The Sin Stock Anomaly
ByLevi Milan Jano1
MSc Finance Thesis
Focus area: Sustainable Society
Supervisor: Dr. A. Dalò January 11th, 2021
1Contact address: l.m.jano@student.rug.nl; student number: S2523957; course: Master's Thesis Finance;
2
Abstract
This research examines the manifestation of groupthink, measured by the dimension of collectivism, as a cultural driver behind investment in ‘sin’ stocks. Specifically, this research addresses the triumvirate of sin that consists of firms operating in alcohol, gaming, and tobacco industries. Time series analyses are conducted to test for the existence of pricing anomalies in sin stock portfolios, and whether these risk-adjusted returns can be fully captured by the risk factors of investment and profitability. I analyze self-constructed sin stock portfolios using conventional asset pricing models in a global sample of 25 countries during 1997-2019. I find evidence of a sin stock anomaly but no evidence for attributing groupthink to the risk-adjusted returns of sin stock portfolios. Nonetheless, I show that results are subjective to region and portfolio construction methodology.
3
Acknowledgment
4
Section I
Introduction
Fueled by abundant scientific evidence and the spread of information, climate change now concerns a global audience. Loss of sea ice, rising water temperatures, and longer and more intense heat waves, as predicted by scientists in the past, are now omnipresent (see IPCC, 2007; IPCC, 2013). Moreover, today’s world is characterized by news that can rapidly spread, or go
viral, through the advancements and the likes of smartphones, social media, and the internet.
Action against climate change has been embraced by a plethora of national governments. This has resulted in the Paris Climate Agreement (2015) to halt global warming below 2°C above pre-industrial levels. However, as Mark Carney, former governor of the Bank of England, eloquently put forth in his speech, titled addressing Lloyd’s of London (Bank of England, 2015): “…the catastrophic impacts of climate change will be felt beyond the traditional horizons of most actors – imposing a cost on future generations that the current generation has no direct incentive to fix”. In other words, resolving the issue of climate change might require a need for long-term policies that are beyond the long-term orientation of policymakers and the business cycle.
Evidently, the physical effects of climate change pose risk to investors. Besides positive effects on climate change mitigation, climate change interventions involve transitional risk for investors as well. To address both the physical and transitional risk of climate change, investors can explore and employ socially responsible investing (SRI) strategies. SRI combines the investor’s objective of creating financial returns with a positive societal impact considering environmental, social and corporate governance (ESG) criteria. According to the US Social Investment Forum (2018), more than a quarter of assets under management in the US find their origin in SRI. With an eighteen-fold expansion between 1995-2018 and a compounded annual growth rate (CAGR) of 13.6 %, SRI is associated with an ascending growth path. Furthermore, between 2016 and 2018, SRI in the US increased by more than 36% from $8.7 trillion to $11.8 trillion. Globally, SRI has followed corresponding growth trends with a 34% increase to $30.7 trillion between 2016 and 2018.
5
whilst providing financial returns to ensure financial stability. Prior literature has shown that companies play an important role to reduce greenhouse gas emissions to meet the Paris Climate Agreement target (Krabbe et al., 2015). Therefore, SRI, particularly in the environmental dimension, might provide a convenient tool to break the tragedy of the horizons and, as such, might aid climate change mitigation and other ESG issues that require long term orientation. The empirical relation between the impact of Corporate Social Responsibility (CSR) and SRI on stock performance has been an ongoing debate since the first publication by Bragdon and Marlin (1972) more than 48 years ago. Dam and Scholtens (2015) propose a strong theoretical foundation for a positive relation between CSR and financial performance conditional on the measure of financial performance. Correspondingly, Revelli and Viviani (2015) conclude that globally there is no real cost or benefit to SRI investments, but that financial performance depends on the methodological choices of authors or the capability of fund managers to deliver financial returns. Nonetheless, there has been growing evidence in the literature suggesting a positive relation between CSR and corporate financial performance or stock performance (Margolis, Elfenbein, and Walsh, 2009; Friede, Busch, and Bassen, 2015). This positive relation presents a rational foundation for a doing well (financially) by doing good proposition. If undertaken successfully, SRI has the potential to align financial concerns of shareholders and ESG concerns of all stakeholders. On the contrary, a growing body of research suggests that SRI investors are missing out on financial returns by excluding investments in controversial stocks. These controversial stocks are often labeled ‘sin’ stocks, or less commonly, as ‘vice’ stocks or ‘shunned’ stocks. The economic rationale for the underperformance is as follows. By excluding sin stocks in their portfolios, SRI investors constrain their investable universe. Mean-variance analysis tells us that an optimal portfolio with constraints will lead to lower risk-adjusted returns compared to an optimal portfolio without constraints (Markowitz, 1952). Per definition, the unscreened sin stocks will exhibit higher risk-adjusted returns. Hong and Kacperczyk (2009) indeed show that US sin stock portfolios exhibit a higher risk-adjusted return compared to a control group of stocks with similar characteristics during 1965-2006. They attribute this investor behavior to norm-constrained investing which involves discriminatory practices against certain controversial stocks according to social norms. Attitudes towards what is sinful might have a substantial impact on investor choices vis-à-vis the financial performance of sin stocks. For example, can smoking be regarded sinful in China when more than half of all men are tobacco smokers compared to 15.6% in the US (CDC, 2018; WHO, 2018)? Hofstede (1984) argued there is a difference between individualistic and
collectivistic countries. Here, individualism refers to a sense of self-being and independence
6
The interest in sin stocks is not only limited to the academic literature. Most prominently representing sinful investments is the Vitium Global Fund, formerly known as the Vice Fund. This mutual fund invests at least 80% of its holdings in equity securities of companies with notable earnings in sin industries such as alcohol, tobacco, gaming, and aerospace/defense. Since its inception in 2002 up until 2017, the fund has reportedly outperformed the market by an average annual return of more than 2%. The Vitium Global Fund posted an average annual return of 13%, against the MSCI World 11%. However, the fund has had a calendar year-to-date loss of 14.10%, against a MSCI World loss of 5.5% as of June 30, 2020 (USA Mutuals, 2020). Losses were primarily obtained due to COVID related setbacks concerning long positions in casino stocks. Hence, no conclusions can be drawn about the long-term development of the fund a posteriori.
In the interest of my study I test whether sin stocks outperform their respective market indices. For instance, in the US, whether a portfolio consisting of the triumvirate of sin stocks (sin portfolio) listed in the US outperforms a value-weighted US market portfolio. Hence, exhibiting positive risk-adjusted returns. Additionally, given that sin portfolios might exhibit positive alphas, I will study whether asset pricing models, such as the capital asset pricing model (CAPM - Lintner 1965; Sharpe 1964; Mossin, 1966), the Fama and French three and five factor model (1993, 2015), and the Carhart four factor model (1997), explain well the pricing structure of sin stock returns.
Sin portfolios will be constructed on the country level as well as the group level of countries that share cultural values. Cultural values determine the level of groupthink measured by a country’s level of collectivism. To my knowledge there are no studies concerning Japan, Hong Kong, and Emerging Market countries that analyze country specific drivers for sin stock returns. Emerging Market countries are spread-out throughout the world and share similar levels of economic development yet may vastly differ in levels of the individualism index. Therefore, the interest of my thesis extends to whether collectivism, as a cultural driver for groupthink, influences the performance of sin portfolios. My thesis aims to answer the following main research question:
Is groupthink a driver for the risk-adjusted returns of ‘sin’ stock portfolios associated with smoking, drinking, and gambling?
