University of Amsterdam
Faculty of Economics and Business
Finance and Organization
Bachelor Thesis
Socially Responsible Investing: The effect of
investing in LGBT-friendly companies on
investor’s return
Anita Yee Kei Kwong
10094911
Supervisor: Mw. Simin He
June 2015, Amsterdam
2
STATEMENT OF ORIGINALITY
This document is written by Anita Yee Kei Kwong who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
ABSTRACT
This thesis analyzes the effect of using LGBT-equality as a socially responsible investing (SRI) screen on investor’s return. A portfolio of LGBT-friendly firms in the US outperformed the market by a significant percentage of 3.48% per year using the CAPM model and 3.24% when using the Carhart 4-factor model. These results are controlled for outliers and the four factors from the Carhart (1997) 4-factor model. These findings supports the evidence that certain SRI screens may improve investor’s return.
4
TABLE OF CONTENTS
Abstract………..3
1. Introduction………..………...5
2. Theoretical discussion………..6
2.1 Negative effects of socially responsible investing……..………...………6
2.2 Positive effects of SRI screens………..………7
2.3 Effects of LGBT-supportive policies………..………...7
3. Data………...9
3.1 HRC’s Corporate Equality Index (CEI) reports………..9
3.2 Portfolio i: only LGBT-friendly firms………10
3.3 Portfolio p: LGBT- and non-LGBT-friendly firms………10
4. Model………....11
4.1 Performance evaluation method of portfolio i………11
4.2 Logistic regression model to account for firm characteristics………..12
5. Empirical results………13
5.1 Performance evaluation results: CAPM and Carhart 4-factor model………13
5.2 Logistic regression model results………...14
6. Discussion………..15
5
1. INTRODUCTION
People and businesses are becoming more aware of being socially and environmentally
responsible in their actions. Large investment banks are monitoring this trend and foresee some business opportunities that arise from the growing interest in socially responsible investing (SRI). As a result, many mutual funds are incorporating SRI in their strategies.
On October 2013, Credit Suisse added another ethical investment product to the existing mainstream. It launched its LGBT-equality index. It includes companies on the basis of their corporate policies that support lesbian, gay, bi-sexual and transgender (LGBT) equality. It is the first of its kind that uses LGBT-equality as a SRI screen.
Despite increasing interests in SRI, there are contradicting theories about SRI. Traditional investment theories argue that the use of any SRI screens worsens investor’s portfolio performance, because it restricts investors in their choices (e.g. Markowitz, 1959). While other theories suggests that SRI screens might identify companies with superior performance, leading to a positive effect on investor’s return (e.g. Freeman, 1984; Derwall & Koedijk, 2009).
These conflicting theories has led to many empirical studies in the field of SRI
performance. Many existing studies find a zero or negative effect of SRI screens on investor’s return. However, some studies do find a positive effect. Thus, the conclusion on whether SRI strategies are profitable remains inconclusive.
Remarkably, no study had been done to investigate the effect of using LGBT-equality as a SRI screen on investor’s return. The objective of this thesis is to fill in that gap by evaluating the performance of LGBT-friendly companies relative to the market in the United States. Moreover, it will focus on the relationship between firm characteristics, performance and the level of LGBT-friendliness. It also explains some effects of implementation of LGBT-supportive policies.
The answers to this topic is important for several reasons. First, it will give investors a better impression on the profitability of the LGBT investment product introduced by Credit Suisse. Since this investment vehicle is relatively new and unique, there aren’t many past information available about its performance. Second, it will contribute to the existing debate about whether the use of SRI screens is profitable. Moreover, it will also support existing theories about whether LGBT-supportive policies affects firm performance.
The structure of this study is as follows. Section 2 discusses the literature about the two conflicting sides of using SRI screens. Moreover, it also provides a theoretical link between the characteristics of a firm and its LGBT-friendliness. Section 3 describes the data used. Section 4 provides the models used in the analytical part. The final results are presented in section 5. Finally, concluding remarks and suggestions for future research can be found in section 6.
6
2. THEORETICAL DISCUSSION
This section views the contradicting theories behind the overall use of SRI screens and their findings. Furthermore, it discusses the effect of LGBT-supportive policies on employees and firm performance. At the same time, it also explains the reasoning behind the decision to implement these policies. This provides us a better understanding why investing solely in LGBT-friendly firms could affect investor’s return in the first place.
