Corporate Environmental Responsibility and Corporate Financial Performance; a study on reported energy use and emissions

(1)

Corporate Environmental Responsibility and Corporate

Financial Performance; a study on reported energy use

and emissions

Abstract:

This study investigates the relation between Corporate Financial Performance and firm risk with Corporate Environmental Responsibility for a MSCI Europe sample from 2003 till 2017. This study does not find statistical evidence for the relation between CER and CFP or firm risk examining energy use and reported emissions with ROA, ROE, Tobin’s Q, excess returns, Altman Z-score and volatility. Opposite to the hypotheses of this study states, lower levels of energy use and reported emissions do not have a positive effect on financial performance measures and do not decrease firm risk. Hence, this study provides evidence that there is no substantial relationship between reported energy use and CO₂, NOX, SOX emission and both financial performance in all

industries due to the direct approach this study uses on firms’ energy use and emissions.

Master thesis

Name: Derk J. Kamermans

Study program: Master of Science Finance Focus area: Energy & Finance

University: University of Groningen Supervisor: Prof. Dr. L.J.R. Scholtens Student number: s2361469

Email: d.j.kamermans@student.rug.nl

Date: 10 January 2019

(2)

2

2.3. Hypotheses ... 9 3. Methodology ... 11 3.1. Methodology ... 11 3.2 Variables ... 11 3.2.1 Dependent variables ... 11 3.2.2 Independent variables ... 13 3.2.3 Control variables ... 14 3.3 Econometric model ... 15 4. Data ... 17 4.1 Data collection ... 17 4.2 Data description ... 18 5. Results ... 22 5.1 Results ... 22

5.2 Robustness and sensitivity ... 27

6. Conclusion ... 29

7. References ... 30

8. Appendices ... 33

Appendix A: Table of descriptive statistics of percentage change variables ... 33

Appendix B: Overview of definitions ... 34

Appendix C: Descriptive statistics of TRES ... 36

Appendix D: Absolute values of CER on CFP ... 37

(3)

3

Appendix F: Basic Materials TRES subsample ... 43

Appendix G: Financials TRES subsample ... 46

Appendix H: Full sample excluding Basic Materials TRES ... 49

Appendix I: Full sample excluding Financials TRES ... 52

Appendix J: Jarque-Bera test results of CER measures ... 55

(4)

4

1. Introduction

This study investigates the relation between Corporate Environmental Responsibility (hereafter written as CER) and both Corporate Financial Performance (hereafter indicated by CFP) as well as firm risk for companies listed at the MSCI Europe Index. Doing so, I examine the relation between firms reported energy use, CO₂(Carbon Dioxide), NOX (Nitrogen Oxide), SOX (Sulphur Oxide)

emission and both financial performance measures and firm risk measures. Existing literature finds a positive effect of CER on both CFP and firm risk (Sharfman and Fernando, 2008; Guenster et al., 2011; Jo, Kim and Park, 2015; Cai, Cui and Jo, 2016). As CER engagement tends to increase financial performance, this study investigates the relation between CER and accounting based performance measures such as ROA, ROE and Altman Z-score and Tobin’s Q, excess return and volatility as market based performance measures.

A vast amount of the studies regarding CER and CFP is based on the environmental pillar scores of the ESG scores that are widely used in CSR literature (Orlitzky, Schmidt and Rynes, 2003; Hu, Wang and Xie, 2018). However, CER is studies in a large variety of approaches. Several studies investigate the relation between emission and CFP (Hart and Ahuja, 1996; Konar and Cohen, 2001; Wang, Li and Gao 2014) and market valuation (King and Lenox, 2001). Others examine the effect of environmental costs on CFP (Jo, Kim and Park, 2015), the relation between environmental performance and the cost of capital (Fernando and Sharfman, 2008; El Ghoul et al., 2011; Chava, 2014), and on investment of ‘green’ stock investors (Heinkel, Kraus and Zechner, 2001) and the exposure to environmental risk and its environmental friendliness (Cai, Cui and Jo, 2016; Fernando, Sharfman and Uysal, 2017).

(5)

5 Cohen, 2001) and increases in non-Green R&D naturally leads to substantial poorer environmental performance than for other industries (Lee and Min, 2015). However, they both lack to compare the effect of reduction of energy use and pollutions between high energy intensive and polluting industries and low energy- and pollution-intensive firms.