7
countries in Emerging Markets.2 These are countries that exhibit measurable cultural
differences and for which risk factors are accessible at the Kenneth R. French website3. While
the US is widely studied in the sin stock literature, no research on groupthink and sin stock returns has been performed on Japan, Hong Kong, and Emerging Market countries. My findings indicate that some sin portfolios exhibit positive risk-adjusted returns that are not fully priced or captured by the conventional asset pricing models. In line with Fabozzi et al. (2015) I find that risk-adjusted returns in a sin portfolio of (mostly) US stocks are completely priced away by the investment and profitability risk factors. Finally, ascribing groupthink as a driver for sin stock performance yields conflicting results. Moreover, my findings seem to be subjective to region and methodology.
The remainder of this master thesis proceeds as follows. In section II, I provide a background of the literature on sin stocks and how I identify sin stocks for the analysis. In section III, I explain the methodology and specify the variables used in the regression analyses. In section IV, I present empirical results. I conclude in section V.
Section II
Background of the Literature
2.1 ‘Sin’ stocks
Sin has been defined as “an offense against religious or moral law” (Merriam Webster, n.d.). What defines sin needs to be interpreted with great caution because what is considered respectable behavior in one country or community might be frowned upon in others. Consequently, I do not make value judgments regarding what is sinful or immoral but contend that investors use sin or vice as a criterion to screen investments (see, e.g., Capelle-Blancard and Monjon, 2014; Renneboog, Ter Horst, and Zhang, 2008; Trinks and Scholtens, 2017). Hence, it is imperative that results need to be interpreted specific to country, religion, culture, and time (e.g., Durand et al., 2013; Fauver and McDonald, 2014; Hong and Kacperczyk, 2009; Salaber, 2007). Other articles have included controversial industries such as the adult entertainment, nuclear power, and defense (e.g. Fabozzi, Ma, and Oliphant, 2008; Kim and
2 Countries that belong to emerging markets are taken from the Kenneth R. French website and include:
Argentina, Brazil, Chile, China, Colombia, Czech Republic, Egypt, Greece, Hungary, India, Indonesia, Malaysia, Mexico, Pakistan, Peru, Philippines, Poland, Qatar, Russia, Saudi Arabia, South Africa, South Korea, Taiwan, Thailand, Turkey, United Arab Emirates.
3 I want to thank Kenneth R. French for the availability of the dataset on his website:
8
Venkatachalam, 2011; Lobe and Walkshäusl, 2016). Nonetheless, I focus on the triumvirate of sin. Firstly, because smoking, drinking, and gambling can be associated with sinful behavior in culture and religion, whereas defense is present in all developed countries and has a more neutral stance (see Salaber, 2007; Hong and Kacperczyk, 2009). Moreover, Durand et al. (2013) include defense as a sin industry in their study but find that it does not materially affect the conclusions of their research. Secondly, these activities are considered addictive and unhealthy. Excessive consumption of each of these three sin activities will lead to undesired social outcomes (see e.g. Goodchild, Nargis, and d’Espaignet, 2018; Loudouceur, Boisvert, Pépin, Loranger, and Sylvai, 1994; Maynard and Kennan, 1981).
The sin stock literature primarily concerns studies on the financial performance of sin stocks. What defines financial performance is subjective and depends on methodology choices. In the sin stock literature, most authors define financial performance as stock returns (e.g. Blitz and Fabozzi, 2017; Fabozzi et al., 2008; Hong and Kacperczyk, 2009; Liston, 2016; Richey, 2014; Salaber, 2007), whereas others included equity valuation (Fauver and McDonald, 2014). Further studies have used alternative measures, such as Tobin’s Q to test the operational performance of sin stocks (Visaltanachoti, Zou and Zheng, 2009; Fauver and McDonald, 2014). The literature on sin stocks primarily focuses on the United States. One paper focused on sin stocks listed in Mainland China and Hong Kong. This research focused on pre-financial crisis data between 1995-2007 and analyzed the triumvirate of sin. Both sin portfolios in Hong Kong and Mainland China outperformed their respective market indices and exhibited positive alphas (Visaltanachoti et al., 2009). Other studies took a global approach but failed to include country characteristics that may drive sin stock returns (Fabozzi et al., 2008; Fabozzi and Blitz, 2015; Humphrey and Tan, 2014; Lobe and Walkhäusl, 2016). Salaber (2007) and Visaltanachoti (2009) researched sin stocks outside of the US in respectively European countries, and China and Hong Kong. However, they do not distinguish the cultural drivers to sin beyond religion. Durand et al. (2013) covered countries in the Pacific Basin and looked into whether cultural dimensions, specifically individualism versus collectivism, influenced sin stock possession. They provided evidence that collectivism positively influenced the likelihood of owning sin stocks.
2.2 SRI strategies
9
that concerns this study is negative screening. It encompasses employing negative screens by excluding certain stocks that are involved in harmful, immoral or controversial activities based on specific ESG criteria. Trinks and Scholtens (2015) have analyzed fourteen controversial issues at the firm level, covering tobacco, alcohol, and gaming, inter alia. They find that investing in controversial stocks in most cases increased risk-adjusted returns. Conversely, excluding these stocks may reduce risk-adjusted returns. Therefore, it might be unwise to exclude stocks involved in controversial issues if the objective of the investor is to maximize financial performance, or as phrased by the aforementioned authors, that there are opportunity costs in negative screening. Nevertheless, as Renneboog et al. (2008) suggested, SRI portfolio managers might be willing to partly give up financial performance in exchange for ESG performance. Hence, if SRI investors concurrently pursue financial and social objectives future research should also concern the quantification of social objectives.
2.3 Conventional asset pricing models
Critical reflection on the outcomes of the results of this study necessitates a detailed understanding of the conventional asset pricing models and its limitations. Reflection towards other sin stock studies that employ these asset pricing models contributes to the development of the methodological choices in this study. Moreover, for the sake of brevity in subsequent sections these asset pricing models, used to price the excess returns of sin stocks, need to be addressed.
The capital asset pricing model (CAPM) captures the relation between the expected returns of assets and its systematic risk. One of the assumptions of this model is that the strong form of market efficiency holds. This implies that stock prices fully reflect all information available at all times, hence stock prices are efficient. If this is the case, then expected returns can be fully priced by the systematic and unsystematic (or idiosyncratic) risk components. Moreover, holding a well-diversified portfolio of assets can remove, or at least mitigate to a large extent, the idiosyncratic risk component. Jensen (1968) transformed the CAPM into an ex post model where the intercept, known as Jensen’s alpha, measures the risk-adjusted return above or below the prediction of the CAPM. Despite failing numerous empirical tests, the model remains popular because of its simplicity and is still central in undergraduate finance courses.
10
The FF3 performs better in explaining the cross-section of the average returns than the CAPM. However, this model is unable to explain the short-term persistence in equity mutual funds returns. Carhart (1997) attributed this phenomenon to the momentum risk factor, inter alia. The momentum strategy involves buying stocks that have performed well in the short-term past and selling stocks that have performed poorly. He extended the FF3 to the Carhart four-factor model (FFC) by including the one-year momentum factor (MOM) of Jegadeesh and Titman (1993) to mimic the momentum risk factor. Carhart found that when applying a momentum strategy of buying last year’s winners and selling last year’s losers an expected yield of 8% per annum could be attained. Of this yield, 4.6% of was attributed to the FFC. The work of Carhart did much to explain the short-term persistence in equity mutual fund returns. However, I find conflicting results in the literature regarding the significance of the momentum risk factor. This seems to be subjective to sample region (Fabozzi and Blitz, 2015). This needs to be taken in consideration because this research covers multiple regions. Moreover, the momentum risk factor plays an insignificant role in explaining variation of sin portfolio returns as opposed to virtue portfolio returns (Lobe and Walkshäusl, 2016).
