1
The effect of corporate social performance on the cost of equity
in an international setting post-crisis
Anne Hoekstra
S1889710
ath14@hotmail.com
MSc Finance
MSc International Financial Management University of Groningen
Faculty of Economics and Business Thesis supervisor: Dr. S. Ursu
June 2016
Abstract
In recent years businesses are increasingly driven to socially-responsible behavior by society, however the jury is still out on whether CSR is a value creating activity. Different countries are analyzed as well as a distinction is made between international and domestic firms. The international dimension is addressed both on a country level as well as on a firm level. The research sample is based on the Thomson Reuters ASSET4 database and ranges from 2008-2013. Results from the pooled OLS regression model show a significant link between CSR and the cost of equity, however the effect varies depending on the cost of equity construct.
JEL Codes: F32; G32; F34; M14
2
1. Introduction
Businesses have been striving for more socially-responsible decision making in the recent years as corporations make an effort to be socially-responsible members of society (Callan & Thomas, 2009; Jha & Cox, 2015). This is more commonly known as corporate social responsibility (CSR), which is defined as “the economic, legal, moral and philanthropic actions of the firm that influence the quality of life of the relevant stakeholders (Hill, Ainscough, Shank, & Manullang, 2007).” Over time CSR has become one of the most dominant topics in business (Goss & Roberts, 2011).
At the conceptual level CSR is nothing new under the horizon, dealing with the environmental, social and economic impacts, as well as stakeholders, has always been an integral part of doing business (Dahlsrud, 2008). However, an elaborate literature study by Aguinis and Glavas (2012) acknowledges that interest in CSR is accelerating rapidly. In particular, the financial crisis of 2007-8 lead to a reformulation of the relationship between society and business (Kemper & Martin, 2010). The crisis has changed the perception of the public towards the role of corporations in society and their corporate social responsibility (Jacob, 2012). The international capital markets are increasingly starting to show appreciation for ethical corporate conduct and are now increasingly rewarding towards corporate social responsibility (Lopatta & Kaspereit, 2014). Hence, CSR has become an increasingly important part of, for example operations of U.S. firms, due to either incorporation into the firms its vision or due to pressure from activist groups (Deng, Kang, & Low, 2013). For example, recently activist groups urged investors into selling their shares of companies that actively contribute to climate changes (The Economist, 20151). Whether the devotion to CSR is voluntarily or not, it becomes apparent that this longing for ethical corporate conduct has the potential of influencing firm performance and valuation. In this paper the main focus will lie on whether stockholders require a lower return from firms with a higher CSR rating. Hence, the main research question is: “What is the effect of CSR on the cost of equity?”. This study focuses on the post-crisis period and is conducted upon a research sample running from 2008 to 2013, which is extracted from the Thomson Reuters Asset4 database.
The sharp increase in CSR has sparkled interest among researchers to investigate the link between CSR and the cost of capital. There is limited empirical evidence on the relationship,however the literature states that U.S. firms with a higher CSR score enjoy a lower cost of equity (El Ghoul, Guedhami, Kwok, & Mishra, 2011) and a lower cost of debt (Goss & Roberts, 2011). Hence, the weighted average cost of capital (WACC) is reduced for companies that are considered to uphold a high ethical standard. Moreover, Jha and Cox (2015) show that the environment in which a firm conducts business is also be influential regarding the impact of CSR on financial performance. These studies raise the interesting question of whether the negative relationship between CSR and the cost of equity holds
1
3
for other countries as well. The literature present no evidence on whether financial managers in Europe or Asia also face this negative relationship. Is the same amount of investor appreciation for CSR activities present in these countries? Furthermore, which elements within society drive this investor appreciation and are these elements extended internationally?
This thesis further explores the relationship between CSR and the cost of equity internationally by making use of stakeholder theory, institutional theory and the resource-based view (RBV). This study addresses the international dimension in two ways. Firstly, continental and country samples provide further insights into the relationship between CSR and the cost of equity globally. The effect is expected to differ internationally as firms face stakeholders with varying priorities under a different institutional setting. Secondly, the international orientation of a firm is taken into account as this thesis discusses the relationship between cost of equity and CSR for both domestic and international firms. Companies with significant operations abroad operate under a brighter spotlight and have a complex system of stakeholders to take into account. Therefore, CSR activities are expected to differ between international and domestic firms as well as the consequences for the cost of equity. Hence, it is interesting to research whether managers need to take the international orientation of their firm into account when evaluating CSR activities.
4
2. Theory and hypotheses development
Firms increasingly regard CSR as a high profile public issue, as corporations expand public good provision and seriously address reductions of negative externalities (Kitzmueller & Shimshack, 2012). Nowadays, managers are put under enormous pressure by society to engage in actions that show their intent to engage in socially responsible behavior (McWilliams & Siegel, 2010). These endeavors may even drive firms to go beyond what is conceived in the best interest of the firm from a shareholder perspective, as firms sacrifice profits for social welfare (Bénabou & Tirole, 2010). For example, Di Giuli and Kostovetsky (2014) find in their research that there is a negative correlation between high CSR ratings and future stock prices, while also operating performance suffers from elaborate CSR activities. A generally accepted belief in corporate finance and valuation is that any management action that does not create cash flow, does not create value (Goedhart, Koller, & Wessels, 2015). Hence, which manager in his right mind would engage in such value-destroying activities? The literature turns to different theoretical frameworks in order to explain CSR undertakings as a generally accepted theoretical framework regarding CSR is yet to be established (Lindgreen & Swaen, 2010; Russo & Perrini, 2010). Three dominant theories are applied here to serve as the theoretical foundation for the hypotheses established in this thesis. The resource-based view, stakeholder theory and institutional theory are elaborated upon in order to investigate the motives of a firm to engage in CSR activities. In combination these theories provide a powerful theoretical basis for the research questions at hand.
5
inimitable intangible resource. This theoretical perspective thus allows for the RBV to acknowledge CSR initiatives as a source of sustainable competitive advantage.
An additional benefit of superior stakeholder management is the mitigation of information asymmetry and opportunistic behavior. Companies with high CSR scores benefit from disclosing information regarding their CSR practices and are more likely to utilize the acquired public goodwill by engaging in a seasoned equity offering (Dhaliwal, Li, Tsang, & Yang, 2011). Moreover, a study by Cheng, Ioannou, and Serafeim (2014) shows that high CSR scores lead to easier access to finance due to decreased agency costs. These studies underline the relationship between the approval of stakeholders and CSR. These stakeholders such as consumers, employees and investors are showing an increasing tendency of rewarding good corporate citizens and punishing bad corporate citizens (Du, Bhattacharya, & Sen, 2010), a process that has accelerated since the financial crisis (Lopatta & Kaspereit, 2014). With this in mind, companies engage in CSR activities in order to acquire a favorable perception among key stakeholders (Lindgreen & Swaen, 2010), which aids a corporation to achieve a sustained strategic advantage over their rivals (Du et al., 2010). Hence, in line with this view, stakeholder engagement through CSR is a positive net present value action which should hence be perceived in a positive light by shareholders (Krüger, 2015). An example of this is provided by Flammer (2013), who finds that shareholders react negatively to eco-harmful behavior by a firm, while rewarding eco-friendly initiatives. Similarly, Sharfman and Fernando (2008) find evidence that environmental risk management has a positive effect on the cost of equity of a firm. Edmans (2011) provides another example by showing that CSR initiatives aimed at employee satisfaction also significantly boost corporate performance. CSR thus holds the potential for creating significant financial benefits, which may not necessarily directly relate to the CSR expenditure. Hence, the seemingly altruistic act of engaging in CSR activities has the potential of being a significant driver of financial performance.