Neoclassicals such as Friedman (1970), argue that social responsibilities are not meant for businesses and the only responsibility of the firm is to increase its profits and serve stakeholders’ interest. However, the world is changing and nowadays people are more concerned about the environment and the contribution of society and companies. Friede, Busch and Bassen (2015) find in a recent meta-analysis using over 2000 empirical studies that corporate environmental responsibility has a significant positive outcome with corporate financial performance of a company. Therefore, this study states the following research question;

Research question: Can corporate environmental responsibility increase corporate financial performance and reduce firm risk by reducing energy use and CO₂, NOx, SOx emission?

(6)

6

2. Literature review

This section describes how Corporate Environmental Responsibility within the literature is defined. Next to that, existing literature is summarized and finance theory is evaluated to relate why CER could be able to increase financial performance and diminish risk. Furthermore, I elaborate on the influence of environmentally responsible on financial performance and explain the reasoning that companies present in an energy- and emission intensive industry might face a stronger impact of corporate environmental responsibility on financial performance than companies that are doing business in low energy- and pollution-intensive industries. Lastly, I state my hypotheses.

2.1. Defining Corporate Environmental Responsibility

CER is becoming increasingly important in literature and corporate landscape, both in how management can adopt CER in their business strategy as well as to what extent CER will have a positive effect on financial performance (Berchicci and King, 2007). Academics have widely concluded that CER is an essential part of Corporate Social Responsibility (Porter and van der Linde 1995, Crane, Matten and Moon (2004) and will become an increasingly highlighted topic in the future. As CSR is measured by Environmental, Social and Corporate Government (ESG) measures of companies CER focuses on the environmental pillar of CSR. CER is generally in the three main categories resource efficiency, pollution reduction and environmental R&D or innovation by well-established databases and literature.

2.2. Corporate Environmental Responsibility and Corporate Financial Performance

(7)

7 ROS as a measure of CFP and Emission Reduction an independent variable in which this variable reflects a summary of reported emissions of selected pollutants from U.S. manufacturing companies. Their conclusion states that a reduction in emissions gives firms a cost advantage since it improves efficiency as a results of improvement of reputation and a decrease in the cost of capital. Nevertheless, the question reverse causality in the relation between emissions reduction by profitable firms remains unanswered. Lower emissions could lead to increasing profitability or profitable firms allocate larger share of their time and money on pollution prevention and emissions reducing undertakings.

Following up on the environmental friendly managements of firms, Konar and Cohen (2001) extend the research on social issues of management literature. Their research examines manufacturing firms in the S&P 500 and show that poor environmental performance has a negative effect on intangible assets. This effect is called the social impact and reputation-building hypothesis and suggests that the indirect social impact of the CER could results in endogeneity problems in Finance research. Konar and Cohen report that a reduction in toxic chemicals emissions has a significant positive effect on market value of a company based on a study of fines in environmental lawsuits. In addition, King and Lenox (2001) find evidence between lower emission levels and Tobin’s Q analysing 652 U.S. manufacturing firms over the time period 1987-1996. Similar to Konar and Cohen, they question the problem of causality between emissions and CFP although they suggest that specific characteristics and strategic position of firms could be the cause of the relation. It remains unclear whether causality in pollution and emission numbers is evident to relate to financial performance under the assumption that it captures CER engagement of firms. Opposing to Hart and Ahuja (1996) and Konar and Cohen (2001), Wang, Li and Gao (2014) find that Tobin’s Q often correlates with higher GHG emissions across all industry sectors. They argue that the contradicting finding is due to Australian market specifics such as the large presence of metal and mining companies which are emission intensive, strong lobby groups from emission-intensive industries among other reasons that promote economic development over environmental performance.