In 2015, Fama and French extended the FF3 to the Fama-French five-factor model (FF5). This model includes additional risk factors for profitability (RMW) and investment (CMA). The profitability risk factor is mimicked by the difference in average return between portfolios with robust and weak profitability, while the investment risk factor is mimicked by the difference in average returns between portfolios that invest conservative and aggressively. Their rationale for including the new risk factors was that empirical evidence suggests that much of the variation in the average returns with respect to profitability and investment was left unexplained in their earlier FF3. The main limitation of this new model is its failure to capture the low average returns of small stocks that invest aggressively despite having low profitability. Furthermore, while the FF3 factors SMB and HML have economic intuition because they might capture the risk of financial distress, Fama and French (2015) do not attempt to explain whether higher average returns for firms with robust profitability and low investments are due to higher (default) risk or a manifestation of (market) mispricing.
2.4 Social norms and sin stock performance
11
by social norms. If social norms play an important role in affecting investor choices, then we should expect to find similar findings for other Western countries that share similar social norms. Following up on the finding of Hong and Kacperczyk (2009) that sin stocks experience less analyst coverage, Kim and Venkatachalam (2011) theorize that the neglect of sin stocks might be due to poorer financial reporting. Contrary to their expectation, the authors find that sin stocks have higher quality financial reporting than non-sin stocks as well as better financial performance. This is consistent with the argument that managers of sin stocks improve financial reporting to acquire a larger investment and analyst base in a country that shuns sin stocks due to social norms. The implication is that in spite of higher financial returns and superior financial reporting quality, investors are willing to shun sin stocks and give up financial performance to conform to social norms.
12
thread through sin stock performance. In order to analyze sin stock performance, I acknowledge the existence of idiosyncratic social norms in countries and communities, such as religious groups, that influence attitudes towards what is sinful.
2.5 Sin stock performance studies
Visaltanachoti et al. (2009) examined the financial and operating performance of the triumvirate of sin in China and Hong Kong using data from 1995-2007. They find that sin stocks have higher abnormal returns. In their finding they confirm that investors that do not invest in sin industries experience a financial cost. Conversely, investors in sin stocks require a premium for giving up being socially constrained. Although the authors mention that the concept of sin is treated differently in China and Hong Kong compared to the US, this research lacks to link the implications of these differences with the empirical results of the study. Likewise, Richey (2014) examined sin stocks returns for the triumvirate of sin (including defense) in the US using data from 2007-2013. His results are similar to those of Visaltanachoti et al. (2009) in that he finds an abnormal return for a given level of systematic risk. The author argues that dead firms are omitted from the dataset because the market proxy, the S&P500, also omits dead firms. However, he is the only author that made this methodological choice. Hence, this study might be subject to survivorship bias (Brown, Goetzmann, Ibottson and Ross, 1992).
Fabozzi et al. (2008) dived deeper into why sin stock’s adjusted returns might differ from the market return and attribute this to several theories. In a global study of 21 countries they provide evidence that admired stocks, in terms of reputation to investors, have lower returns than shunned stocks. As such, sin stocks are expected to have higher returns and find that sin stocks tended to outperform the market on a risk-adjusted basis. Research has shown that the Fama and French asset pricing model is not appropriate for Japan (Kubota and Takehara, 2015). Not surprisingly, they find that for Japan the one factor alpha is not significant. Blitz and Fabozzi (2015) provide similar findings for the risk-adjusted returns of sin stocks in Japan. However, in general, for a diverse sample of countries, authors find that sin stocks exhibit positive abnormal returns.This leads to my first testable hypothesis:
H1: Sin portfolios exhibit positive risk-adjusted returns
13
include nuclear power as a controversial industry in their studies of SRI performance (e.g. Derwall et al., 2010; Trinks and Scholtens, 2015). Hence, the choice to include nuclear power stocks in a sin index might not be ex nihilo but the study lacks justification for the allocation of almost half of the portfolio to nuclear power stocks. Preceding the outcomes of Lobe and Walkhäusl (2016), Durand, Koh and Limkriangkrai (2013) provide evidence that a ‘sinner’, or pure play sin portfolio, outperforms the market on a risk-adjusted basis. In contrast, they do not find underperformance for the ‘saints’, or pure play SRI portfolio.
Sin stocks are strongly tied to addictive products, such as cigarettes and alcoholic beverages. Studies have shown that in particular tobacco related goods exhibit a relatively inelastic price elasticity of demand (e.g. Chaloupka, 1999) - where price elasticity of demand is defined as the percentage change in the quantity demanded of a product that results from a 1% increase in price. It would be interesting to observe how stocks relating to these products behave over the business cycle. Salaber (2009) finds that sin stocks outperform the market during bear markets in the US. However, these stocks are just as recession-proof as some other industry-comparable stocks. In light of this study, SRI investors would miss out on returns when they exclude sin stocks based on ESG criteria but would not necessarily fare worse during recessions by investing in industry-comparable stocks.
A novel approach in the empirical analysis of sin stock returns was introduced by Liston (2015). He intertwined the realms of behavioral finance and sin stock performance by adding investor sentiment as a control variable to the asset pricing models of Fama and French (1993) and Carhart (1997). In his study he evaluated the financial performance of triumvirate of sin stocks in the US and found positive abnormal returns. Moreover, the inclusion of investor sentiment completely priced away the alphas. In a similar fashion, although through a different mechanism, Blitz and Fabozzi (2015) find that the asset pricing model of Fama and French (2015), with the inclusion of the profitability and investments premiums, completely price away alphas in sin stock returns. They conducted their research on a global sample with sub samples for Europe, Japan, and the US. The findings of Fabozzi and Blitz (2015) lead to the second testable hypothesis:
H2: Risk-adjusted returns of sin portfolios are completely priced away and captured by the risk
factors of investment and profitability
2.6 Individualism index and groupthink
14
“… a preference for a loosely-knit social framework in which individuals are expected to take care of only themselves and their immediate families. Its opposite, collectivism, represents a preference for a tightly-knit framework in society in which individuals can expect their relatives or members of a particular in-group to look after them in exchange for unquestioning loyalty (it will be clear that the word "collectivism" is not used here to describe any particular political system). The fundamental issue addressed by this dimension is the degree of interdependence a society maintains among individuals. It relates to people's self-concept: 'I' or ‘we’.” (Hofstede, 1984, pp 84)
Durand et al. (2013) analyzed sin stocks in the Pacific Basin that exhibit a variety of social norms. They suggest, considering measurable differences in cultural characteristics, that the price of sin is a manifestation of groupthink (Janis, 1982; Taffler and Tucket, 2010). Groupthink, or also called herding mentality, assumes that when a group is operating, individuals do not act upon their own ideas, but rather seek comfort in the efforts of the group. Under the wing of the group individuals pursue group harmony and adhere to conformity which might cloud judgment and result in an irrational decision-making process. Through the mechanism of groupthink, the authors argue that collectivistic countries herd towards sin stocks while individualistic countries herd away from sin stocks. This might have a meaningful impact on the risk-reward ratio of these stocks. According to the shunned stock hypothesis, this causes sin stocks to be underpriced, ceteris paribus, and to exhibit higher expected returns (Derwall, Koedijk, Ter Horst, 2011). In turn, herding towards sin stock might have a negative relation vis-à-vis stock returns whilst herding away might have a positive relation.