2.1 Negative effects of socially responsible investing
Investors that uses SRI screens in their investment strategies don’t only look at returns, but also take other factors into account. For example, a firm, that produces high returns at the cost of its employees and other stakeholders, will not meet the criteria that a socially responsible investor imposes on its portfolio.
Traditional investment theories argue that investors are being restricted during their stock selection process when reckon with social responsibility (Derwall & Koedijk, 2009). As a result, investors can only choose from a limited number of stocks to be included in their portfolio. Putting a constraint on an investment opportunity set prevents an investor to choose its optimal portfolio. Using SRI screens could only lead to sub-optimal choices. This mindset stems from the theory developed by Markowitz (1959).
According to Markowitz (1959), investors are risk-averse. All they want is earning a high return with the lowest possible risk attached to it. A portfolio is efficient if it maximize its returns given a level of risk or minimize its risk level given a rate of return. In other words, an efficient portfolio can’t do better without accepting a higher risk or lower return. Every efficient portfolio is located on the efficient frontier.
Every investor has an indifference curve. Each point on the indifference curve is a combination of risk and return that yield he same utility to the investor. The goal of the investor is to maximize its utility by moving to a higher indifference curve. The point where the efficient frontier and the indifference curve meet is the point of the optimal portfolio. This process will be disturbed by putting a constraint to its portfolio (Derwall & Koedijk, 2009). And thereby,
eliminate its efficiency which further leads to sub-optimal choices.
The theory of Markowitz (1959) is also being supported by Hong and Kacperczyk (2009). They studied the effect of not investing in sin stocks because of moral norms. Sin stocks are stocks that are related to the weapon, alcohol and tobacco industry. They find that the decision of not investing in these stocks would lower its investor’s return.
Other thoughts about firms being socially responsible is that it comes at a price. The reasoning behind this is that firms will allocate the cost of being ethical to its consumers. As a
7 result, creating competitive disadvantages and lower profitability (Walley & Whitehead, 1994). This indicates that socially responsible firms could harm shareholders value.
2.2 Positive effect of SRI screens
Empirical studies have shown that not all SRI screens negatively affect portfolio returns. Edmans (2011) constructed a portfolio using employee satisfaction as his SRI screen. After controlling his portfolio return for risk using the Carhart (1997) 4-factor model, he concludes that his portfolio indeed earned an abnormal return. Even after controlling for industry risk and firm characteristics.
Another study conducted by Derwall et al. (2005) finds that a portfolio of highly ranked eco-efficient companies outperformed its lower ranked counterpart. The difference remained significant even after controlling for the market, industry-specific factors, investment style and transaction costs.
These findings support the alternative theory about the effect of SRI screens on investor’s return. This theory is called Stakeholder Theory and is developed by R. Edward Freeman (1984). Freeman (1984) suggests that firms distinguish themselves by expressing socially responsible behavior. This behavior will positively affect the firm’s relationships with its stakeholders and yield significant positive effects on future profitability. These firms gain long term reputational benefits which are often overlooked by investors who are focused on short-term performance (Derwall & Koedijk, 2009).
According to traditional investment theory, the available investment opportunities are limited by using SRI screens as an investment criteria. This will negatively affect investor’s return. However, Stakeholder Theory suggests that this effect will be offset. The reasoning is that investment opportunities are being filtered by using SRI screens, leaving only superior
investments (Freeman, 1984). These investments are superior, because they are focused on long term performance. Even though the available pool of investments is limited, the higher quality of companies offsets this limitation.
2.3 Effect of LGBT-supportive policies
After viewing the two sides behind SRI screens in general, the focus will now lie on the effect of implementing LGBT-supportive policies on firm performance and employees. This section provides us a better understanding of how the specific use of LGBT-equality as a SRI screen could affect investor’s return. Some studies find that these policies will benefit the firm, while others show the opposite. If these policies leads to better firm performance, than investing in LGBT-friendly companies would positively affects investor’s return. Thus, supporting the use of LGBT-equality as a SRI screen.