6
(1997) find that a competitive disadvantage arises for low CSR firms as they create tensions with stakeholders and have increased chances of explicit costs. The study by Cheng et al. (2014), that shows decreased access to finance for low CSR rating firms, also underlines this increase in perceived agency costs. Fatemi, Fooladi, and Tehranian (2015) discuss this at length and acknowledge that a more mindful production process and engagement with stakeholders lowers the risk profile of a company as a wide array of stakeholders is positively addressed through CSR, resulting in positive valuation effects through multiple valuation fundamentals. Hence, frictions between the firm and the stakeholders are avoided due to a high CSR commitment, leading to the expectation that a negative relationship between CSR and the cost of equity exists. This effect is expected to be strengthened in this study relative to the study by El Ghoul et al. (2011) as the post-crisis period has seen an increasing outcry by the public for CSR activities. Therefore, the first hypothesis states that:
Hypothesis 1: The CSR rating of a firm is negatively related to the firm’ cost of equity.
7
and institutions that contribute to social justice is observed (Dobers & Halme, 2009), while emerging markets are often also characterized by lacking business standards due to a turbulent socio-economic, instable political conditions and a weak legal system (Lenssen et al., 2011). Furthermore, Campbell (2007) extensively studies reasons for differences in social responsible ways and notes that when normative calls for responsible behavior are institutionalized, firms are more likely to engage in socially responsible behavior. Hence, the weaknesses of structures and institutions that aid social justice in emerging markets moderate the negative relationship between cost of equity and CSR. Therefore, the second hypothesis states that:
Hypothesis 2: The relationship between the CSR rating and the cost of equity is weakened for less developed emerging markets.
8
makes investors more likely to take CSR into consideration and reward high CSR firms with a lower cost of equity. Hence, the third hypothesis states that:
9
3. Methodology
In this section an overview is provided of the methodology that is applied in this study. Firstly, an elaborate discussion is provided on how to the cost of equity is estimated in this study, based upon best practices in the existing literature. Secondly, the theoretical underpinnings that are required to estimate the cost of equity are discussed in detail. In this section, mainly, the dividend discount model is elaborated upon as it provides an essential intuitive basis for the cost of equity estimations applied in this thesis. Thirdly, the abnormal earnings growth (AEG) model is briefly discussed as it is provides the basis for the cost of equity estimations used in this thesis. Fourthly, the capital asset pricing model (CAPM) is discussed as an alternative to the dividend discount model (DDM) based measures. Lastly, based on the previous discussion, an outline is provided of the quantitative techniques that are applied in this thesis.
3.1 Estimating the cost of equity
10
to using analysts’ earnings forecasts, Damodaran (2012) suggests that valuable information for forecasting earnings is extracted from looking at the current and past values. Historical data on key performance drivers is hence a good starting point for estimating a required rate of return, since investors take historical values into account when making future investment decisions. Therefore, in this thesis earnings forecasts are created based on historical values as an alternative to analysts data since analyst data is not widely available and presents several limitations.
11 3.2 Dividend discount model
An often applied method for valuing the stock price of a company is the popular dividend discount model, also known as the Gordon growth model (Gordon & Shapiro, 1956). With this method the stock value of a firm is equal to the discounted value of all future dividend payments. If the paid dividends are assumed to be continuously discounted, then the stock price at t=0 can be represented by equation (1).
𝑃0 = ∫ 𝐷𝑡𝑒−𝑘𝑡 ∞
0
𝑑𝑡 (1)
In this equation, 𝑃0 is the price of a share at time t, 𝐷𝑡 is the expected dividend at time t and 𝑘 stands for
the discount rate. In order to estimate the discount rate, or the cost of equity, that equates the future stream of dividends to the current stock price, it is necessary to estimate the Dt for t = 1,2,…, ∞. Since such information is not known, the dividend discount model estimates the cost of equity by making use of the current stock price and the expected future dividend payments. These dividends are expected to growth over time at rate g, which is equal to the return that a firm earns times the retention rate b. The retention rate in this case is defined as the percentage of net income that is not paid out as dividends but retained within the company in order to grow the business.
𝐷𝑡= 𝐷0𝑒𝑔𝑡 , with 𝑔 = 𝑏𝑟 (2)
In equation (2) 𝑏 is the retention rate and r is the return on equity, which is a measure of overall corporate profitability. The retention rate measures what percentage of income is reinvested into the firm in order to generate future growth. On the other hand, the return on equity provides an insight into the efficiency of management, as it shows the conversion rate of equity into net income. If a firm has a return on equity of 20%, then each dollar of equity is translated by the firm into $0.20 cents of net income. Now, equation (2) is substituted into equation (1) in order to derive the common equation of the dividend discount model. 𝑃0= 𝐷0∫ 𝑒−𝑡(𝑘−𝑔) ∞ 0 𝑑𝑡 (3) 𝑃0 = 𝐷𝑜 𝑘 − 𝑔 (4)
12
cost of equity is equal to the dividend yield of a stock plus the expected dividend growth rate in perpetuity.
𝑘 = 𝐷0 𝑃0
+ 𝑔 (5)
Hence, by acquiring the stock of a firm, an investor entitles himself to the future cash flows that a firm generates. This action is most profitable for the investor when he can entitle himself to such earnings at the lowest cost possible. Hence, as Ohlson and Juettner-Nauroth (2005) describe it: ”making money in the stock market reduces to the idea that investors want to buy future earnings “as cheaply as possible” for a given risk-level.” The fundamental question to an investor hence becomes: “Which firm, ceteris paribus, its earnings can relatively be most cheaply acquired?” In order to find those shares that hold great potential for investors, the theoretical underpinnings of the DDM are regarded as a valuable starting point. We now turn to the AEG model, which builds upon the insight that the present value of the expected dividends per share is the main determinant of value per share.
3.3 Abnormal earnings growth model
For practitioners that aim to calculate the cost of equity for an individual firm, using the DDM comes with practical difficulties. Firstly, the model is best suited for mature firms that enjoy a constant and sustainable growth rate combined with a fixed pay-out ratio. Alternative specifications that account for the different growth phases that a firm enjoys require many subjective assumptions in regard to the aforementioned variables. Secondly, the DDM has the obvious limitation that firms that pay no dividends hold no value according to the model.
The RIV and AEG model are accounting based and build upon the intuition that the valuation of a firm consists of an initial value plus the present value of any residual income (or abnormal earnings) which the firm obtains or is expected to obtain in the future. As previously discussed, in the cost of equity literature, the Gebhardt et al. (2001) and Claus and Thomas (2001) the model are often used to calculate the implied cost of equity based on the RIV model2. In these models the anchor value is the
2Implied cost of equity based on the Claus and Thomas (2001) estimation:
𝑃𝑡= 𝐵𝑡+ ∑ 𝐹𝐸𝑃𝑆𝑡+𝑖− 𝑅𝐶𝑇∗ 𝐵𝑡+𝑖−1 (1 + 𝑅𝐶𝑇)𝑖 5 𝑖=1 + (𝐹𝐸𝑃𝑆𝑡+5− 𝑅𝐶𝑇∗ 𝐵𝑡+4)(1 + 𝑔𝑙𝑡) (𝑅𝐶𝑇− 𝑔𝑙𝑡)(1 + 𝑅𝐶𝑇)5
13
book value per share at t=0. However, in this thesis we turn to the AEG models by Ohlson and Juettner-Nauroth (2005) and Easton (2004) in order calculate the cost of equity.
Ohlson and Juettner-Nauroth (2005) derive a square root formula that allows for calculating the required rate of return without imposing contrived restrictions on the dividend policy. The value of a share is determined by the EPS in the next period and the expected EPS growth by allowing for both a short-term measure as well as a long-term measure. By developing this setting, the OJ-model expands upon the constant growth model by allowing for an additional degree of freedom for growth. Theoretically the model is founded upon the Modigliani-Miller dividend irrelevance assumption, as it anchors the valuation upon forward EPS and subsequent growth. Both these measures hold no direct relation with the set dividend policy. However, the smoothly diminishing growth in the expected eps-sequence is adjusted for dividend in case these are paid out. Overall, the OJ model is well suited for research focusing on individual firms (Ohlson & Juettner-Nauroth, 2005). More information on the cost of equity estimations based on the AEG models by Ohlson and Juettner-Nauroth (2005) and Easton (2004) is presented in Appendix A.