Interestingly, the findings from Durand et al. (2013) for Australia and New Zealand, that can be considered Western countries, come close to the findings for Western countries in the paper by Hong and Kaczperzyk (2009). In the former two countries investors herd away from sin stocks. An explanation for this phenomenon is that in individualistic countries investors might have the belief that they can positively impact the world by taking responsibility through shunning sin firms. For other countries that are less individualistic the herding is the other direction towards sin firms. Herding towards sin can be accounted to various social norms. It may be the case that in collectivistic countries groupthink is manifest in deviating from social norms when others in the group behave similarly. Durand et al. (2013) use the individualism index, specifically its lower bound, to determine the level of collectivism in a country. The individualism index ranges from 0-100, where 0 represents a completely collectivistic country and 100 a completely individualistic country. They argue that a collectivistic country is more exposed to groupthink. As a result, if groupthink drives the sin stock anomaly, one would expect higher abnormal returns in collectivistic countries and lower abnormal returns in individualistic countries. Therefore, the third testable hypothesis is:
15
Section III
Methodology and Data
3.1 Sin portfolios
3.1.1 Identifying sins stocks
Hong and Kacperczyk (2009) identified sin stocks in the US based on SIC (Standard Industrial Code) and NAICS (North American Industry Classification System). Even so, Thomson Reuters Business Classification (TRBC) provides a richer dataset than selecting stocks based on SIC. Additionally, TRBC is applicable to stocks listed outside the US as well. Hence, in order to select sin stocks, stocks listed in Emerging Market countries, Hong Kong, Japan and the US, are selected by business activity using TRBC sectors that include alcohol, tobacco, and gaming.4 Delisted stocks are included in the sample until they disappear to account for
survivorship bias. The sample consists of primary quotes only and excludes over-the-counter (OTC) securities. With regard to the sample timespan, I follow Fama and French (2017) who conducted an international test for the validity of their five-factor model using data between 1990-2015. A similar starting point is used for my sample period. However, I find that sin stock data in Emerging Markets is sparse in the early nineties. Thus, my sample spans from July 1997 to December 2019 (Henceforth 1997-2019). In all, the above-mentioned data selection criteria lead to a raw sin stock sample of 432 firms in 25 countries
3.1.2 Identifying sin stock performance
After identifying sin stocks, monthly total returns are retrieved from Thomson Reuters EIKON as a measure for stock performance. The total return index provides a more accurate representation of a stock’s performance than the return on stock prices because it tracks capital gains and dividend yield concurrently.
3.1.3 Data cleaning procedure
Duplicate stocks and stocks with no return data are deleted from the sample. Moreover, a firm is only included in the sample if it has at least 12 months of continuous total returns which is in line with Salaber (2009, 2013). Stocks that produce ‘0-returns’ require a special treatment. When a stock produces more than three ‘0-returns’ consecutively or twice in a return series the stock is deleted. When the ‘0-returns’ occur at the end of the return series the ‘0-returns’ are deleted until a valid returns series remains. Finally, stocks with no market capitalization data
16
during the sample period are also deleted. Following this procedure 96 stocks are deleted from the sample.5
3.1.4 Sin stock sample
Table 1 reports the number of sin stocks per country in each sin industry after cleaning the raw sin stock sample. China and India are the main contributors of sin stocks in the emerging markets. Notably, China has no listed gaming stocks as the Chinese government prohibits gambling by law. In practice however, gambling is widespread among Chinese citizens via (illegal) online gaming and special administrative regions such as Macau and Hong Kong where gambling is (partly) legalized (Loo, Raylu, and Oei, 2008). This is reflected in the relatively high number of gaming and alcohol stocks in Hong Kong. Furthermore, the positive attitude towards betting in Hong Kong is reflected in the high turnover of betting on horse races, which is the largest globally. However, the exploitation of horse race betting is restricted to a single government-granted monopoly that is not listed on any stock exchange. Next, no sin stocks constitute exchanges in both Qatar and Thailand. The Qatari government strictly prohibits all forms of gambling and the production of alcohol and tobacco, while alcohol consumption is limited to licensed hotels and restaurants. The Thai government prohibits gambling, besides the run lotteries and betting on horse races. Moreover, tobacco production is run by a state-monopoly and alcohol-related firms are not allowed to list on Thai stock exchanges. Consequently, Thailand’s largest (alcoholic) beverage company, ThaiBev, is listed on the Singapore Stock Exchange. In the US more than half of sin stocks constitute gambling stocks. This stems primarily from the deregulation in many states to legalize casino gaming outside Native American casinos in the mid-to-late nineties. Nonetheless, the number of sin stocks in this sample – only 357 in all 26 countries – is very small compared to the universe of thousands of stocks.
17 Table 1
Sin Stock Selection
.
The table reports the stock selection for the 26 countries within the Emerging Markets, the United States, and Japan per sin industry. The first column denotes the ISIN abbreviation of each country and the second column shows the full country name. The final columns show the total sin stocks per country. The final row shows the total sin stocks per sin industry.
Alcohol Tobacco Gaming Total
AR Argentina 1 2 - 3 BR Brazil 3 1 - 4 CL Chile 6 1 4 11 CH China 38 1 - 39 CO Colombia 2 1 - 3 CR Czech Republic 5 1 1 7 EG Egypt 1 1 - 2 GR Greece 2 2 6 10 HU Hungary 1 - - 1 IN India 32 6 4 42 ID Indonesia 3 6 - 9 MY Malaysia 3 2 8 13 MX Mexico 3 1 - 4 PK Pakistan - 3 - 3 PE Peru 6 1 - 7 PH Philippines 3 - 12 15 PL Poland 9 2 - 11 RU Russia 5 - - 5 ZA South Africa 9 - 7 16 KR South Korea 11 2 4 17 TW Taiwan - - 4 4 TH Thailand - - - 0 TR Turkey 4 - - 4 QA Qatar - - - 0
Subtotal Emerging Markets 147 33 50 230 HK JP Hong Kong Japan 18 11 1 1 23 10 42 22 US United States 23 19 62 104 Total 199 54 145 398
3.1.5 Data to price sin portfolio returns
For the US and Emerging Market countries risk factors, risk free rates, and market returns are downloaded from the Kenneth French website. Risk factors include the performance of size (SMB), book-to-market value (HML), momentum (MOM), profitability (RMW), and investment (CMA) portfolios.6 The market return consists of a region’s value-weighted return
6 Multicollinearity of the risk factors does not seem to be an issue because the risk factors are designed
18
of the market portfolio minus the one-month US treasury bill rate (T-bill). Moreover, the monthly US T-bill rate is used as the risk-free rate to calculate excess returns.
A country specific dataset for the US and Japan is available that includes all Fama and French risk factors and the Carhart momentum risk factor. Whilst for Emerging Market countries regional risk factors are available. Griffin (2002) finds that country-specific risk factor models explain much more variation in the time series and have lower pricing error than global risk factor models. Global risk factors are thus not adequate in performance analysis of sin stocks in Hong Kong. However, Cakici, Fabozzi and Tan (2013) provide evidence that regional risk factors for emerging markets do a better job in explaining value and momentum effects than global aggregate and US factors. Hence, regional risk factors are used for Emerging Markets countries in the absence of country-specific risk factors. Global risk factors for Hong Kong are included in the developed market, formerly global, dataset as well as regional risk factors in the Asia Pacific ex Japan dataset. For each sin portfolio the regional or country risk factors are used for which the country or region has the largest market capitalization. Therefore, in this study I only use the risk factors for Emerging Markets and the United States.