8 Badgett et al. (2013) has reviewed a number of studies related to corporate policies and LGBT employees. Some of their findings were less discrimination and a better workplace environment. The implementation of LGBT-supportive policies limits discrimination practices by including nondiscrimination policy based on sexual orientation and gender identity (Badgett et al., 2013). A number of studies has shown that these policies can create a supportive
workplace environment where employees feel comfortable to work in. Because, these policies lead to more openness about being LGBT. It creates a climate where LGBT employees are more likely to disclose their sexual orientation and gender identity (Badgett et al., 2013).
They’ve also found a connection between having LGBT-supportive policies and improved employee’s health outcomes. Their explanation was that the psychological health of the
employees increased due to less discrimination, more openness and a supportive surrounding. This outcome is also being supported by the Human Rights Campaign (2009) survey. They found that employees who openly expressed their sexual orientation in their workplace were less likely to feel depressed, distracted, exhausted and less likely to avoid social events than employees from companies that doesn’t have these policies installed (Badgett et al., 2013).
Other findings were increased job satisfaction and commitment. This could be explained by the social exchange theory (Blau, 1964). This theory is based on reciprocity. The
fundamentals behind this theory is that one should help and not harm the other who have helped them. Employees working for LGBT-friendly companies will feel appreciated and supported by the firm. This could lead to reciprocal behavior in the form of greater job commitment and productiveness (Wang & Schwarz, 2010).
Wang and Schwarz (2010) have another theory, which expects a positive relationship between LGBT-supportive policies and firm performance. Their reasoning is that firms exploiting these policies create a better corporate image. It signals that the firm is concerned about the wellbeing of its employees and is providing a protective environment. Their
reputation against LGBT workers will also improve. This image may attract more qualified and higher educated people to work for the firm. It also stimulates new business relationships and opportunities with businesses that are looking for some sort of association with high reputation organizations. Moreover, this image could also create opportunities to access the LGBT
consumer market. LGBT consumers are more likely to make a purchase from a LGBT-friendly company.
However, there are also some negative sides to the implementation of LGBT-supportive policies. Johnston and Malina (2008) indicate that a significant portion of the US population still views homosexuality and LGBT people as immoral. Consumers and/or investors holding these negative thoughts might restrain themselves from LGBT-friendly companies, as it is against their
9 beliefs. The same also implies for employees, as some of them might not want to work with LGBT people.
Another concern is that the implementation of these policies comes at a substantially high cost (Johnston & Malina, 2008). A part of LGBT-supportive policies is providing domestic partner benefits. This implies that the partner of a LGBT employee is being recognized as a legal partner. Thereby, he or she receives the right to obtain benefits just like any other partner of a non-LGBT employee. In the United States, there is a possibility for employees to cover their partners under the health insurance plan of its employer. This will lead to an increase in company’s health insurance costs for its employees by also providing this benefit to the LGBT minority group.
Even though these policies are costly, it will be offset by the benefits from less discrimination and improved employee’s health outcomes. Enclosure of nondiscrimination policies may bring the company more in-line with federal or state regulations (Wang & Schwarz, 2010). Hence, it limits the probability of lawsuits and legal costs. Improvements in employees’ health lowers the healthcare expenditures and improves productivity.
3. DATA
This section describes the data used for the analysis part. First, there will be a brief explanation about the essence of the Human Rights Campaign’s Corporate Equality Index reports. Followed by a description on the formation of the sample portfolio’s used further in this study.
3.1 HRC’s Corporate Equality Index (CEI) reports.
My main data source are from the Corporate Equality Index (CEI) reports retrieved from the Human Rights Campaign (HRC) website. Since 2002, the HRC publish a yearly CEI report which rates companies on their policies and practices against LGBT employees. These are the same reports used by Credit Suisse in the formation of their LGBT-equality portfolio.
Every year, the HRC sends out a CEI survey to previous and prospective respondents. The surveys are sent to the chief executive officer of the firm, as well as the highest level executive responsible for the human resource department (Human Rights Campaign, 2014). Questions in these surveys are related to topics such as non-discrimination policies and equal employee benefits. Based on how well the company does in each field, it can earn a CEI score between zero and 100. A score of 100 means that the firm is highly LGBT-friendly. An important notice is that the HRC is not affiliated with the companies who responds to the surveys, else it may have incentives to manipulate the results.