14 3.4 Capital asset pricing model
In order to calculate the required rate of return of an investor, the model that is most commonly referred to in finance textbooks is the capital asset pricing model (CAPM). It was independently developed by Sharpe (1964), Lintner (1965) and Mossin (1966), who build on the portfolio theory model by Markowitz (1952). Despite the many simplifying assumptions underlying the model and its shaky empirical performance, CAPM is still held in high regard (Fama & French, 2004). The reason behind this is that the CAPM provides an intuitive and simple way of interpreting the relationship between risk and the expected return. The model by Markowitz (1952) assumes that investors are risk averse and are only concerned with the mean and the variance of their expected investment return. Investors are hence expected to pick ‘mean-variance efficient’ portfolio, which implies maximizing the expected return provided a given level of variance or minimizing the level of variance for a given level of expected return. This set of minimum variance portfolios is assumed to be similar for all investors as the model postulates homogenous expectations. The point where the capital market line, with intercept 𝑅𝑓, and the
efficient frontier are tangent is called the tangency portfolio. In accordance with the mutual fund theorem, regardless of risk preferences, all investors will hold a combination of the tangency portfolio and the riskless asset. If all investors hold this very same portfolio of risky assets, then it must be the case that the aggregate of all these identical portfolios equals the total market portfolio. Hence, it becomes important to note that if investors hold such a highly diversified portfolio, then their main concern is systematic risk as nonsystematic risk diminishes to zero. For individual securities risk beta is a good proxy of systematic risk, which measures security risk as the expected change in 𝑅𝑖 provided
a change in 𝑅𝑚. A more elaborate derivation is provided in Appendix A, which leads to the familiar
CAPM relationship as presented in equation (6).
𝐸[𝑅𝑖] = 𝑅𝑓+ 𝛽𝑖( 𝐸[𝑅𝑚] − 𝑅𝑓 ) (6)
15 3.5 Research design
The research design in this thesis is similar to the previously discussed cost of equity studies found in the literature. However, the data sample differs significantly in this study as it deviates from the, in general, U.S. oriented literature base. This study significantly contributes to existing literature due to its global dataset which is now more readily available compared to several years ago. For example El Ghoul et al. (2011) limit their research to U.S. firms with a time interval of 1992-2007 mainly due to data unavailability. However, the premise here is that, beyond this time interval the relation between society and business has changed across the globe. Besides the theoretical and intuitive basis for this premise, the data itself signals increased interest in this relationship. Relative to years before the crisis, the data quality and availability has significantly improved. This allows for the option to research the role of CSR on the cost of equity in an international setting post-crisis. Countries that have sufficient cross-sections and observations available to be studied in isolation are examined individually. These countries are the U.S., Canada, France, Germany, Switzerland, the U.K., Japan, Australia, South Korea, Hong Kong, India and Taiwan. In line with hypothesis 2, the latter three countries are considered less developed emerging markets. Alternatively, in order to produce more reliable results, based on a larger sample, companies are grouped based on geographical location. This however requires a restrictive ceteris paribus assumption across a continent to hold. The descriptive statistics and further elaboration upon the sample selection is provided in the next section. After providing the reader with further details on the data, both on continental and country level, the data is analyzed by means of a pooled OLS regression. By means of a multivariate regression analysis the effect of CSR on the cost of equity is evaluated. Equation (7) shows the regression equation that is applied for the multivariate analysis.
𝑅𝑖,𝑡 = 𝛼 + 𝛽1𝐶𝑆𝑅𝑖,𝑡+ ∑ γCONTROLS𝑖,𝑡+ 𝜀𝑖,𝑡 (7)
16
4. Data
4.1 Data sample
17
Table 1
Sample breakdown by country and year.
Panel A. North America
18
Panel D. Latin America
Country N % Year N % Chile 87 24.23 2008 29 8.08 Brazil 180 50.14 2009 42 11.70 Mexico 92 25.63 2010 68 18.94 Total 359 100 2011 71 19.78 2012 68 18.94 2013 81 22.56 Total 359 100
This table provides a breakdown of the research sample. The time interval of the research sample runs from 2008-2013 containing a total of 12,022 firm-year observations. On the left hand side, Panels A-D provide a breakdown per continent as the sample observations are widespread internationally. On the right hand side of the table an overview is provided that gives further insight into how the observations are spread out over time.
The information provided in Table 1 on the unbalanced research sample is broken down into four different geographical regions. It is noteworthy that the continental samples are often dominated by the large countries within the continent. For Europe the U.K. provides a significant share of the observations, for North-America the U.S. is the dominant factor in the subsample while for Asia Japanese observations constitute the majority of the sample. Data provision has increased significantly post-crisis as is reported in Table 1. Especially in Asia and Latin-America an increasing amount of firms have been placed within the scope of the CSR database. This is in line with the expectation that CSR activities have become a more important factor post-crisis when evaluating a firm.
4.2 Dependent variable
19 𝑟𝑜𝑗= 𝐴 + √𝐴2+ 𝐸𝑃𝑆𝑡+1 𝑃0 ∗ ( 𝐸𝑃𝑆𝑡+2− 𝐸𝑃𝑆𝑡+1 𝐸𝑃𝑆𝑡+1 − 𝑔𝑙𝑡) (8) where, 𝐴 =1 2 (𝑔𝑙𝑡+ 𝐷𝑃𝑆1 𝑃0 )
For these calculations earnings per share (𝐸𝑃𝑆𝑡), the market price as of 31st of December (𝑃𝑡) and
dividend per share (𝐷𝑃𝑆𝑡) are extracted from the Datastream database. The abnormal earnings growth
that is feasible in the long run (𝑔𝑙𝑡) is set equal to the country-specific inflation rate (K. C. Chen et al.
(2009). These country-specific inflation rates are manually gathered from the Worldbank database. The Ohlson and Juettner-Nauroth (2005) measure for calculating the cost of equity implied from analyst forecasts is adapted here due to data unavailability. Since no data is available on analyst forecasts, forecasted earnings per share (FEPS) is substituted for by 𝐸𝑃𝑆𝑡+𝑇 calculated on the basis of
historical values. This is an alternative to the limiting perfect foresight of forecasts assumed by other studies that face similar data unavailability. For example, H. Chen, Chen, Lobo, and Wang (2011) make use of realized earnings in their implied cost of equity estimations in their paper researching the effect of audit quality on the cost of equity in China. This method of assuming perfect foresight is not only theoretically restrictive but also leads to practical difficulties. In order to see this, it should be noted that the implied cost of equity models used in this thesis require that 𝐹𝐸𝑃𝑆𝑡+2 > 𝐹𝐸𝑃𝑆𝑡+1 > 0. This restrictive
assumption leads to significant observation losses in the case of using realized earnings as a substitute for analyst forecasts. Hence, in this thesis short-term growth rates are estimated based on standard financial theory in which earnings per share growth is estimated by the retention rate times the return on equity (Damodaran (2012). 𝐹𝐸𝑃𝑆𝑡+1and 𝐹𝐸𝑃𝑆𝑡+2 are hence calculated based upon the most recent
value for the return on equity and the payout ratio for which the data is acquired through Datastream. Calculating the cost of equity based on the Easton measure requires solving the quadratic formula in equation (9) for 𝑟𝑒𝑠.
𝑃𝑡 =
(𝐸𝑃𝑆𝑡+2− 𝐸𝑃𝑆𝑡+1+ 𝑟𝑒𝑠∗ 𝐸𝑃𝑆𝑡+1∗ 𝑃𝑂𝑈𝑇)
𝑟𝑒𝑠2
(9)
The inputs are similar as for the Ohlson and Juettner-Nauroth (2005) measure. Earnings per share (𝐸𝑃𝑆𝑡)
and the market price as of 31st of December (𝑃
𝑡) are extracted from Worldscope through Datastream.