3.1.6 Individualism index
Cultural dimension scores have been retrieved from the website of Geert Hofstede.7 The data
concerns the latest cultural dimensions scores from 2015. Throughout the years these cultural scores have changed as countries have (economically) developed. However, changes in cultural scores seem to be absolute instead of relative (Beugelsdijk, Maseland, and Van Hoorn, 2015). This implies that the cultural distance between countries appears to be stable through time. Henceforth, the above-mentioned justifies my choice to use these scores statically throughout the sample period. Appendix D. displays each of the countries used in the construction of sin portfolios and their respective individualism index score (IVD). The total sample consists of 25 countries.8 Table 2 depicts the division of countries based on IVD. According to Hofstede
(1984), when a country exceeds a score of 50 it is considered individualistic, whilst a score below 50 indicates a collectivistic country. India and Japan have intermediate IVDs around the 50-mark, because they share both cultural traits of individualism and collectivism. This complicates the manner to attribute the individualism score to the level of groupthink. However, in Japan and India group harmony plays an import role in society, as opposed to countries where group harmony is more restricted to extended family or the in-group, such as Argentina (Hofstede, 2001). Hence, I expect in India and Japan that investor attitudes are affected by group think. Five sin portfolios (SP1-SP5) are composed based on the region and the level of
7 I want to thank the late Geert Hofstede and Gert Jan Hofstede for the availability of cultural dimension
19
IVD. The construction is based on dividing the IVD scores in five percentile ranks in intervals of two deciles to include five levels of individualism.9
3.1.7 Portfolio construction
Hong and Kacperczyk (2008) construct equally weighted sin portfolios. However, employing equally-weighted portfolios might bias results towards smaller firms. Hence, value-weighted portfolios are constructed by weighting firms relative to market capitalization in US Dollars. In addition, equally-weighted portfolios are also constructed as a robustness check. Fama and French (1993) suggest portfolio returns are calculated as value-weighted from July of year t to June of t+1, where rebalancing occurs in June t+1.10 The value-weighted sin portfolios consist
9 Note that intermediate here is subjective to sample and cannot be extrapolated as the intermediate level of individualism, considering that Hofstede (2001) defines countries individualistic below scores of 50. 10Value-weighted portfolios that rebalance in December are also constructed but provide no meaningful differences in the outcomes of this study
Table 2
Individualism index score (IVD) sin portfolios.
The table reports the 24 countries in the sample ranked on the IVDs. The first column denotes the percentile with score between brackets. The second column reports the sin portfolio (SP) name composed of the adjacent group of countries ranked on IVD.
Percentile Portfolio Country Individualism IVD 20th (17.4) SP1 Colombia Low 13
Indonesia 14
Pakistan 14
Peru 16
Taiwan 17
40th (25.8) SP2 South Korea Low-intermediate 18
20
of all sin stocks for which market equity data is available for December of t-1 and June t, and positive book equity data for t-1.
The value weights of each security for all sin portfolios can be calculated as: 𝑊!,# = $%!,#
∑$!%&$%!,# (3)
where: 𝑊!,# is the weight of stock i at time t; 𝑀𝐸!,# is the market capitalization of stock i at time
t.
Such that the return for each value-weighted sin portfolio can be calculated as: 𝑟',#()= ∑ 𝑊
!,#𝑟!,# *
!+, (4)
where: 𝑟',#() is the value-weighted return of portfolio 𝑝 at month 𝑡; 𝑊
!,# is the weight of stock 𝑖
at time 𝑡; 𝑟!,# is the return of stock 𝑖 at month 𝑡.
The equally weighted portfolio is calculated as follows: 𝑟',#-) = ∑$!%'.!,#
* (5)
where: 𝑟',#-) is the equally-weighted return of portfolio 𝑝 at month 𝑡; 𝑟
!,# is the return of stock 𝑖
at month 𝑡.
3.1.8 Descriptive statistics
Table 3 reports the summary statistics and Sharpe Ratios of the excess returns of the five sin portfolios. The Sharpe ratio is a measure of risk-adjusted return. The Sharpe ratio describes how well the return of a portfolio compensates the investor for the risk taken. In the simple CAPM framework, consisting of the universe of risky assets, mean variance analysis dictates that the portfolio with the highest Sharpe ratio is the optimal portfolio of risky assets. Therefore, it makes economic sense to describe portfolios based on Sharpe ratios. It is calculated as:
𝑆𝑅',# = .(,#/.),#
0(,# (3)
where: 𝑆𝑅' is the Sharpe ratio of portfolio 𝑝; 𝑟',#− 𝑟1,# is the excess return of portfolio 𝑝 at
time 𝑡; and 𝜎',# is the standard deviation of 𝑟',# at time 𝑡.
21
exhibits higher excess returns than SP3vw. The Jarque-Bera probabilities indicate that there is
insufficient evidence to reject the null hypothesis of normality in all portfolios. Skewness is relatively moderate in all portfolios except for SP3vw and SP1ew. The latter portfolios exhibit
22 Table 3
Descriptive statistics.
Table 3 describes the descriptive statistics of average monthly excess returns, i.e. the average monthly return (%) of the portfolio minus the one-month Treasury bill (𝑟*+𝑟,) of the sin portfolios. Panel A concerns value-weighted sin portfolios. Panel B concerns
equally-weighted sin portfolios. The penultimate column reports the Sharpe ratios of the sin portfolios. The final column reports the p-values for the Jarque-Bera test for normality. The data for the sin portfolios is beforehand winsorized at the 1st and 99th percentile to control for
extreme values. The sample period concerns July 1997 – 2019 (end of year). Panel A
Average monthly value-weighted excess returns Portfolio Variable Obs Mean Std.
Dev.
Min Max p1 p99 Skew. Kurt. Sharpe Ratio JB prob. SP1vw 𝑟*+𝑟, 270 1.388 9.453 -25.389 33.098 -25.389 33.098 .483 5.104 0.147 0.000 SP2vw 𝑟*+𝑟, 270 .957 5.099 -14.847 17.29 -14.847 17.29 -.041 4.35 0.188 0.010 SP3vw 𝑟*+𝑟, 270 .522 4.847 -14.895 23.708 -14.895 23.708 1.189 9.527 0.108 0.000 SP4vw 𝑟*+𝑟, 270 .497 4.841 -13.76 11.747 -13.76 11.747 -.328 3.392 0.103 0.039 SP5vw 𝑟*+𝑟, 270 .859 4.415 -12.636 11.864 -12.636 11.864 -.561 3.898 0.195 0.000 Panel B
Average monthly equally-weighted excess returns Portfolio Variable Obs Mean Std. Dev.
23
3.2 Empirical model
3.2.1 Testing for pricing anomalies
After constructing sin portfolios, I test for evidence of pricing anomalies by means of the conventional asset pricing models. Accordingly, this leads to the estimation of the following models:
CAPM: 𝑟!,#− 𝑟$,# = 𝛼!+ 𝛽%,!'𝑟&,#− 𝑟$,#( + 𝜀!,# (6)
FF3: 𝑟!,# − 𝑟$,#= 𝛼!+ 𝛽%,!'𝑟&,#− 𝑟$,#( + 𝛽',!𝑆𝑀𝐵#+ 𝛽(,!𝐻𝑀𝐿#+ 𝜀!,# (7)
FFC: 𝑟!,#− 𝑟$,# = 𝛼!+ 𝛽%,!'𝑟&,#− 𝑟$,#( + 𝛽',!𝑆𝑀𝐵#+ 𝛽(,!𝐻𝑀𝐿#+ 𝛽),!𝑀𝑂𝑀#+𝜀!,# (8)
FF5: 𝑟!,# − 𝑟$,#= 𝛼!+ 𝛽%,!'𝑟&,#− 𝑟$,#( + 𝛽',!𝑆𝑀𝐵#+ 𝛽(,!𝐻𝑀𝐿#+ 𝛽)*𝑅𝑀𝑊#+ 𝛽+*𝐶𝑀𝐴#+ 𝜀!,# (9) where: 𝑟',# − 𝑟2,# indicates the return of portfolio 𝑝 in excess of the risk-free rate in month 𝑡;
𝛼' is the abnormal return of portfolio 𝑝; 𝑟$,#− 𝑟2,# is the return of the market portfolio in excess of the risk free rate in month 𝑡; 𝑆𝑀𝐵#, 𝐻𝑀𝐿#, 𝑀𝑂𝑀#, 𝑅𝑀𝑊#, and 𝐶𝑀𝐴# indicate respectively
the risk factors for size, value, momentum, profitability, and investment in month 𝑡; Lastly, 𝜀',#
is the error term of portfolio 𝑝 in month 𝑡.