10 Another factor to be considered is that the information gathered by these surveys is not objective enough. A firm who is not LGBT-friendly can provide false answers or it may decide not to respond to the survey, because they expect to earn a low CEI score which could harm the firm’s reputation. To control for this factor, the HRC also collects publicly available information and cross-check for any false assumptions, before giving the company its official rating.
Companies who repeatedly decide not to respond to the survey will receive an unofficial rating based on their findings from other information resources. Thereby, the CEI reports includes reliable, objective information and CEI scores of both LGBT- and non-LGBT-friendly firms.
3.2 Portfolio i: only LGBT-friendly firms
The portfolio of companies that was created by Credit Suisse is not publicly available. Based on the CEI reports described above, I form my own portfolio of only LGBT-friendly firms which is named “portfolio i”. The criteria I use to determine if a company is LGBT-friendly or not is being based on their CEI score. Companies with a CEI score between 80 and 100 are considered LGBT-friendly. This is the same criteria used by Credit Suisse. Therefore, the portfolio only consists of companies that consistently satisfied the CEI score criteria between the period 2010 and 2014. This means, for example, that a firm which scored a CEI of 90 in year 2014 will not be included if it scored a CEI of 70 in year 2010. The reason behind this is that I want to mimic the portfolio created by Credit Suisse as much as possible. Moreover, the main focus will be on firms that are listed in the S&P 500. This yields a total of 118 companies who are marked as LGBT-friendly.
The monthly returns of this portfolio will then be found in database CRSP. The returns are from January 2005 to December 2014. CRSP makes a distinction between two weighting methodologies: value- and equally-weighted returns. Value-weighted returns are weighted by their proportion of market capitalization. Returns of firms with a relatively high market cap will receive a higher weight than firms with a lower market cap. I construct both value- and equally-weighted portfolio’s to make sure that my findings are not influenced by different weighting methodologies.
3.3 Portfolio p: LGBT- and non-LGBT-friendly firms
For the second part of the analysis, I want to investigate whether there is a relation between firm characteristics and the level of LGBT-friendliness of the company. Therefore, I will need a portfolio of LGBT and non-LGBT firms. This portfolio will be named “portfolio p”. It consist of 118 LGBT-friendly companies as described above and 121 non-LGBT companies. These figures represent approximately the top 15% and bottom 15% of the companies who received a CEI score in the CEI report of 2014, respectively. All of these companies are listed in the S&P 500. I use database CRSP and COMPUSTAT to search for firm characteristics of these companies.
11 Studies related to firm characteristics and LGBT is limited. Based on my own intuition and studies related to SRI, I’ll identify some factors that might provide a link.
Firms with high market cap tend to care a lot about their reputation. They are more likely to express their socially responsible awareness. This could be reasoned by the fact that irresponsible behavior could damage its reputation, and thereby hurting its share price and firm value.
Verwijmeren and Derwall (2010) find that companies that are socially responsible tends to have a low debt-to-equity ratio. It wants to avoid bankruptcy costs and signal to the market that it has a good financial reputation.
Derwall and Koedijk (2009) explained that the social responsibility expressed by firms could lead to intangible sources for superior performance. I split these intangibles into two categories: intangible assets and goodwill. Intangible assets are related to things like patents, licenses and trademarks. Goodwill is associated with brand image and reputation.
Firm size also plays a role on the decision of whether to implement LGBT-supportive policies or not. Companies with a lot of employees might want to express their concerns about their employees and willingness to adapt. Firms, who supports LGBT, strives for diversity and hopes that its actions could lead to greater job satisfaction and commitment.
4. MODEL
This section describes the models used for our analysis. First, there will be a focus on the performance evaluation method of portfolio i. The second part will provide a description of the method used to determine if there is a relation between firm characteristics and the level of LGBT-friendliness of a firm.