(8)

(9)

9

2.3. Hypotheses

Reducing energy use and emissions could have a positive impact when costs of electricity, gas, oil and possible Emissions Trading Scheme (ETS) results in decreasing costs while public awareness of CER could increase the image and reputation of companies. On the other hand, incremental investments in CER (e.g. use of more efficient machines, reduce waste by recycling activities, changing to cleaner energies) which are required for the reduction of energy use and emissions could possibly decrease returns. Among others, Fernando, Sharfman and Uysal (2017) find lower values of Tobin’s Q for green firms, “indicating that corporate expenditures to enhance greenness beyond the mitigation of environmental risk exposure do not increase firm value”. Among others, Hu, Wang and Xie (2018) do find evidence between CER and both ROA and Tobin’s Q, while Hart and Ahuja (1996) found significant evidence between emission reduction and both ROA and ROE since environmental performance decreases operational costs, increases brand image and social reputation and consequently increases CFP. Friede, Busch and Bassen (2015) find mainly favourable relations (58,7% vs. 4,3%) regarding environmental studies related to CFP in their meta-analysis. Therefore, this study states the following hypothesis for CER and CFP;

Hypothesis 1: Corporate Environmental Responsibility increases Corporate Financial

Performance

Hence, this study test whether a reduction in energy use and CO₂, NOX and SOX emission increases

ROA, ROE, Tobin’s Q and excess returns. In following of the reasoning why CER can increase financial performance because it enhances social reputation, it also reduces conflicts of interest between managers and environmental activists. Sharfman and Fernando (2008) suggest that CER has positive effect on firm risk. Improved environmental risk management is negatively related with cost of capital by alleviation of the adverse influence on firms’ cash flows from expected financial, social, or environmental crisis. Additionally, shareholders prefer the ‘insurance-like’ protection for the intangible asset (Godfrey, 2005), others state that CER lowers the cost of capital (e.g. El Ghoul et al., 2011; Chava, 2014; Ng and Rezaee, 2015) which indicates the level of risk of the company. Consequently, reductions in the cost of capital and an insurance-like protection reduce market risk to shareholders and the risk of bankruptcy of the company. Hence, this study states the following hypothesis concerning CER and firm risk;

Hypothesis 2: Corporate Environmental Responsibility decreases firm risk

Therefore, this study tests whether a reduction in energy use and CO₂, NOX and SOX emission

(10)

10 industries experience a stronger effect, others suggests that is it a result of regional or industry characteristics. However, they lack to provide investigate the characteristics of energy intensity of industries or pollution- or emission-levels of industries. Hart and Ahuja (1996) report that pollution intensive sectors such as petrochemicals and forest product did not achieved dramatic levels of emission reduction and therefore did not picked the “low hanging fruits” yet. CER could have stronger effect for companies that are more environmentally hazardous than banking companies that are the most eco-friendly sector of the low energy- and pollution-intensive financial services sector (Jo, Kim and Park, 2015). Cai, Cui and Jo (2016) suggest that CER initiatives and policies reduce risk for U.S. manufacturing firms are environmental hazardous while CER engagement increases risk for firms active in service industries. Companies active in the chemicals, metals & mining, construction materials, paper & forest, tends to be high energy-intensive (e.g. in use of oil, gas and electricity) and high polluting (e.g. CO2, NOX and SOX) industries. On the other hand, the

financial services sector is a relatively low energy- and polluting-intensive industry. The losses of market value do due Toxic Release reports or environmental lawsuits are largest for chemical (31.2%), manufacturing (29.7%), primary metals (27.7%) and paper (21.1%) industries (Konar and Cohen, 2001). This finding of Konar and Cohen suggests that, the larger the intensity of energy and pollution of the industry, the larger the possible impact of CER. Therefore, this study states the following hypothesis concerning differences in industry sectors based on the level of their energy- and pollution-intensity.

Hypothesis 3: The effect of Corporate Environmental Responsibility on both CFP and firm risk

for high energy- and intensive sectors is larger than for low energy- and pollution-intensive sector.

(11)

11

3. Methodology

This section describes the methodology used in this study to answer the research question. I elaborate on the assumptions regarding the research method, the errors and the regression model. Next, I discuss the dependent firm performance variables, the independent variables and control variables. Lastly, I discuss the econometric model that this research uses to test significance in the estimation model and to answer the research question.