To determine whether abnormal returns exists one starts at estimating the CAPM. If the alpha is insignificant or zero, there are no abnormal returns to begin with. According to Fabozzi and Blitz (2015) one should expect to observe diminishing alphas when controlling for additional risk factors. Moreover, positive exposures to the latest investment and profitability risk factors are expected to completely price away alpha.
Section IV
Results
4.1 The existence of pricing anomalies
4.1.1 Value-weighted sin portfolios
To investigate whether sin portfolios exhibit positive risk-adjusted returns I run the CAPM, FF3, FFC, FF5 models for the value-weighted sin portfolios. Table 4 presents the results of the time series regressions for all five sin portfolios (SP1-SP5).
24
begin with. Thus, my first hypothesis, that sin stocks exhibit risk-adjusted returns, is not proved based on the results of SP3 and SP4. Nonetheless, the remaining sin portfolios do exhibit positive risk-adjusted returns that are statistically significant. Hence, attempting to answer my first hypothesis provides conflicting results. Notably, the market beta coefficient is positive but below 0.5 for the significant portfolios SP1 and SP2 with the highest alphas. This indicates a low market beta portfolio. From an investor perspective, due to its low beta, this sin portfolio might be more suitable for investors with relatively more risk-aversion and a long-term orientation who tend to rely less on market perils. Additionally, research has shown that low-beta assets have higher alphas and Sharpe rations than portfolios of high-low-beta assets. 11This is
reflected in this study, where the portfolios with the highest betas have the lowest alphas. However, SP5, the portfolio with the lowest beta has the highest Sharpe ratio of the value-weighted sin portfolios.
In line with my results, Blitz and Fabozzi (2015) find that the inclusion of the risk factors of investment and profitability completely renders the alphas insignificant for a US based sin portfolio. SP5 is a special case because it consists of stocks mostly listed in the US in number and market capitalization. I find conflicting results where alphas diminish for some portfolios while increasing for others when controlling for additional risk factors. However, when controlling for investment and profitability in the FFC model, the statistically significant alphas are rendered insignificant compared to the FF3 model. Therefore, I provide evidence for my second hypothesis in this setting.12
Notably, the sin portfolio with the lowest IVD scores (SP1) outperforms the sin portfolio with the highest IVD scores (SP5) in terms of risk-adjusted returns. If a pattern in the alphas of the sin portfolios exists, it appears to be the reverse of my expectation. Therefore, I am unable to proof my third hypothesis that sin stocks in countries with high individualism have higher risk-adjusted returns than sin stocks in collectivistic countries.
4.1.2 Risk factor loadings
The analysis of the second hypothesis immediately disqualifies SP3 and SP4 because the one-factor alphas are insignificant to begin with. However, in discussing the portfolio excess returns and its loadings to the conventional risk factors I include all sin portfolios to examine which factors drive sin stock excess returns in the various models. The adjusted r-squared tends to be fairly low for all model specifications, especially in SP1. This suggests that much of the variation in the excess returns is left unexplained in these sin portfolios. In this regard, this
11see Frazzini and Pedersen (2014) for an excellent discussion
12Notably, but not included in this research, a portfolio consisting of a US based value-weighted sin
25
should not be a large surprise; the literature is conclusive in dismissing the performance of the Fama and French five-factor model in explaining Emerging Markets average excess returns 13
The market beta coefficients are highly significant and smaller than unity. This hints to the anecdotal belief that sin stocks are more recession-proof because of their addictive nature and their use of more private debt compared to non-sin stocks (Hong & Kaczperzyk, 2009). However, Salaber (2009) provides evidence that sin stocks do only as good as industry comparables, such as pharmaceuticals and utilities during recessions.
The risk factor for size is statistically significant only for SP2 and SP4. For these portfolios the Emerging Markets set of risk factors was used because these portfolios have the largest share of market capitalization in Emerging Market countries. This hints to the suggestion that there is a slight size effect, a size premium in small stocks over large stocks, in the excess returns in Emerging Markets sin stocks in line with previous work by Leite, Klotzle, and Pinto (2018). The value risk factor is only significant in SP5. The significant value risk factor implies the existence of a value effect in SP5. The value effect here, concerns higher expected returns of value stocks over growth stocks. The rationale is that value stocks have been established and have less investment opportunities than growth stocks. Hence, higher expected returns vis-à-vis higher cost of capital. However, in the FF5 specification, the risk factors of profitability and investment render the value risk factor insignificant in line with Fama and French (2015). This implies that the addition of the investment and profitability risk factors describe better the variation in the excess returns to such an extent that the value risk factor becomes insignificant. None of the sin portfolios exhibit statistically significant momentum risk factors. Thus, momentum performs abysmally in explaining the variation of the excess returns of sin stock portfolios. As mentioned, this is not unexpected as other authors show similar results concerning momentum (e.g. Blitz and Fabozzi, 2015; Salaber, 2007). Furthermore, the profitability and investment risk factors are only significant in SP5 in line with Blitz and Fabozzi (2015).
26 Table 4
Return performance of value-weighted sin portfolios
This table reports average coefficients of the time-series regression of five sin portfolios (SP1-SP5)(the monthly excess return for an value-weighted portfolio of sin stocks – alcohol, tobacco, and gaming) on common risk factors. Sin portfolios are composed of Emerging Markets, Hong Kong, Japan, and the US, where each sin portfolio is based on the level of a country’s individualism index. The sample period concerns July 1997 – 2019 (end of year). The first column denotes the model specification. The second column denotes the SP. MKTRF indicates the excess returns of the market. ALPHA, the intercept, indicates the risk-adjusted return of the sin portfolio. SMB, HML, MOM, RMW indicate respectively the risk factors for size, value, momentum, profitability, and investment, as described by Fama and French (2015). Robust std. errors are between brackets. *** 1% significance; ** 5% significance; and * 10% significance.
Model Portfolio ALPHA MKTRF SMB HML MOM CMA RMW Adj. R2
SP1vw 1.177** 0.375*** 0.059 (0.567) (0.0919) SP2vw 0.750*** 0.370*** 0.206 (0.276) (0.0452) CAPM SP3vw 0.374 0.263*** 0.114 (0.279) (0.0406) SP4vw 0.327 0.302*** 0.151 (0.274) (0.0476) SP5vw 0.520** 0.567*** 0.324 (0.228) (0.0595)
Model Portfolio ALPHA MKTRF SMB HML MOM CMA RMW Adj. R2
SP1vw 1.265* 0.449*** 0.628* -0.142 0.068 (0.669) (0.0970) (0.342) (0.439) SP2vw 0.650** 0.349*** -0.156 0.149 0.207 (0.327) (0.0467) (0.158) (0.209) SP3vw 0.363 0.279*** 0.146 0.0126 0.110 FF3 (0.278) (0.0439) (0.147) (0.146) SP4vw 0.311 0.344*** 0.374** 0.0162 0.166 (0.305) (0.0484) (0.161) (0.152) SP5vw 0.473** 0.617*** -0.107 0.355*** 0.389 (0.217) (0.0571) (0.0908) (0.0841)
Model Portfolio ALPHA MKTRF SMB HML MOM CMA RMW Adj. R2
SP1vw 1.386* 0.431*** 0.630* -0.166 -0.145 0.068 (0.727) (0.0984) (0.341) (0.438) (0.242) SP2vw 0.623* 0.353*** -0.156 0.155 0.0318 0.204 (0.342) (0.0472) (0.158) (0.209) (0.132) SP3vw 0.460 0.265*** 0.147 -0.00645 -0.116 0.111 FFC (0.317) (0.0418) (0.148) (0.151) (0.106) SP4vw 0.322 0.342*** 0.374** 0.0140 -0.0133 0.163 (0.315) (0.0531) (0.161) (0.149) (0.120) SP5vw 0.514** 0.590*** -0.0964 0.327*** -0.0654 0.391 (0.218) (0.0633) (0.0905) (0.0836) (0.0502)
Model Portfolio ALPHA MKTRF SMB HML MOM CMA RMW Adj. R2
27 4.1.3 Equally-weighted portfolios
As a robustness check, equally-weighted equivalents of sin portfolios SP1-SP5 are constructed in line with Hong and Kacperczyk (2009). The results of these portfolios are shown in Table 5. All sin portfolios of the equally-weighted portfolios exhibit positive and highly significant risk-adjusted returns. Hence, the existence of a sin stock anomaly in each of the portfolios.