4.1 Performance evaluation method of portfolio i
The first implication of this thesis is to find if there is a positive relation between the use of LGBT-equality as a SRI screen and investor’s return. It will be tested using the following two models:
CAPM: 𝑅𝑅𝑖𝑖𝑖𝑖 − 𝑅𝑅𝑓𝑓𝑖𝑖 = 𝛼𝛼 + 𝛽𝛽𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑖𝑖+ 𝜀𝜀𝑖𝑖𝑖𝑖
Carhart: 𝑅𝑅𝑖𝑖𝑖𝑖 − 𝑅𝑅𝑓𝑓𝑖𝑖 = 𝛼𝛼 + 𝛽𝛽𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑖𝑖+ 𝛽𝛽𝑆𝑆𝑀𝑀𝑆𝑆𝑆𝑆𝑀𝑀𝑆𝑆𝑖𝑖+ 𝛽𝛽𝐻𝐻𝑀𝑀𝐻𝐻𝐻𝐻𝑀𝑀𝐻𝐻𝑖𝑖+ 𝛽𝛽𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑖𝑖+ 𝜀𝜀𝑖𝑖𝑖𝑖
The CAPM model is mostly used in studies to measure mutual fund performance. It uses only the market proxy to explain the variation of stock returns. However, recent literature (e.g. Fama and French, 1993) has raised some questions about the use of the CAPM model and
12 suggests that the use of a multi-factor asset-pricing model leads to a better explanation of
mutual fund performance (Bauer et al., 2005, p. 1760).
Therefore, I will also make use of the Carhart (1997) 4-factor model. This model is an extension of the Fama and French (1993) 3-factor model. It explains the variation of portfolio returns using additional risk proxies, other than the market proxy in the CAPM model. The Carhart (1997) model takes into account that certain investment strategies might explain some of the variation of portfolio returns (Bauer et al., 2005, p. 1760). These investment strategies includes investing in small cap versus large cap portfolio (SMB), growth stock versus value stock portfolio (HML) and winners versus losers portfolio (MOM).
There will be two regressions. First with the CAPM model and second with the Carhart (1997) 4-factor model to make sure that any outperformance of LGBT-friendly firms is not due to risk. The dependent variable is the value-weighted or equal-weighted portfolio return less the risk-free rate. The intercept α captures the abnormal return adjusted for market (MKT), size (SMB), value (HML) and momentum (MOM) risk. The variable of interest is the alpha. A positive alpha tells you how good the portfolio did compared to the market and a negative alpha tells you the opposite. The factors 𝑅𝑅𝑓𝑓𝑖𝑖 , MKT, SMB, HML and MOM are taken from Ken French’s website.
4.2 Logistic regression model to account for firm characteristics
The second implication of this thesis is to test whether there is a link between firm
characteristics and the LGBT-friendliness of a firm. The logistic regression model can be applied to test for this type of question. In this case our dependent variable is binary, namely if the firm is LGBT-friendly or not. Our independent variables are the firm characteristics described in section 3.3. The model is as follows:
LGBT = 𝛽𝛽0+ 𝛽𝛽1𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 + 𝛽𝛽2𝑚𝑚𝑒𝑒𝑒𝑒𝑒𝑒𝑚𝑚𝑒𝑒 + 𝛽𝛽𝑑𝑑𝑚𝑚𝑑𝑑𝑚𝑚 3𝑒𝑒𝑖𝑖𝑚𝑚𝑚𝑚𝑖𝑖𝑖𝑖𝑒𝑒𝑑𝑑𝑖𝑖𝑚𝑚 + 𝛽𝛽4𝑖𝑖𝑔𝑔𝑔𝑔𝑑𝑑𝑔𝑔𝑒𝑒𝑖𝑖𝑖𝑖 + 𝛽𝛽5𝑚𝑚𝑚𝑚𝑚𝑚𝑖𝑖𝑔𝑔𝑒𝑒𝑚𝑚𝑚𝑚𝑒𝑒 + 𝜀𝜀𝑖𝑖𝑖𝑖
The dependent variable, LGBT, can only take the value 1 if the firm is LGBT-friendly and 0 if it is not. Considering the fact that each of the independent variables are continuous, we must first divide each of them into three categories: 1=low; 2=mid; 3=high. Now, the model becomes easier to interpreted.