The payout ratio (POUT) is calculated by dividing the dividend per share by the earnings per share. The equally weighted average of both implied cost of equity measures is then calculated in order to acquire the 𝑟𝑎𝑣𝑟. Furthermore, in line with Hail and Leuz (2006), outliers outside the 2.5th and 97.5th percentile
20
In order to avoid bias concerns due to relying fully on this implied cost of equity measure, we also calculate the cost of equity based on the CAPM. This additional estimation of the cost of equity allows us to critically evaluate the effect of CSR on the cost of equity further. The formula for calculating the cost of equity based on the CAPM model is slightly adjusted compared to the formula presented in the previous section as equation (7).
𝐸[𝑅𝑖] = 𝑅𝑓+ 𝛽𝑖( 𝐸[𝑅𝑚] + 𝐶𝑅𝑃 + − 𝑅𝑓 ) (10 )
In equation (10) the expected return on equity for stock 𝑖 (𝐸[𝑅𝑖]) is calculated with this formula which
required a risk free rate 𝑅𝑓 and an expected market return 𝐸[𝑅𝑚]. These variables are both extracted
from the database provided by Damodaran3. The 10-year Treasury Bond rates serve as a proxy for the risk free rate. The beta (𝛽) requires extensive calculations as it is calculated by dividing the covariance of the return on stock 𝑖 with the market return by the variance of the market return. The weekly stock returns are extracted from Datastream for every stock in the sample and the total return index of MSCI ACWI is used as a proxy for the total market portfolio. Furthermore, the required rate of return which is estimated based upon the CAPM is adjusted to incorporate a country-specific risk premium. This premium is required since the equity risk premium is higher in countries that have significant country specific risks (e.g. macroeconomic or political risks). These annually calculated country risk premiums are hand-collected from the public data archives of Damodaran. An overview of the country-risk premiums is found in Appendix C.
Table 2 Panel A provides descriptive statistics on the implied cost of equity for the research sample. The cost of equity is highest during the aftermath of the financial crisis in 2008, after which the cost of equity gradually decreased. In 2011 another spike is observed which was due to fears in the stock market related to the sovereign debt of multiple European countries and possible contagion within the integrated financial markets (Lane, 2012). The shrewd reader may have noticed that the cost of equity is surprisingly low for Asia and Latin America compared to the more developed continents Europe and Northern-America. Appendix D and Appendix E show a similar trend for the country specific descriptive statistics. Taiwan, for example, is reported to have a lower average cost of equity than developed markets like France, Germany and the U.S. for both CAPM and the implied cost of equity estimate. However, after a thorough examination of the cost of equity literature, this result is not so surprising. Hail and Leuz (2006) and Attig, Guedhami, and Mishra (2008) report similar findings for their research samples. It is important to note that these averages are not necessarily suitable for cross-country comparison due to sample bias. Attig et al. (2008) report that these averages are not reliable proxies for the cost of equity within the country as a whole. The significantly smaller sample for Taiwan relative to the U.S. could for example indicate sample bias as for the Taiwanese sample it may be the
21
case that only the larger, more notable, firms are covered. Hence, such omitted factors can bias results, which is why we extend the research to evaluating the relationship between CSR and cost of equity within a particular country.
4.3 Independent variable
22
Table 2
Descriptive statistics for 𝒓𝒂𝒗𝒓 and the CSR rating.
Panel A. Descriptive statistics for the implied cost of equity estimate (𝒓𝒂𝒗𝒓)
North America Europe Asia Latin America
Yea r Mean Me d ian St. Dev . Mean Me d ian St. Dev . Mean Me d ian St. Dev . Mean Me d ian St. Dev . 2008 11.70 10.79 5.56 11.75 10.88 5.92 8.52 7.84 4.27 11.79 10.72 4.93 2009 8.87 7.87 4.81 8.86 7.89 4.78 7.69 6.76 4.47 9.21 9.19 4.34 2010 9.41 8.44 4.86 9.09 8.35 4.73 8.23 7.31 4.63 9.63 8.98 4.13 2011 10.28 9.47 5.01 9.71 8.92 4.75 8.75 7.98 4.66 9.21 8.49 4.16 2012 9.57 9.10 4.69 8.95 8.32 4.64 8.03 7.27 4.20 7.62 6.91 3.39 2013 8.41 7.76 4.13 7.96 7.03 4.40 7.90 6.85 4.39 7.96 7.32 3.91 Total 9.62 8.86 4.91 9.31 8.35 4.98 8.19 7.31 4.46 8.91 8.28 4.19
Panel B. Descriptive statistics for the corporate social responsibility rating
North America Europe Asia Latin America
Yea r Mean Me d ian St. Dev . Mean Me d ian St. Dev . Mean Me d ian St. Dev . Mean Me d ian St. Dev . 2008 59.28 58.80 28.06 65.06 76.85 29.32 41.90 39.57 30.72 47.14 48.19 32.79 2009 57.90 55.75 28.78 67.39 78.50 28.79 36.78 28.89 28.78 48.40 50.51 31.87 2010 60.01 61.17 27.65 69.86 82.11 27.55 36.77 27.05 29.17 49.56 54.68 31.80 2011 59.34 60.11 28.13 69.55 82.04 27.57 39.60 32.92 29.97 46.45 46.41 31.20 2012 55.97 54.91 28.79 68.77 79.61 26.88 39.40 32.61 30.10 51.99 59.10 31.59 2013 57.92 57.30 28.48 67.98 79.12 27.57 41.07 36.00 30.51 51.58 61.79 31.47 Total 58.37 58.53 28.33 68.25 79.90 27.88 39.34 32.54 29.98 49.53 53.09 31.51
This table reports the descriptive statistics for the dependent variable and for the variable of interest in this study. The sample contains 12,022 firm-year observations ranging from 2008 to 2013 and spread over four continents. The dependent variable, 𝑟𝑎𝑣𝑟, is the implied cost of equity based on an equally weighted average of the Ohlson and Juettner-Nauroth (2005) model and
the Easton (2004) model and is reported in Panel A, while in Panel B the descriptive statistics are provided of the corporate social responsibility score. In both panels the mean, median and standard deviation are reported for the respective variables.
23
from Japan and the U.S. conduct business in a domestic market with a lower levels of CSR. This would also explain why the differences in the U.K. are not that large since Table 2 Panel B indicated that Europe is a relatively high CSR environment. Therefore, the international firms in the U.K. are likely to already meet the expectations of the additional stakeholders.
Table 3
Regression variables for international/domestic sample split.
United States
International Domestic
Mean Median Std. Dev. Mean Median Std. Dev.
rcapm 8.67 8.41 2.47 8.12 7.74 2.59 CSR 63.84 69.61 27.44 48.06 41.70 27.49 SIZE 15.89 15.69 1.42 16.03 15.87 1.40 MTB 4.02 2.54 12.21 4.24 1.87 36.56 LEV 1.30 0.48 10.14 2.36 0.76 23.47 United Kingdom International Domestic
Mean Median Std. Dev. Mean Median Std. Dev.
rcapm 7.88 7.68 2.39 7.56 7.47 2.30 CSR 71.74 82.11 24.57 68.41 76.65 24.02 SIZE 15.51 15.18 1.94 14.81 14.65 1.49 MTB 3.40 2.21 7.39 4.49 1.56 15.77 LEV 0.90 0.50 1.87 1.70 0.56 4.34 Japan International Domestic
Mean Median Std. Dev. Mean Median Std. Dev.