Looking at the CAPM, in line with the results for the value-weighted portfolios, there does not seem to be a pattern in the alphas ascending from SP1 to SP5. This is evident in the first sin portfolio which has the largest risk-adjusted return. Hence, in the case of the equally weighted portfolios I am unable to confirm my third hypothesis as well. Constructing portfolios based on sin stocks from countries with higher levels of individualism does not deliver higher risk-adjusted returns.
The alphas seem to diminish when controlling for additional risk factors in some sin portfolios while in others they seem to increase in both the FF3 and FFC specifications. I hypothesized that risk-adjusted returns of sin stocks can be captured and fully priced away in the FF5 specification. Contrary to my expectation, the alphas remain positive and significant. Moreover, the risk factors of investment and profitability are all insignificant except for SP5. This suggests that portfolio or country-specific idiosyncrasies outside the scope of this research influence the risk-adjusted returns of sin stock portfolios. While the results for the equally-weighted portfolios deviate from the results of the value-weighted portfolios this is not per se problematic. I want to emphasis that equally-weighted portfolios, although used in the sin stock literature, are typically deviating from portfolio construction in practice by institutional and retail investors alike.14 Moreover, the results of the value-weighted portfolios in this study are
line with the findings of Blitz and Fabozzi (2015).
28 Table 5
Return performance of equally-weighted sin portfolios.
This table reports average coefficients of the time-series regression of five sin portfolios (SP1-SP5)(the monthly excess return for an equally-weighted portfolio of sin stocks – alcohol, tobacco, and gaming) on common risk factors. Sin portfolios are composed of Emerging Markets, Hong Kong, Japan, and the US, where each sin portfolio is based on the level of a country’s individualism index. The sample period concerns July 1997 – 2019 (end of year). The first column denotes the model specification. The second column denotes the SP. MKTRF indicates the excess returns of the market. ALPHA, the intercept, indicates the risk-adjusted return of the sin portfolio. SMB, HML, MOM, RMW indicate respectively the risk factors for size, value, momentum, profitability, and investment, as described by Fama and French (2015). Robust std. errors are between brackets. *** 1% significance; ** 5% significance; and * 10% significance.
Model Portfolio ALPHA MKTRF SMB HML MOM CMA RMW Adj. R2
SP1ew 2.552*** 0.319*** 0.116 (0.333) (0.0501) SP2ew 1.276*** 0.640*** 0.338 (0.345) (0.0547) CAPM SP3ew 1.007*** 0.561*** 0.332 (0.299) (0.0495) SP4ew 1.802*** 0.508*** 0.254 (0.334) (0.0529) SP5ew 1.208*** 0.780*** 0.485 (0.222) (0.0495)
Model Portfolio ALPHA MKTRF SMB HML MOM CMA RMW Adj. R2
SP1ew 2.593*** 0.376*** 0.497** -0.0717 0.136 (0.366) (0.0508) (0.196) (0.181) SP2ew 1.506*** 0.647*** 1.148*** -0.360 0.442 (0.400) (0.0530) (0.217) (0.273) SP3ew 1.141*** 0.0271*** 0.725*** -0.211 0.382 FF3 (0.307) (0.0039) (0.193) (0.147) SP4ew 1.794*** 0.559*** 0.456** 0.0023 0.267 (0.348) (0.0512) (0.181) (0.158) SP5ew 1.140*** 0.750*** 0.325*** 0.248*** 0.543 (0.207) (0.0452) (0.0923) (0.0747)
Model Portfolio ALPHA MKTRF SMB HML MOM CMA RMW Adj. R2
SP1ew 2.602*** 0.375*** 0.497** -0.0734 -0.0106 0.133 (0.370) (0.0525) (0.196) (0.178) (0.125) SP2ew 1.455*** 0.785*** 1.148*** -0.350 0.0617 0.440 (0.396) (0.0619) (0.217) (0.268) (0.111) SP3ew 1.180*** 0.642*** 0.725*** -0.219 -0.0467 0.382 FFC (0.322) (0.0580) (0.192) (0.150) (0.134) SP4ew 1.821*** 0.555*** 0.456** -0.0030 -0.0326 0.264 (0.375) (0.0535) (0.182) (0.160) (0.140) SP5ew 1.221*** 0.696*** 0.345*** 0.192*** -0.128*** 0.557 (0.203) (0.0468) (0.0793) (0.0677) (0.0419)
Model Portfolio ALPHA MKTRF SMB HML MOM CMA RMW Adj. R2
29
Section V
Conclusion
In this thesis I attempt to answer the research question whether groupthink is a driver for the risk-adjusted returns of sin portfolios. Additionally, I examined whether the conventional asset pricing models explain well the excess returns of sin portfolios. The answer to this question might have meaningful implications for (institutional) investors with objectives that involve investments based on social norms while executing a fiduciary duty to its investors. Moreover, the outcome of this study has relevance to SRI investors that apply negative screens to sin stocks, and by doing so might potentially miss out on financial returns. I employed time series regressions using conventional asset pricing models to price sin portfolio excess returns in a sample of 338 sin stocks in 25 countries, over the period 1997-2019.
The conventional asset pricing models demonstrated weak performance in explaining the excess returns of sin stock return in line with prior research (Blitz and Fabozzi, 2015). Moreover, my study has resulted in conflicting findings considering different groups of countries sorted on the individualism index score. I provided evidence for the existence of pricing anomalies in sin stock portfolios. However, I showed inconsistencies among various sin portfolios and countries. My findings indicate that in SP3 and SP4 there are no significant one-factor risk-adjusted returns to begin with, while in SP1, SP2, and SP5 the risk-adjusted returns are completely priced away and captured by the investment and profitability risk factors. Furthermore, equally-weighted portfolios showed outcomes deviating from the findings of the value-weighted portfolios. Nonetheless, the results for the equally-weighted portfolios were in line with the findings of Hong and Kacperczyk (2009) because all equally-weighted portfolios exhibited positive and significant risk-adjusted returns.
30
31
References
Akerlof, G. A. (1980). A theory of social custom, of which unemployment may be one consequence. The quarterly journal of economics, 94(4), 749-775.
Bank Of England, 2015. Breaking the tragedy of the horizon - climate change and financial stability - speech by Mark Carney. https://www.bankofengland.co.uk/speech/2015/breaking-the-tragedy-of-the-horizon-climate-change-and-financial-stability Accessed 17-09-2020.
Beugelsdijk, S., Maseland, R., & Van Hoorn, A. (2015). Are Scores on H ofstede's Dimensions of National Culture Stable over Time? A Cohort Analysis. Global Strategy Journal, 5(3), 223-240. Blitz, D., & Fabozzi, F. J. (2017). Sin stocks revisited: Resolving the sin stock anomaly. The
Journal of Portfolio Management, 44(1), 105-111.
Bragdon, J. H., & Marlin, J. (1972). Is pollution profitable. Risk management, 19(4), 9-18.