The beta coefficients represents the logistic odds ratio’s. The interpretation of these beta’s are not the same as in ordinary least squares (OLS) models. Beta’s in OLS models states that a unit increase of independent variable X will lead to a β increase in dependent variable Y. This doesn’t applies to logistic models. In this case, the independent variables are divided into three categories. An increase of one unit in X means a shift from one category to another.
For example, when a company goes from low market cap (category 1) to a mid-market cap (category 2) than the odds of becoming more LGBT-friendly is β times higher relative to a
13 low market cap firm. If the odds ratio (β) is smaller than 1, than this would mean that the odds of LGBT=1 decreases.
5. EMPIRICAL RESULTS
This section provides a summary of the empirical results found through the regression models described in the previous section. It will also provide an explanation on how to interpret the results.
5.1 Performance evaluation results: CAPM and Carhart 4-factor model
Table 1 presents a summary of the core results found for the entire 2005-2014 period. Panel A contains the results from the CAPM model and Panel B from the Carhart (1997) model. The dependent variable is the value-weighted or equal-weighted portfolio return less the risk-free rate. The returns of the portfolio are winsorized at the xth and (100-x)th percentile across the
sample period.
Table 1: Risk-adjusted returns period 2005-2014
x = 0 x = 5 x = 10
Value Equal Value Equal Value Equal
Panel A: CAPM model 𝛼𝛼 𝛽𝛽𝑀𝑀𝑀𝑀𝑀𝑀 -0.03 (-0.96) 1.00 (128.20)*** -0.10 (-0.57) 0.94 (29.39)*** 0.08 (0.97) 0.85 (43.60)*** 0.13 (0.81) 0.89 (24.62)*** 0.29 (2.88)*** 0.69 (30.61)*** 0.32 (2.06)** 0.70 (19.90)*** R2 0.99 0.88 0.94 0.84 0.89 0.77
Panel B: Carhart model 𝛼𝛼 𝛽𝛽𝑀𝑀𝑀𝑀𝑀𝑀 𝛽𝛽𝑆𝑆𝑀𝑀𝑆𝑆 𝛽𝛽𝐻𝐻𝑀𝑀𝐻𝐻 𝛽𝛽𝑀𝑀𝑀𝑀𝑀𝑀 -0.03 (-0.87) 1.00 (106.86)*** -0.01 (-0.58) -0.01 (-0.53) -0.02 (-1.72)* -0.03 (-0.25) 0.79 (29.41)*** 0.22 (8.65)*** -0.02 (-0.68) -0.18 (-7.32)*** 0.07 (0.85) 0.86 (37.27)*** -0.01 (-0.17) 0.03 (0.76) 0.05 (2.31)** 0.15 (1.14) 0.77 (21.28)*** 0.47 (7.35)*** -0.04 (-0.05) -0.04 (-1.20) 0.27 (2.77)*** 0.71 (26.49)*** 0.04 (0.88) -0.01 (-0.13) 0.06 (2.47)** 0.33 (2.52)*** 0.60 (17.05)*** 0.46 (7.38)*** 0.01 (0.23) 0.02 (0.60) R2 0.99 0.94 0.94 0.89 0.90 0.84 #obs 120 120 120 120 120 120 Notes:
*: significant at the 10% level; **: significant at the 5% level; ***: significant at the 1% level The sample period is January 2005 – December 2014. t-Statistics are in parentheses.
14 Column x=0 states the results without excluding any outliers. It shows that the alpha is slightly negative, but not significantly different from zero. This outcome is the same for both models and weighted portfolio’s. This indicates that portfolio i didn’t performed better or worse than the market.
Column x=5 presents the results after winsorizing at the 5th and 95th percentile. In this
case, the alpha is slightly positive, but also not significantly different from zero. This leads to the same conclusion as before, that portfolio i didn’t under- or outperformed the market.
Column x=10 generates the results after winsorizing at the 10th and 90th percentile. It
shows that the alpha is positive and significantly different from zero in both models. The CAPM model predicts an alpha of 0.29 for the value-weighted returns. This means that the portfolio performed 0.29% higher than the monthly market return. This would yield to a yearly return of 3.48% above the market. The same goes for the Carhart (1997) model which predicts an alpha of 0.27, indicating a yearly return of 3.24% above the market. Both alpha’s are significant at the 1% level.