rcapm 8.43 8.54 2.60 8.14 8.34 2.52
CSR 48.32 51.52 29.52 34.23 25.51 27.14
SIZE 15.50 15.40 1.41 15.33 15.25 1.76
MTB 1.92 1.44 1.46 5.22 1.48 55.43
LEV 0.77 0.49 1.05 0.91 0.43 2.06
24 4.4 Control variables
The control variables are selected based upon often used variables in the cost of equity related literature. Firstly, firm size is used in this study as a control variable, the natural logarithm of total assets in U.S. dollars is used as a proxy for firm size. The general consensus is that large firms often have a proven track record and carry less risk than smaller firms. Secondly, the market-to-book ratio is recognized as an explanatory factor of cost of equity. This is the inverse of the book-to-market ratio, which is in this study included to capture differences in growth opportunities. If the market value significantly exceeds the book value then it is likely that the market expects future growth. Thirdly, leverage is included as a control variable, which is calculated as total debt divided by the market value of equity. Companies with a higher level of debt relative to their equity are in general considered to be increasingly risky. Lastly, the beta is used as a control variable. As discussed previously in section 4.2, this variable is manually calculated and serves as a proxy for the risk of a stock relative to the market. In line with prior research, the predicted signs of these control variables are BETA (+), SIZE (+) , MTB (-) and LEV (+). With the exception of the stock beta, these explanatory variables are extracted from the Worldscope database through Datastream. The descriptive statistics for these variables are presented in Table 4 Panel A, while also a correlation matrix of the regression variables is provided in Table 4 Panel B. Overall, Table 4 reports that there are low correlations between the control variables while the values of the control variables make theoretical sense. For the sample underlying hypothesis 3, Table 3 reports high standard deviations for the market-to-book value and the leverage ratio. The main reason for this are significant outliers, which are mainly internet based companies that have limited book value. The median value is hence intuitively more appropriate for the samples in which outliers are not removed. These outliers are not removed in this study since they do in fact represent the actual companies. Appendix D provides an overview of the descriptive statistics for the regression variables of the country specific samples based on the CAPM estimate, Appendix E similarly reports the descriptive statistics based on the average implied cost of equity measure. The standard deviations reported based on 𝑟𝑎𝑣𝑟
are reported to be significantly lower, this is due the fact that in this sample outliers are removed based on the 𝑟𝑎𝑣𝑟 construct. These extreme values are removed since they did not make any theoretical sense
and it is not possible to validate the construct variable 𝑟𝑎𝑣𝑟. The implied cost of equity formulas are best
25
Table 4
Regression variables for the continental sample.
Panel A. Descriptive statistics for control variables
North America Asia
Variable Mean Median Std.
Dev. Mean Median
Std. Dev. BETA 0.97 0.91 0.45 0.86 0.85 0.41 SIZE 15.93 15.75 1.50 15.74 15.63 1.65 MTB 2.88 2.16 2.81 2.41 1.55 4.90 LEV 0.87 0.52 1.88 0.85 0.50 1.23
Europe Latin America
Variable Mean Median Std.
Dev. Mean Median
Std. Dev. BETA 0.98 0.94 0.43 0.83 0.75 0.45 SIZE 16.01 15.71 1.84 16.33 16.19 1.40 MTB 2.53 1.82 3.51 2.87 2.30 2.50 LEV 1.26 0.61 2.10 1.36 0.78 1.69
Panel B. Pearson correlation coefficients between regression variables
North America Asia
𝑟𝑎𝑣𝑟
CSR_
SCORE BETA SIZE MTB 𝑟𝑎𝑣𝑟
CSR_
SCORE BETA SIZE MTB
CSR 0.08 0.01
BETA 0.04 -0.11 0.01 -0.05
SIZE -0.06 0.42 -0.08 0.07 0.30 -0.02
MTB 0.34 0.01 -0.05 -0.21 0.09 0.04 -0.03 -0.14
LEV 0.04 -0.01 0.04 0.20 0.19 0.12 0.034 -0.02 0.37 0.08
Europe Latin America
𝑟𝑎𝑣𝑟
CSR_
SCORE BETA SIZE MTB 𝑟𝑎𝑣𝑟
CSR_
SCORE BETA SIZE MTB
CSR -0.00 0.00
BETA 0.02 0.08 0.31 0.13
SIZE -0.04 0.38 0.17 0.24 0.35 0.38
MTB 0.15 0.02 -0.17 -0.18 0.05 -0.09 -0.10 -0.27
LEV 0.06 0.04 0.06 0.43 0.05 0.27 0.05 0.29 0.53 0.23
This table provides further insight into the regression variables of the sample that contains a total of 12,022 firm-year observations and ranges from 2008-2013. In Panel A. descriptive statistics are presented of the control variables for each continent separately. For each variable the provided descriptive statistics are mean, median and standard deviation. The variables reported are respectively the beta of the firm, firm size, market-to-book value and the leverage ratio. Panel B. shows the Pearson correlation coefficients of the regression variables. CSR is the corporate social responsibility score as
provided by Thomson Reuters ASSET4 and
𝑟𝑎𝑣𝑟 is the implied cost of equity based on an equally weighted average of the Ohlson and Juettner-Nauroth (2005) model
26
5. Empirical Results
5.1 Multivariate Analysis
In order to test the effect of CSR on the cost of equity, an ordinary least squares test is conducted with the ravr as dependent variable. The model includes four firm-specific control variables as well as
industry and year effects. The sample is broken down in line with Table 1, based on the continental subdivision of the Thomson Reuters database. The alternative model specification with the cost of equity based upon the standard CAPM as the dependent variable is not applied to the continental samples. The country risk premium is the main determinant of differences in the cost of equity within the continental samples and hence the CAPM has difficulties to explain aggregate results in such a sample.
27
Table 5
Regression results for the continental sample .
North America Europe Asia Latin America
C S R _ S C S R _ S C S R _ S C S R _ S C S R _ S C S R _ S C S R _ S C S R _ S (1) (2) (1) (2) (1) (2) (1) (2) CSR 0.018*** (6.916) 0.018*** (6.125) -0.003 (-0.949) 0.003 (0.827) 0.003 (1.312) -0.001 (-0.296) 0.002 (0.327) -0.011 (-1.479) BETA 0.406** (2.483) 0.820*** (3.964) 0.069 (0.409) 2.644*** (4.612) SIZE -0.053 (-0.820) -0.171*** (-2.583) 0.127*** (2.658) 0.180 (0.768) MTB 0.542*** (1.995) 0.186*** (7.365) 0.082*** (5.764) 0.172* (1.737) LEV 0.023 (0.573) 0.182*** (3.528) 0.333*** (5.527) 0.294 (1.615) INTER-CEPT 8.704*** (2.584) 7.580*** (7.120) 10.418** * (2.970) 11.062** * (10.877) 8.364*** (3.603) 6.282*** (7.991) 10.656** * (12.520) 3.401 (0.874) Year
effects Yes Yes Yes Yes Yes Yes Yes Yes Industry
effects Yes Yes Yes Yes Yes Yes Yes Yes N 4510 4121 3689 3326 5269 4208 438 359 Adj. R2 0.087 0.182 0.051 0.076 0.006 0.045 0.073 0.194
This table provides an overview of the regression results, in which the corporate social responsibility score is used as an explanatory factor next to the more conventional control variables in order to explain cost of equity differences within the sample ranging from 2008-2013. The regression is conducted separately for respectively North America, Europe, Asia and Latin America. Model specification (1) does not contain any control variables, while the second model specification contains the full set of control variables. The control variables reported are respectively the beta of the firm, firm size, market-to-book value and the leverage ratio. Statistical significance is denoted by *, ** and *** which respectively denotes 10%, 5% and 1% significance. Within parentheses the t-statistics are provided.
In order to further evaluate the effects of CSR on the cost of equity internationally and to overcome the aforementioned issues, similar ordinary least squares regressions are performed upon individual country samples. An additional required return estimation (CAPM) is used in order to avoid over-reliance on the implied cost of equity ravr. The country samples are selected based upon the criteria
28
Table 6
Regression results for the country samples.