3.1. Methodology

Similar to Jo, Park and Kim (2015) in their paper concerning CER and Firm Performance, this study uses a regression model with Ordinary Least Squares (OLS) containing of longitudinal data. First, I perform Hausman-test of random-effects for both entity effects and period effects. The test rejects Hausmans’ null hypothesis, the random effects model is not appropriate and fixed entity and fixed period effects model is preferred in the models with; ROA, ROE, Altman Z-score, excess return and volatility as dependent variables. The Hausman test fails to reject the null hypothesis for random period effects for Tobin’s Q dependent variable. Consequently, the regressions of this dependent variable will include solely period fixed effects. Prior to the analysis, the Gustav-Markov assumption of OLS are checked. This study uses the Jarque-Bera test to analyse extreme outliers and fat tails that cause skewness and kurtosis in the data testing for normal distribution. The Jarque-Bera null hypothesis of normality is rejected for all dependent variables. As in many other studies amongst which Cai, Cui and Jo (2016) and Hu, Wang and Xie (2018), I use winsorizing at the 1st percentile and the 99th percentile to exclude extreme outliers that the cause a non-normal distribution. This study uses White’s test to identify heteroskedasticity in the data. I analyse the Durban-Watson test statistic for detection of serial correlation. In all tests I do not find significant evidence for serial correlation, for which reason I do not indicate further issues. Thus, I apply White’s heteroskedastic consistent standard error standard errors whenever detected.

3.2 Variables

3.2.1 Dependent variables

(12)

12 and Fouts, 1997) and a decrease of cost of capital (El Ghoul et al., 2011; Chava, 2014; Ng and Rezaee, 2015).

𝑅𝑒𝑡𝑢𝑟𝑛 𝑜𝑛 𝐴𝑠𝑠𝑒𝑡𝑠 = 𝑁𝑒𝑡 𝑖𝑛𝑐𝑜𝑚𝑒

𝑉𝑎𝑙𝑢𝑒 𝑜𝑓 𝑇𝑜𝑡𝑎𝑙 𝐴𝑠𝑠𝑒𝑡𝑠 (1)

Return on Equity (ROE) represent the profitability of the firm on its equity part of the financing part of the firm by dividing net income of the company over the value of the equity of the firm.

𝑅𝑒𝑡𝑢𝑟𝑛 𝑜𝑛 𝐸𝑞𝑢𝑖𝑡𝑦 = 𝑁𝑒𝑡 𝑖𝑛𝑐𝑜𝑚𝑒

𝑉𝑎𝑙𝑢𝑒 𝑜𝑓 𝑇𝑜𝑡𝑎𝑙 𝐸𝑞𝑢𝑖𝑡𝑦 (2)

Edward I. Altman (1968) examines a study and formula to measure the level of likelihood of bankruptcy, the characteristics of business failures and the indicators and predictors of corporate distress. The formula weights financial ratios of companies and aggregates it to a score. This score is compared to a graded scale, a Z-score below 1.8 indicates a high possibility of business failure. Z-scores between 1.8 and 3 indicates moderately change for bankruptcy whereas Z-score of 3 and above are considered financially stable. Characteristics of business failures are included in the formula with indicators for profitability, leverage, solvency, liquidity and predictions of insolvency.

𝐴𝑙𝑡𝑚𝑎𝑛 𝑍 − 𝑠𝑐𝑜𝑟𝑒 = 1.2 ∗ 𝑋1+ 1.4 ∗ 𝑋2+ 3.3 ∗ 𝑋3+ 0.6 ∗ 𝑋4+ 1.0 ∗ 𝑋5 (3)

With 𝑋₁, 𝑋₂, 𝑋₃, 𝑋₄ and 𝑋₅ representing the following ratios:

𝑋₁ =𝑊𝑜𝑟𝑘𝑖𝑛𝑔 𝑐𝑎𝑝𝑖𝑡𝑎𝑙 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠 𝑋₂ = 𝑅𝑒𝑡𝑎𝑖𝑛𝑒𝑑 𝑒𝑎𝑟𝑛𝑖𝑛𝑔𝑠 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠 𝑋₃ = 𝐸𝐵𝐼𝑇 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠 𝑋4 = 𝑀𝑎𝑟𝑘𝑒𝑡 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝑒𝑞𝑢𝑖𝑡𝑦 𝑇𝑜𝑡𝑎𝑙 𝑙𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠 𝑋₅ = 𝑆𝑎𝑙𝑒𝑠 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠

(13)

13 King and Lenox (2001) and El Ghoul et al. (2011) investigate that CER improves the market valuation of a company, therefore CER would increase the firms’ market value of assets and hereby Tobin’s Q. A reduction in energy use and emissions together with an increase in financial performance could find its casualty in environmental investment but this is a speculation that should be investigated.