The alpha is also significant when using the equally-weighted portfolio returns. The CAPM model predicts an alpha of 0.32 and the Carhart (1997) model predicts a value of 0.33. The interpretation is the same as above. The portfolio generates a yearly return of 3.84% or 3.96% above the market, respectively. The alpha is significant at the 5% level when using the CAPM model and significant at the 1% level when using the Carhart (1997) model.
From the table, we can also see that the 𝛽𝛽𝑀𝑀𝑀𝑀𝑀𝑀 coefficient is significant in every case. This
makes sense since all of the companies included in portfolio i are all listed in the S&P 500. The 𝛽𝛽𝑆𝑆𝑀𝑀𝑆𝑆 coefficient is also significant when accounting for equally-weighted returns. This
coefficient suggests that some of the variation of the abnormal return is obtained by investing in stocks with a low market cap. The βMOM coefficient is also significant in some cases. This
coefficient indicates that some variation of the abnormal return is obtained by investing in winner stocks. Winner stocks are stocks that performed well during a twelve month period.
These above results indicates that, when controlling for outliers, investing in LGBT-friendly companies can yield a higher return than the market.
5.2 Logistic regression model results
Table 2 summarize the results from the logistic regression model. Note that the coefficients presented here has a different interpretation than with the linear regression model presented above. The odds ratio represent the odds of a firm becoming more LGBT-friendly (LGBT=1) when the independent variable increases by 1 unit. Keep in mind that the independent variables are divided into three categories: 1=low; 2=mid; 3=high. An increase by 1 unit means a shift from one category to another.
15 From the table, we can see that every variable is significant. LGBT-friendly firms tend to have high market cap, debt/equity ratio, goodwill and employees. Since the odds ratio of all these variables are larger than 1. When a company goes from low market cap to a mid-market cap than the odds that the firm is LGBT-friendly, is 6.88 times higher relative to a low market cap firm. The same goes for the deb-to-equity ratio. When a firm has a high debt-to-equity ratio than the odds that the firm is LGBT-friendly is 2.90 times higher. This also implies for the variables goodwill and employees.
A higher market cap, goodwill and number of employees are in line with our
expectations of the firm characteristics that a LGBT-friendly firm would have as described in section 3.3. However, the expectation was that socially responsible firms would also have a lower debt-to-equity ratio’s. But the opposite is true in this sample.
Another notice is that LGBT-friendly firms tend to possess less intangible assets. When a firm has more intangible assets than the odds that the firm is LGBT-friendly is 0.22 times lower than a firm with less intangible assets.
6. DISCUSSION
The empirical results found in section 5 shows that the alpha is significant when controlling for outliers, market (MKT), size (SMB), value (HML) and momentum (MOM) risk. This indicates that the use of LGBT-equality as a SRI screen yields a positive effect on investor’s return. This
Table 2: Logistic regression results
LGBT Odds ratio Constant Market cap Debt/equity Intangible assets Goodwill Employees 0.02 (-7.28)*** 6.88 (6.17)*** 2.90 (4.47)*** 0.22 (-2.11)** 5.26 (2.42)** 2.07 (3.11)*** # of obs. 235 Pseudo R2 0.386 Notes:
*: significant at the 10% level; **: significant at the 5% level; ***: significant at the 1% level; z-Statistics are in parentheses.
16 strategy could earn the investor an abnormal return between 3.24% and 3.96% per year. The results also show that investing in a LGBT-friendly portfolio doesn’t harm investor’s return even when outliers are not controlled for or being winsorized at a small rate (x=5%). The alpha is not significantly different from zero in these situations. Which suggests that this portfolio didn’t under- or outperformed the market. This study provides supportive evidence that the use of SRI screens is not harmful for investor’s return.
These findings are not in line with the theory of Markowitz (1959). Putting a constraint to the available choice set of stocks doesn’t necessarily harm investor’s return. As a matter of fact, the worst case possible would be that it yields no different returns than the market based on our findings. This suggests that investing in the LGBT-fund created by Credit Suisse would be an investment strategy to consider. It allows investors to expect a return equal to the S&P 500 while also only investing in LGBT-friendly companies.