United States Canada France Germany
rcapm ravr rcapm ravr rcapm ravr rcapm ravr
(3) (4) (3) (4) (3) (4) (3) (4) CSR -0.010*** (-6.425) 0.017*** (5.088) -0.003 (-0.990) 0.024*** (2.934) 0.012** (2.319) -0.019* (-1.774) 0.000*** (3.632) 0.002 (0.123) BETA 0.723*** (3.916) -0.126 (-0.323) 0.512*** (3.808) 1.208** (2.188) SIZE 0.014 (0.413) 0.050 (0.698) -0.406*** (-5.793) -0.568*** (-3.192) -0.035 (-0.376) 0.180* (1.683) -0.530*** (-4.484) 0.160 (0.612) MTB -0.019*** (-5.784) 0.593*** (2.028) 0.027 (0.930) 0.161* (1.831) -0.228*** (-3.856) 0.401** (2.390) 0.001 (0.020) 0.421*** (2.884) LEV 0.028*** (5.716) -0.009 (-0.210) -0.039 (-1.507) 0.578*** (3.196) 0.378*** (5.556) 0.013 (0.082) 0.271** (2.344) 0.264 (0.811) INTER-CEPT 9.604*** (17.896) 4.724*** (3.809) 13.000** * (12.492) 17.587** * (6.604) 7.280*** (5.031) 2.884 (0.991) 15.289** * (8.121) 1.491 (0.349) Year
effects Yes Yes Yes Yes Yes Yes Yes Yes Industry
effects Yes Yes Yes Yes Yes Yes Yes Yes N 4627 3448 1197 673 491 354 398 297 Adj. R2 0.088 0.208 0.187 0.129 0.196 0.088 0.112 0.110
Switzerland United Kingdom Hong Kong Australia
rcapm ravr rcapm ravr rcapm ravr rcapm ravr
(3) (4) (3) (4) (3) (4) (3) (4) CSR 0.004 (0.709) -0.002 (-0.235) -0.004 (-1.260) 0.010 (1.201) -0.007 (-1.546) -0.000 (-0.049) 0.001 (0.534) -0.016*** (-2.991) BETA 1.377*** (3.058) 0.381 (0.933) 0.280 (0.535) 0.049 (0.141) SIZE -0.153 (-1.392) 0.138 (0.869) 0.282*** (6.139) -0.201 (-1.569) -0.263*** (-3.483) 0.138 (0.897) -0.064 (-1.372) 0.359*** (3.518) MTB -0.291*** (-4.897) 0.624*** (6.237) -0.011*** (-2.022) 0.125*** (3.965) -0.054* (-1.710) 0.121* (1.735) 0.024 (0.830) 0.283*** (3.747) LEV 0.374*** (4.051) -0.050 (-0.298) 0.014 (0.674) 0.452*** (4.733) 0.250*** (2.747) -0.023 (-0.110) -0.017 (-0.454) 0.151 (1.230) INTER-CEPT 9.353*** (5.552) 2.961 (1.185) 4.225*** (6.486) 11.062** * (6.643) 13.986** * (11.223) 6.718** (2.529) 7.382*** (9.572) 3.200* (1.899) Year
effects Yes Yes Yes Yes Yes Yes Yes Yes Industry
29
India Taiwan South Korea Japan
rcapm ravr rcapm ravr rcapm ravr rcapm ravr
(3) (4) (3) (4) (3) (4) (3) (4) CSR -0.006 (-0.859) -0.012 (-1.308) -0.001 (-0.443) -0.004 (-0.496) 0.003 (0.746) 0.041*** (4.961) -0.002 (-0.965) -0.002 (-0.470) BETA -1.541** (-2.381) 0.845 (0.624) -0.068 (-0.089) -0.483 (-1.340) SIZE 0.356*** (2.606) 0.493** (2.032) -0.015 (-0.387) 0.678*** (3.599) -0.104 (-1.171) -0.409* (-1.896) 0.002 (0.056) 0.180* (1.935) MTB -0.011 (-0.548) 0.114 (0.896) -0.144*** (-5.591) 0.400*** (3.188) -0.182*** (-3.232) 0.155 (1.109) 0.000 (0.032) 0.126*** (4.500) LEV -0.030 (-0.571) 0.571** (2.578) -0.145*** (-3.024) 0.112 (0.407) -0.080 (-1.075) 0.689*** (4.956) 0.000 (0.003) 0.297** (2.164) INTER-CEPT 7.903*** (3.636) 2.149 (0.575) 11.724** * (14.298) -2.792 (-0.699) 13.117** * (8.954) 10.710** * (2.957) 10.033** * (18.404) 6.102*** (4.114) Year
effects Yes Yes Yes Yes Yes Yes Yes Yes Industry
effects Yes Yes Yes Yes Yes Yes Yes Yes N 340 250 530 407 380 273 1743 1289 Adj. R2 0.308 0.135 0.508 0.054 0.337 0.171 0.303 0.038
This table provides an overview of the regression results, in which the corporate social responsibility score is used as an explanatory factor next to the more conventional control variables in order to explain the relationship between CSR and the cost of equity. The regression is conducted separately for twelve different countries for which sufficient data is available. The time interval ranges from 2008-2013. Under model specification (3) cost of equity is estimated using the CAPM, while model specification (4) provides the results of the regression that uses the Ohlson-Juettner Naurtoth (2004) model to determine the cost of equity. The full set of control variables is used in these regressions. The control variables reported are respectively the beta of the firm, firm size, market-to-book value and the leverage ratio. Statistical significance is denoted by *, ** and *** which respectively denotes 10%, 5% and 1% significance. Within parentheses the t-statistics are provided.
The tests all include the same control variables as in the continental samples, whereas also industry and year effects are controlled for. The descriptive statistics for these samples are found in Appendix D. Overall, the model fits the data better for the individual country samples as the R2 indicates. The sign and significance varies based upon the model specification indicating that the effects of CSR on the cost of equity are highly dependent on sample construction. For example for the U.S. the effect is found to be negative for the CAPM specification but positive for the implied cost of equity measure. In both instances the effect is highly significant. For the CAPM specification this implies that an increase of 1 unit of the CSR rating leads to a decrease in the cost of equity of 0.01, in contrast the implied cost of equity specification forecast an increase of 0.017. None of the sample countries have significant results with a similar sign for both model specifications. Before sensitivity testing these results, we turn to the evaluation of the CSR and cost of equity relationship based on international orientation of the companies within the sample.
30
relationship between CSR and the cost of equity is found for the domestic firms in all three countries. This is in line with expectations, however a strengthened effect for international firms is not supported.
Table 7
Regression results for the international/domestic sample split.
United States United Kingdom Japan
International Domestic International Domestic International Domestic
(5) (6) (5) (6) (5) (6) CSR -0.012*** (-5.454) -0.014*** (-5.862) 0.002 (0.602) -0.015*** (-3.384) -0.000 (-0.140) -0.008** (-2.576) SIZE -0.088* (-1.882) 0.162*** (3.177) 0.092 (1.484) 0.778*** (8.085) 0.120 (1.567) 0.027 (0.551) MTB -0.021*** (-2.820) -0.018*** (-4.931) -0.026** (-2.249) 0.002 (0.365) 0.053 (0.868) 0.000 (0.155) LEV 0.026*** (2.890) 0.028*** (4.925) -0.053 (-1.085) 0.007 (0.290) -0.088 (-1.019) 0.016 (0.403) INTERCEPT 12.006*** (15.081) 7.211*** (8.989) 7.035*** (7.711) -2.730* (-2.002) 8.490*** (7.373) 9.660*** (11.977) Year effects Yes Yes Yes Yes Yes Yes Industry effects Yes Yes Yes Yes Yes Yes
N 2125 2239 833 471 654 733
Adj. R2 0.048 0.134 0.081 0.248 0.282 0.281
This table provides an overview of the regression results, in which the CSR rating is used as an explanatory factor next to the more conventional control variables in order to evaluate the relationship between CSR and the cost of equity. The regression is conducted separately for the domestic (6) and the international companies (5) within the U.S., the
and Japan sample, ranging from 2008-2013. Under both model specifications the cost of equity is based on the CAPM measure of the cost of equity. The control variables reported are respectively firm size, market-to-book value and the leverage ratio. Statistical significance is denoted by *, ** and *** which respectively denotes 10%, 5% and 1% significance. Within parentheses the t-statistics are provided.
