ENV = Bloomberg score based on the extent of a company's Environmental disclosure. GOV = Bloomberg score which measures the quality of corporate governance.

, we regress the disaggregated ESG measure into its three categories, ENV, SOCIAL, and GOV, with the analysts’ recommendations (REC). These analyses will provide us with a better understanding: first, which category of ESG (ENV, SOCIAL, or GOV) affects the analysts’ recommendation of the firm, second, how the interplay between royal family directors and the three categories of ESG (ENV, SOCIAL, or GOV) affects analysts’ recommendations.

Table 9 shows six columns of regressions models: columns 1, 3, and 5 for ENV, SOCIAL, and GOV individually, and columns 2, 4, and 6 for the interaction terms between ENV, SOCIAL, and GOV and RFP. The results show that ENV and SOCIAL performance are positively related to analysts’ recommendations, and their coefficients are 0.009 and 0.007 and significant at p < 0.05, respectively. However, we did not find that governance positively related to analysts’ recommendations.

The interaction terms (ENV * RFP) and (SOCIAL * RFP) are both negative and significant at p < 0.05. This might suggest that firms that have a higher presence of royal family directors and report higher ENV and SOCIAL disclosures are likely to use the corporate social responsibility engagement as a vehicle for impression management. If this is the case, the analysts can see through the corporate spin and react with skepticism in giving their recommendations. In fact, these results contribute to our findings that ENV and SOCIAL activities are less structured and regulated in GCC countries. The royal families, therefore, have more discretion to influence these types of practices in GCC countries.

On the other hand, the corporate governance disclosure (GOV) is applied in all GCC countries with less skepticism since it is highly mandated and regulated as well as being closely monitored by various authorities, such as Central Banks, auditors, stock market regulators, and governmental ministries. Thus, we assume that the royal family and analysts perceive GOV disclosure differently. In fact, we find in Table 9 that the corporate governance category (GOV) of ESG and the interaction term (GOV * RFP) have no relationship with analysts’ recommendations. This result is also consistent with prior studies that provide evidence on the royal family having less discretion of corporate governance in GCC countries (Al-Yahyaee and Al-Hadi 2016; Al-Hadi et al. 2015).

7. Conclusions

There has been a debate in academia regarding whether capital markets benefit from positive investment recommendations arising from the social and environmental concerns of the CSR practices. First, we are interested to see whether better ESG disclosure influences sell-side analysts’ recommendations or not. Second, we want to see how financial analysts view the firms with royalty political connections in terms of their ESG disclosure. There are only a few prior studies that have focused on the link between ESG and investment recommendations. This has created a gap in the literature. We fill this important lacuna by bringing in new evidence from different cultures. We find that connected firms gain more optimistic recommendations than non-politically connected firms. As far as we are aware, this is the first time that a study that contributes to the politically connected firm literature shows that sell-side analysts tend to issue favorable recommendations to firms with political connection, particularly, those firms with royal family directors. Importantly, we also find that ESG for royal politically connected firms is higher than for non-connected firms (mean of 13.18 compared to mean 11.6), which is significant at two-tailed (p < 0.1). We also find that there is a positive relationship between ESG disclosure and sell-side analysts’ recommendations. These findings are in line with Ioannou and Serafeim (2015). This study confirms the outcomes of the study of Eccles et al. (2011), which found that the capital market shows a high interest in the level of CSR disclosures, and Luo et al. (2015), who found that analysts factor CSR information into their investment recommendations. It also found that ESG disclosures in companies with political connections are perceived negatively by sell-side analysts.

Generally, analysts act as an important intermediary channel between the company and the stakeholders. As industry experts, financial analysts reduce the information asymmetry between firms and investors by incorporating firm CSR information into their recommendations. We extend the literature from other capital market settings and cultures.

It is also noticeable that the ESG disclosure is very low. The pressure of stakeholders and government bodies on firms to disclose ESG information is very weak. To improve this, policymakers and stakeholders can be advised to play key roles by enforcing firms and issuing some compulsory rules on ESG disclosure. The securities commissions can also make some obligatory rules and follow some practices of some Western and Asian securities commissions that obligate firms to disclose ESG information. Due to our small sample size of 537 firm observations, it can be difficult to make inferences. Another limitation is using the Bloomberg ESG rating as the proxy for ESG disclosure. There is a divergence between different ESG ratings and every rating agency can rank a particular company differently (Chatterji et al. 2016; Dorfleitner et al. 2015). Future research can conduct surveys or interviews with financial analysts to study their perception about the existence of royal family directors on the board of directors. It is also worth conducting future research to study the relationship between foreign ownership and the performance of firms by comparing firms with royal family directors and firms which have no royal family directors.