With the use of the logistic regression model, this study also finds some firm
characteristics that were linked to the LGBT-friendliness of a firm. In correspondence with our intuition and main findings in existing literature: LGBT-friendly firms tend to have a high market cap, goodwill and large firm size. These factors might explain the performance of these
companies which require some further research. However, the expectation was that LGBT-friendly firms would also have a lower debt-to-equity ratios. This might be explained by the possibility that LGBT-friendly firms require higher debt to fulfil his socially responsible job. It could also be that LGBT-friendly firms tend to be in industries that require high debt-to-equity ratios. Another point of interest is that LGBT-friendly firms tend to have less intangible assets. This might be explained by the fact that intangible assets are more to be found in technological industries. This could indicate that technological industries are less likely to incorporate LGBT-supportive policies into their corporate culture.
However, this study does have some limitations. First, we failed to take fund fees into account. If Credit Suisse imposes a high fee on its fund, then this would negatively affect
investor’s return. Thereby, it increases the chance of making a loss. Second, control variables for industry risks are not included in the model, which could affect our findings. There is a chance that some LGBT-friendly firms are from industries that happens to enjoy some major up- or downturn which affects their monthly returns. Suppose that a relatively large portion of LGBT-friendly firms are from industries that experienced a boom in our sample period which resulted in high returns. Then, the outperformance of our portfolio is more affected by industry risks rather than the LGBT-friendliness of a firm. Third, the R2 in our logistic regression model is
relatively small. This means that our model is affected by omitted variable bias. This endogeneity problem can be solved by including more control variables. These limitations, discussed above, provide some ground suggestions for future research.
17
7. REFERENCES
Badgett, M.V.L., Durso, L.E., Kastanis, A., Mallory, C. (2013) The Business Impact of LGBT- Supportive Workplace Policies. Retrieved from The Williams Institute website: http://williamsinstitute.law.ucla.edu/wp-content/uploads/Business-Impact-of-LGBT-Policies-May-2013.pdf
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. Blau, P. M. (1964). Exchange and power in social life. New York: Academic Press.
Carhart, M. M. (1997). On persistence in mutual fund performance. Journal of Finance 52, 57–82. Derwall, J., Guenster, N., Bauer, R., Koedijk,K. (2005). The eco-efficiency premium puzzle.
Financial Analysts Journal 61, 51–63.
Derwall, J., & Koedijk, K. (2009). Socially responsible fixed-income funds. Journal of Business Finance & Accounting, 36(1) & (2), 210–229.
Edmans, A. (2011). Does the stock market fully value intangibles? Employee satisfaction and equity prices. Journal of Financial Economics, Volume 101, Issue 3, Pages 621–640. Fama, E.F., French, K.R. (1993), “Common Risk Factors in the Returns on Stocks and Bonds”,
Journal of Financial Economics 33: 3-56.
Freeman, R.E. (1984). Strategic Management: A stakeholder approach. Boston: Pitman. ISBN 0- 273-01913-9.
Hong, H., Kacperczyk, M. (2009). The price of sin: the effect of social norms on markets. Journal of Financial Economics 93, 15–36.
Human Rights Campaign (2009). Degrees of equality: A national study examining workplace climate for LGBT employees. Retrieved from Human Rights Campaign Web site: http://www.hrc.org/files/assets/resources/DegreesOfEquality_2009.pdf. Human Rights Campaign (2014). Corporate equality index: On gay, lesbian, bisexual and
transgender social responsibility. Retrieved June 2, 2015, from www.hrc.org/documents/2014CEIReport.pdf
Johnston, D., & Malina, M. A. (2008). Managing sexual orientation diversity: The impact on firm value. Group and Organization Management, 33(5), 602–625.
Markowitz, H. (1959). Portfolio Selection : Efficient Diversification of Investments. John Wiley & Sons, New York.
Verwijmeren, P., Derwall, J. (2010). Employee well-being, firm leverage, and bankruptcy risk. Journal of Banking and Finance 34, 956-964.
18 Walley, N., & Whitehead, B. (1994). It’s not easy being green. Harvard Business Review, Vol. 72,
No. 3, 46-52.
Wang, P., & Schwarz, J.L. (2010). Stock price reactions to GLBT nondiscrimination policies. Human Resource Management, 49, 195– 216.