𝑇𝑜𝑏𝑖𝑛′𝑠 𝑄 = 𝑀𝑎𝑟𝑘𝑒𝑡 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝑎𝑠𝑠𝑒𝑡𝑠

𝐵𝑜𝑜𝑘 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝑎𝑠𝑠𝑒𝑡𝑠 (4)

As market valuation of companies increase, total returns of the firms’ stock rise. Increasing stock returns drive excess returns, returns that exceed the risk-free rate. The total return index calculates the total return including dividend pay-outs and currency fluctuations. The risk-free rate is based on the ten-year German government bond rate, which is a commonly used risk-free rate in European corporate valuation practices.

𝐸𝑥𝑐𝑒𝑠𝑠 𝑟𝑒𝑡𝑢𝑟𝑛𝑠 = 𝑇𝑜𝑡𝑎𝑙 𝑟𝑒𝑡𝑢𝑟𝑛 − 𝑅𝑖𝑠𝑘 𝑓𝑟𝑒𝑒 𝑟𝑎𝑡𝑒 (5)

This study uses standard deviation as a measure of a firms’ market risk. The higher the volatility of the security, the higher the risk. The yearly volatility is based on the Total Return Index also used to measure excess returns. Sharfman and Fernando (2008) argue that firms reduce risk by avoiding public social and environmental crisis with CER engagement. This would relate to lower standard deviations in the market returns, hence decrease volatility. Yearly volatility is calculated by multiplying the monthly Total Return Index (TRI) and the square root of the number of trading months.

𝑌𝑒𝑎𝑟𝑙𝑦 𝑣𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦 = 𝑀𝑜𝑛𝑡ℎ𝑙𝑦 𝑇𝑅𝐼 ∗ √𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑡𝑟𝑎𝑑𝑖𝑛𝑔 𝑚𝑜𝑛𝑡ℎ𝑠 (6)

3.2.2 Independent variables

As I defined CER in Section 2.1 of this study, I elaborate on the CER measures that this study uses more extensively in this section. The independent variables are based on Thomson Reuters comprehensive ESG database, which included over 400 metrics of over 6000 public companies to measure their ESG Score and data points. Thomson Reuters claims to be one of the leading ESG data providers with over 150 trained ESG content research analysts. ESG data is based on company annual reports, company websites, NGO websites, stock exchange filings, CSR reporting and news sources. Energy use, CO₂ emission, NOX emissions and SOX emission are 4 of the 114

(14)

14 study uses to investigate environmental responsibility are, in contradiction to most other studies, incremental in their origin.

This study measures CER in an incremental value of the energy used by companies in gigajoules and CO₂, NOX and SOX emission in tonnes per fiscal year. Most large companies report and

monitor their energy use in respond to increasingly interest of consumers in their own and businesses carbon footprint and use of energy (Deloitte, 2018)1_{. Companies are currently not}

legally obligated to report energy use which could be a potential endogeneity problem that extreme high, therefore poor in the context of this study, are not reported. However, the United Kingdom already announced that large companies need to report concerning energy use and carbon footprint from April 2019 on. Next to transparency endogeneity issue, the energy used by a company could be renewable energy in some form of wind, solar or hydropower. Data concerning renewable energy percentages or incremental data on renewable energy use is largely absent in Thomson Reuters Eikon and excluded from this research.

All the remaining explanatory variables include emission measures. The most public established emission measure is CO₂ (Carbon Dioxide). Hart and Ahuja (1996) investigate the effect of emission and pollution reduction on firm performance. In their research they report that emission reduction is developed by a formula of the Toxic Release Inventory although they do not report how this is derived to the emissions reduction variable. To specify and investigate different classes of emissions I include fairly well known Nitrogen Oxide (NOX) emission and Sulphur Oxide (SOX)

emission to the explanatory variables. Methane (CH4) would be interesting to investigate since this

the emission of Methane is a broadly discussed topic in the agricultural industry for some time (DNB, 2018)2. However, Thomson Reuters Eikon ESG database does not provide data on Methane emission.