5.3 Robustness tests
31
of these robustness tests. The results are in line with those previously reported based upon the weighted average CSR score.
Secondly, the results so far may suffer from a sample bias regarding the time interval that is taken into account. In Table 2 it is reported that at the end of 2008 the financial markets were still going through the aftermath of the financial crisis, which resulted in high cost of equity estimates for especially the developed markets. Furthermore, the sovereign debt crisis in 2011 is possibly also a factor leading to changing market conditions which could bias the results of this study. Hence, additional tests are performed in which 2008 and 2011 are removed from the sample, separately and simultaneously. The results are not significantly different from those previously reported and are omitted from this thesis due to space constraints.
Thirdly, in this thesis the estimation of the implied cost of equity is conducted using current values of the return on equity and the pay-out ratio to make predictions about short-term growth rates. This assumption comes with the limitation that near-term growth estimates may be biased due to shocking events or, for example, for less mature firms that do not experience constant growth. One could argue that it is more realistic to use average or median values in order to estimate future EPS. Hence, alternative model specifications are tested that use the 5-year mean and median return on equity, instead of the most recent value, to forecast growth in EPS. An advantage of applying this method is that it limits outliers based upon a particular poor or good year in terms of EPS/ROE andlimits bias related to accounting practices. Under these alternate specifications the results remain similar to those previously reported4.
Lastly, the results are further cross-validated by using different cost of equity estimation techniques. The multivariate regression is repeated using different combinations of calculating the implied cost of equity estimations as the weighted average dependent variable of this study. For example, the OJ-model measures based on different near-term growth expectations are used as the weighted average implied cost of equity estimate. Furthermore, an extreme case of the Easton (2004) specification, the forward EP ratio, that assumes no abnormal earnings growth is used as a proxy for the cost of equity. These different specifications deliver similar results as those reported in this thesis4. Hence, overall the results are robust to alternative model specifications and remain unchanged in the conducted sensitivity tests.
4 Robustness checks using median and mean return on equity to forecast near-term EPS growth are
32
6. Conclusions
The thesis set out to investigate the effect of CSR on the cost of equity in an international setting post-crisis. The study expands the existing literature by testing the relationship for multiple continents and countries, while also making a distinction between international and domestic firms. The hypotheses, based on current literature, formulated the expectation of a significant positive relationship between CSR and the cost of equity (H1), with a strengthened relationship for developed markets (H2) and internationally oriented firms (H3). Firstly, based on a continental sample of 12,022 firm year observations ranging from 2008 to 2013, no conclusive evidence is found for a significant positive relationship between CSR and the cost of equity. Secondly, when we dive further into the sample and test the relationship for individual countries, we similarly find no evidence supporting the hypothesis of this study. Thirdly, despite finding significantly higher levels of CSR for international firms, no significant evidence is found that supports the hypothesis that international firms enjoy a stronger cost of equity benefit from CSR activities than domestic firms. All conducted tests include firm-specific explanatory variables and control for industry and year effects. Alternative specifications that include different cost of equity measures or CSR constructs lead to similar results. This thesis underscores the crucial role that the estimation of the cost of equity plays in this line of research. Significant results are presented, however the sign of the significant relationship depends on the construct.
This thesis has reaffirmed that the strategic investment decision to engage in CSR activities is a pressing concern for financial managers. The results of the El Ghoul et al. (2011) study are not easily replicated internationally and across time, which indicates that the negative relationship of CSR on the cost of equity is not beyond doubt. It is critical that financial managers that aim to raise capital in the financial markets take into account firm characteristics but even more importantly that they take into account the specific needs of their stakeholders. The required rate of return by investors is not necessarily found to be similar across the globe, while investors may also be more rewarding to particular CSR activities. Hence, it is important that managers evaluate their environment and keep a close eye on the increasing public pressure for CSR activities.
33
on the cost of equity have to deal with the limitation that the cost of equity is not directly observed. It is shown in this thesis that the method of estimation of the cost of equity construct has a significant impact on the results. Hence, it is very important to carefully evaluate the procedure of measuring the cost of equity and to include robustness tests based on alternative specifications. In this thesis historic values are used as a substitute for unavailable analyst forecasts, which may bias the implied cost of equity estimates used here. Therefore, it may be interesting to repeat this study using the I/B/E/S analyst forecasts. However, as this study shows, the method of estimating the cost of equity is highly influential, hence it may also be insightful to apply further estimation techniques to the El Ghoul et al. (2011) data to further verify these results. Simultaneously using the Thomson Reuters CSR data with the KLD data is also an option to further verify the results.
34
7. References
Aguinis, H., & Glavas, A. (2012). What we know and don’t know about corporate social responsibility a review and research agenda. Journal of Management, 38(4), 932-968.
Attig, N., Guedhami, O., & Mishra, D. (2008). Multiple large shareholders, control contests, and implied cost of equity. Journal of Corporate Finance, 14(5), 721-737.
Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1), 99-120.
Bénabou, R., & Tirole, J. (2010). Individual and corporate social responsibility. Economica, 77(305), 1-19.
Brammer, S., Jackson, G., & Matten, D. (2012). Corporate social responsibility and institutional theory: New perspectives on private governance. Socio-Economic Review, 10(1), 3-28.
Branco, M. C., & Rodrigues, L. L. (2006). Corporate social responsibility and resource-based perspectives. Journal of Business Ethics, 69(2), 111-132.
Callan, S. J., & Thomas, J. M. (2009). Corporate financial performance and corporate social performance: an update and reinvestigation. Corporate Social Responsibility and Environmental Management, 16(2), 61-78.
Campbell, J. L. (2007). Why would corporations behave in socially responsible ways? An institutional theory of corporate social responsibility. Academy of Management Review, 32(3), 946-967. Cao, Y., Myers, J. N., Myers, L. A., & Omer, T. C. (2015). Company reputation and the cost of equity
capital. Review of Accounting Studies, 20(1), 42-81.
Chapple, W., & Moon, J. (2005). Corporate social responsibility (CSR) in asia a seven-country study of CSR web site reporting. Business & Society, 44(4), 415-441.
Chen, H., Chen, J. Z., Lobo, G. J., & Wang, Y. (2011). Effects of Audit Quality on Earnings Management and Cost of Equity Capital: Evidence from China*. Contemporary Accounting Research, 28(3), 892-925.
Chen, K. C., Chen, Z., & Wei, K. J. (2009). Legal protection of investors, corporate governance, and the cost of equity capital. Journal of Corporate Finance, 15(3), 273-289.
Cheng, B., Ioannou, I., & Serafeim, G. (2014). Corporate social responsibility and access to finance. Strategic Management Journal, 35(1), 1-23.
Clarkson, M. E. (1995). A stakeholder framework for analyzing and evaluating corporate social performance. Academy of Management Review, 20(1), 92-117.
Claus, J., & Thomas, J. (2001). Equity premia as low as three percent? Evidence from analysts' earnings forecasts for domestic and international stock markets. Journal of Finance, 56(5), 1629-1666. Contrafatto, M. (2014). The institutionalization of social and environmental reporting: An Italian
35
Dahlsrud, A. (2008). How corporate social responsibility is defined: an analysis of 37 definitions. Corporate Social Responsibility and Environmental Management, 15(1), 1.
Damodaran, A. (2012). Investment valuation: Tools and techniques for determining the value of any asset (Vol. 666): John Wiley & Sons.