Brown, S. J., Goetzmann, W., Ibbotson, R. G., & Ross, S. A. (1992). Survivorship bias in performance studies. The Review of Financial Studies, 5(4), 553-580.
Cakici, N., Fabozzi, F. J., & Tan, S. (2013). Size, value, and momentum in emerging market stock returns. Emerging Markets Review, 16(3), 46-65.
Capelle‐Blancard, G., & Monjon, S. (2014). The performance of socially responsible funds: does the screening process matter?. European Financial Management, 20(3), 494-520.
Carhart, M. M. (1997). On persistence in mutual fund performance. The Journal of finance, 52(1), 57-82.
CDC, 2018. Current Cigarette Smoking Among Adults in the United States. https://www.cdc.gov/tobacco/data_statistics/fact_sheets/adult_data/cig_smoking/index.htm#n ation Accessed 21-09-2020.
Chaloupka, F. J. (1999). Macro-social influences: the effects of prices and tobacco-control policies on the demand for tobacco products. Nicotine & Tobacco Research, 1(Suppl_1), S105-S109. Cheung, W. M. Y., & Lam, D. (2015). Comparing the price of sin: Abnormal returns of cross-listed
casino gaming stocks in the Hong Kong and US markets. International Journal of Hospitality Management, 45, 73-76.
Chong, J., Her, M., & Phillips, G. M. (2006). To sin or not to sin? Now that's the question. Journal of Asset Management, 6(6), 406-417.
Dam, L., & Scholtens, B. (2015). Toward a theory of responsible investing: On the economic foundations of corporate social responsibility. Resource and Energy Economics, 41, 103-121. Delmas, M. A., & Burbano, V. C. (2011). The drivers of greenwashing. California management
review, 54(1), 64-87.
Derwall, J., Koedijk, K., & Ter Horst, J. (2011). A tale of values-driven and profit-seeking social investors. Journal of Banking & Finance, 35(8), 2137-2147.
Durand, R. B., Koh, S., & Limkriangkrai, M. (2013). Saints versus Sinners. Does morality matter?. Journal of International Financial Markets, Institutions and Money, 24, 166-183.
32
Fabozzi, F. J., Ma, K. C., & Oliphant, B. J. (2008). Sin stock returns. The Journal of Portfolio Management, 35(1), 82-94.
Fama, E. F. (1970). Efficient capital markets: A review of theory and empirical work. The journal of Finance, 25(2), 383-417.
Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1), 3-56.
Fama, E. F., & French, K. R. (2015). A five-factor asset pricing model. Journal of Financial Economics, 116(1), 1-22.
Fama, E. F., & French, K. R. (2017). International tests of a five-factor asset pricing model. Journal of financial Economics, 123(3), 441-463.
Fauver, L., & McDonald IV, M. B. (2014). International variation in sin stocks and its effects on equity valuation. Journal of corporate finance, 25, 173-187.
Friede, G., Busch, T., & Bassen, A. (2015). ESG and financial performance: aggregated evidence from more than 2000 empirical studies. Journal of Sustainable Finance & Investment, 5(4), 210-233.
Goodchild, M., Nargis, N., & d'Espaignet, E. T. (2018). Global economic cost of smoking-attributable diseases. Tobacco control, 27(1), 58-64.
Griffin, J. M. (2002). Are the Fama and French factors global or country specific?. The Review of Financial Studies, 15(3), 783-803.
GSIA, 2018. 2018 Global Sustainable Investment Review.
https://www.ussif.org/files/GSIR_Review2018F.pdf Accessed 16-09-2020.
Hofstede, G. (1983). National cultures in four dimensions: A research-based theory of cultural differences among nations. International Studies of Management & Organization, 13(1-2), 46-74.
Hofstede, G. (1984). Cultural dimensions in management and planning. Asia Pacific journal of management, 1(2), 81-99.
Hofstede, G. (2001). Culture's consequences: Comparing values, behaviors, institutions and organizations across nations. Sage publications.
Hong, H., & Kacperczyk, M. (2009). The price of sin: The effects of social norms on markets. Journal of Financial Economics, 93(1), 15-36.
Humphrey, J. E., & Tan, D. T. (2014). Does it really hurt to be responsible?. Journal of business ethics, 122(3), 375-386.
IPPS, 2007. Climate Change 2007: Impacts, Adaptation, and Vulnerability. https://www.ipcc.ch/site/assets/uploads/2018/03/ar4_wg2_full_report.pdf Accessed 17-09-2020
IPPS, 2013. Climate Change 2013: The Physical Science Basis. https://www.ipcc.ch/site/assets/uploads/2018/02/WG1AR5_all_final.pdf Accessed 17-09-2020.
33
Jegadeesh, N., & Titman, S. (1993). Returns to buying winners and selling losers: Implications for stock market efficiency. The Journal of finance, 48(1), 65-91.
Jensen, M. C. (1968). The performance of mutual funds in the period 1945–1964. The Journal of finance, 23(2), 389-416.
Kim, I., & Venkatachalam, M. (2011). Are sin stocks paying the price for accounting sins?. Journal of Accounting, Auditing & Finance, 26(2), 415-442.
Krabbe, O., Linthorst, G., Blok, K., Crijns-Graus, W., Van Vuuren, D. P., Höhne, N. & Pineda, A. C. (2015). Aligning corporate greenhouse-gas emissions targets with climate goals. Nature Climate Change, 5(12), 1057-1060.
Kubota, K., & Takehara, H. (2018). Does the Fama and French five‐factor model work well in Japan?. International Review of Finance, 18(1), 137-146.
Ladouceur, R., Boisvert, J. M., Pépin, M., Loranger, M., & Sylvain, C. (1994). Social cost of pathological gambling. Journal of gambling studies, 10(4), 399-409.
Leite, A. L., Klotzle, M. C., Pinto, A. C. F., & da Silva, A. F. (2018). Size, value, profitability, and investment: Evidence from emerging markets. Emerging Markets Review, 36, 45-59.
Lintner, I. (1965). The valuation of risk assets and the selection of risky investments in stock portfolios and capital budgets. Review of Economics and Statistics, 47(1), 13-37.
Liston, D. P. (2016). Sin stock returns and investor sentiment. The quarterly review of economics and finance, 59, 63-70.
Liston, D. P., and Soydemir, G. (2010). Faith-based and sin portfolios: An empirical inquiry into norm-neglect vs norm-conforming investor behavior. Managerial Finance, 36(10), 876–885. Lobe, S. & Walkshäusl, C. (2016). Vice versus Virtue Investing Around the World. Review of
Managerial Science, 10(2), 303-344.
Loo, J. M., Raylu, N., & Oei, T. P. S. (2008). Gambling among the Chinese: A comprehensive review. Clinical Psychology Review, 28(7), 1152-1166.
Margolis, J. D., Elfenbein, H. A., & Walsh, J. P. (2007). Does it pay to be good? A meta-analysis and redirection of research on the relationship between corporate social and financial performance. Ann Arbor, 1001, 48109-1234.
Markowitz, H. M. (1952). Portfolio selection. Journal of Finance, 7(1), 77-91.
Maynard, A., & Kennan, P. (1981). The economics of alcohol abuse. British journal of addiction, 76(4), 339-345.
Merriam Webster. (n.d). Sin. In Merriam-Webster.com dictionary. https://www.merriamwebster.com/dictionary/sin Accessed 10-09-2020.
Mosoeu, S., & Kodongo, O. (in press). The Fama-French five-factor model and emerging market equity returns. The Quarterly Review of Economics and Finance. Retrieved from https://www.sciencedirect.com/science/article/pii/S1062976920301460?casa_token=nRPo8H Q2RJoAAAAA:wlR_0MCETJCl7xyCPK8MBrdxLFLvED4bveuVHXK1m2TfGHtdQLGyvb dSTfB3Dz3CPmgc0DUEY_A