Deng, X., Kang, J.-k., & Low, B. S. (2013). Corporate social responsibility and stakeholder value maximization: Evidence from mergers. Journal of Financial Economics, 110(1), 87-109. Dhaliwal, D., Heitzman, S., & Li, O. Z. (2006). Taxes, leverage, and the cost of equity capital. Journal
of Accounting Research, 44(4), 691-723.
Dhaliwal, D., Li, O. Z., Tsang, A., & Yang, Y. G. (2011). Voluntary nonfinancial disclosure and the cost of equity capital: The initiation of corporate social responsibility reporting. The Accounting Review, 86(1), 59-100.
Di Giuli, A., & Kostovetsky, L. (2014). Are red or blue companies more likely to go green? Politics and corporate social responsibility. Journal of Financial Economics, 111(1), 158-180.
Dobers, P., & Halme, M. (2009). Corporate social responsibility and developing countries. Corporate Social Responsibility and Environmental Management, 16(5), 237-249.
Donaldson, T., & Preston, L. E. (1995). The stakeholder theory of the corporation: Concepts, evidence, and implications. Academy of Management Review, 20(1), 65-91.
Du, S., Bhattacharya, C. B., & Sen, S. (2010). Maximizing business returns to corporate social responsibility (CSR): The role of CSR communication. International Journal of Management Reviews, 12(1), 8-19.
Easton, P. D. (2004). PE ratios, PEG ratios, and estimating the implied expected rate of return on equity capital. The Accounting Review, 79(1), 73-95.
Easton, P. D. (2006). Use of forecasts of earnings to estimate and compare cost of capital across regimes. Journal of Business Finance & Accounting, 33(3‐4), 374-394.
Easton, P. D., & Sommers, G. A. (2007). Effect of analysts' optimism on estimates of the expected rate of return implied by earnings forecasts. Journal of Accounting Research, 45(5), 983-1015. Edmans, A. (2011). Does the stock market fully value intangibles? Employee satisfaction and equity
prices. Journal of Financial Economics, 101(3), 621-640.
El Ghoul, S., Guedhami, O., Kwok, C. C., & Mishra, D. R. (2011). Does corporate social responsibility affect the cost of capital? Journal of Banking & Finance, 35(9), 2388-2406.
Fama, E. F., & French, K. R. (2004). The capital asset pricing model: Theory and evidence. Journal of Economic Perspectives, 18, 25-46.
Fatemi, A., Fooladi, I., & Tehranian, H. (2015). Valuation effects of corporate social responsibility. Journal of Banking & Finance, 59, 182-192.
36
Flammer, C. (2015). Does product market competition foster corporate social responsibility? Evidence from trade liberalization. Strategic Management Journal, 36(10), 1469-1485.
Freeman, R. E. (1983). Strategic management: A stakeholder approach. Advances in strategic management, 1(1), 31-60.
Gebhardt, W. R., Lee, C. M., & Swaminathan, B. (2001). Toward an implied cost of capital. Journal of Accounting Research, 39(1), 135-176.
Goedhart, M. H., Koller, T., & Wessels, D. (2015). Valuation: measuring and managing the value of companies: John Wiley & Sons.
Goedhart, M. H., Koller, T. M., & Williams, Z. D. (2002). The real cost of equity. McKinsey & Company, 5, 11-15.
Gordon, M. J., & Shapiro, E. (1956). Capital equipment analysis: the required rate of profit. Management Science, 3(1), 102-110.
Goss, A., & Roberts, G. S. (2011). The impact of corporate social responsibility on the cost of bank loans. Journal of Banking & Finance, 35(7), 1794-1810.
Hail, L., & Leuz, C. (2006). International differences in the cost of equity capital: Do legal institutions and securities regulation matter? Journal of Accounting Research, 44(3), 485-531.
Hess, D., Rogovsky, N., & Dunfee, T. W. (2002). The next wave of corporate community involvement: Corporate social initiatives. California Management Review, 44(2), 110-125.
Hill, R. P., Ainscough, T., Shank, T., & Manullang, D. (2007). Corporate social responsibility and socially responsible investing: A global perspective. Journal of Business Ethics, 70(2), 165-174. Hong, H., Kubik, J. D., & Solomon, A. (2000). Security analysts' career concerns and herding of
earnings forecasts. The Rand Journal of Economics, 31(1), 121-144.
Hou, K., Van Dijk, M. A., & Zhang, Y. (2012). The implied cost of capital: A new approach. Journal of Accounting and Economics, 53(3), 504-526.
Islam, M. A., & Deegan, C. (2010). Media pressures and corporate disclosure of social responsibility performance information: a study of two global clothing and sports retail companies. Accounting and Business Research, 40(2), 131-148.
Jacob, C. K. (2012). The impact of financial crisis on corporate social responsibility and its implications for reputation risk management. Journal of Management and Sustainability, 2(2), p259. Jha, A., & Cox, J. (2015). Corporate social responsibility and social capital. Journal of Banking &
Finance, 60, 252-270.
Jones, T. M. (1995). Instrumental stakeholder theory: A synthesis of ethics and economics. Academy of Management Review, 20(2), 404-437.
Kemper, A., & Martin, R. L. (2010). After the fall: The global financial crisis as a test of corporate social responsibility theories. European Management Review, 7(4), 229-239.
37
Krüger, P. (2015). Corporate goodness and shareholder wealth. Journal of Financial Economics, 115(2), 304-329.
Lane, P. R. (2012). The European sovereign debt crisis. The Journal of Economic Perspectives, 26(3), 49-67.
Lenssen, G., Blagov, Y., Bevan, D., Peters, S., Miller, M., & Kusyk, S. (2011). How relevant is corporate governance and corporate social responsibility in emerging markets? Corporate Governance: The International Journal of Business in Society, 11(4), 429-445.
Lindgreen, A., & Swaen, V. (2010). Corporate social responsibility. International Journal of Management Reviews, 12(1), 1-7.
Lintner, J. (1965). The valuation of risk assets and the selection of risky investments in stock portfolios and capital budgets. The Review of Economics and Statistics, 13-37.
Lopatta, K., & Kaspereit, T. (2014). The world capital markets’ perception of sustainability and the impact of the financial crisis. Journal of Business Ethics, 122(3), 475-500.
Maon, F., Swaen, V., & Lindgreen, A. (2015). One vision, different paths: an investigation of corporate social responsibility initiatives in Europe. Journal of Business Ethics, 8(8), 1-18.
Markowitz, H. (1952). Portfolio selection. the Journal of Finance, 7(1), 77-91.
McWilliams, A., & Siegel, D. S. (2010). Creating and capturing value: Strategic corporate social responsibility, resource-based theory, and sustainable competitive advantage. Journal of Management, 37(5), 1480-1495.
Mossin, J. (1966). Equilibrium in a capital asset market. Econometrica: Journal of the Econometric Society, 34(4), 768-783.
Ohlson, J. A., & Juettner-Nauroth, B. E. (2005). Expected EPS and EPS growth as determinantsof value. Review of Accounting Studies, 10(2-3), 349-365.
Post, J. E., Preston, L. E., & Sachs, S. (2002). Managing the extended enterprise: The new stakeholder view. California Management Review, 45(1), 6-28.
Russo, A., & Perrini, F. (2010). Investigating stakeholder theory and social capital: CSR in large firms and SMEs. Journal of Business Ethics, 91(2), 207-221.
Scott, W. R. (2004). Institutional theory. Encyclopedia of Social Theory, 11, 408-414.
Sharfman, M. P., & Fernando, C. S. (2008). Environmental risk management and the cost of capital. Strategic Management Journal, 29(6), 569-592.
Sharfman, M. P., Shaft, T. M., & Tihanyi, L. (2004). A model of the global and institutional antecedents of high-level corporate environmental performance. Business & Society, 43(1), 6-36.
Sharpe, W. F. (1964). Capital asset prices: A theory of market equilibrium under conditions of risk. Journal of Finance, 19(3), 425-442.
38