3.2.3 Control variables

Company size and capital expenditures do not one-on-one relate with dependent financial performance measures. Consistent with existing literature, this study uses the logarithm of total assets to control for size, the logarithm of capital expenditures (Konar and Cohen, 2001; Wang, Li and Gao, 2014; Jo, Kim and Park, 2015). The financing structure of a company has impact on financial performance measures, hence this study uses the leverage ratio of the company as a control variable. Lastly, this study controls for the R&D intensity ratio following Wang, Li and Gao (2014). R&D investments result in knowledge enhancement, which leads to product and

1_{Deloitte energy study of business and residential consumers (May 2018)}

https://www2.deloitte.com/insights/us/en/industry/power-and-utilities/energy-study-of-businesses-and-residential-consumers

2_{De Nederlandsche Bank, 2018, De prijs van transitie, een analyse van de economische gevolgen van CO₂-belasting.}

(15)

15 process innovation because it is considered to an investment in technical capital. Moreover, R&D increases productivity and therefore it proxies for long term growth or long-run economic performance (McWilliams and Siegel, 2000).

𝐿𝑜𝑔𝑆𝑖𝑧𝑒 = 𝐿𝑜𝑔(𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠) (7) 𝐿𝑜𝑔𝐶𝑎𝑝𝐸𝑥 = 𝐿𝑜𝑔(𝐶𝑎𝑝𝑖𝑡𝑎𝑙 𝐸𝑥𝑝𝑒𝑛𝑑𝑖𝑡𝑢𝑟𝑒𝑠) (8) 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒 = 𝐿𝑜𝑛𝑔 𝑡𝑒𝑟𝑚 𝑑𝑒𝑏𝑡 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠 (9) 𝑅&𝐷𝑖𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦 = 𝑅𝑒𝑠𝑒𝑎𝑟𝑐ℎ 𝑎𝑛𝑑 𝑑𝑒𝑣𝑒𝑙𝑜𝑝𝑚𝑒𝑛𝑡 𝑐𝑜𝑠𝑡𝑠 𝑆𝑎𝑙𝑒𝑠 (𝑟𝑒𝑣𝑒𝑛𝑢𝑒) (10) 3.3 Econometric model

The model looks as follows;

𝐶𝐹𝑃𝑖,𝑡 = 𝛼𝑖,𝑡+ 𝛽1𝐶𝐸𝑅𝑖,𝑡+ 𝛽2𝐿𝑜𝑔𝑆𝑖𝑧𝑒𝑖,𝑡+ 𝛽3𝐿𝑜𝑔𝐶𝑎𝑝𝐸𝑥𝑖,𝑡 + 𝛽4𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡+

𝛽₅𝑅&𝐷𝑖𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦_𝑖,𝑡 + 𝜀_𝑖,𝑡 (11)

The model is estimated with Panel Ordinary Least Squares. The constant, 𝛼_𝑖 (alpha) controls for the fixed effect and 𝜀𝑖𝑡 for the error term of the model. 𝛽1 captures the amount that CFP increase

as a result of the CER factors; 𝐸𝑛𝑒𝑟𝑔𝑦𝑈𝑠𝑒_𝑖,𝑡, 𝐶0_2𝑖,𝑡, 𝑁0_{𝑋 𝑖,𝑡}, 𝑆0_{𝑋𝑖,𝑡}. 𝛽₂, 𝛽₃, 𝛽₄ and 𝛽₅ display the coefficient of respectively; logarithm of size, logarithm of capital expenditures, leverage and R&D intensity on the dependent CFP measure. The dependent variable 𝐶𝐹𝑃𝑖,𝑡 in this model includes

Return on Assets, Return on Equity, Altman Z-score, Tobin’s Q, excess returns and volatility. In addition, this study regresses energy use, CO₂, NOx, SOx emission increase or reduction measures for CER. All percentage changes are calculated by:

∆𝐶𝐸𝑅_𝑖,𝑡 = 𝐶𝐸𝑅𝑖,𝑡−𝐶𝐸𝑅𝑖,𝑡−1

𝐶𝐸𝑅𝑖,𝑡−1 ∗ 100 % (12)

After which the model looks as follows;

(16)

(17)

17

4. Data

This section firstly outlines how the data of this study is collected. The second part of this section describes the dataset, the source of the data and descriptive tables to give an overview of the data sample this study uses.

4.1 Data collection

To collect the data, this study uses the Eikon database from Thomson Reuters. Thomson Reuters claims to have one of the most established database concerning ESG factors, providing up to 114 environmental KPI’s for over 6000 public listed firms around the world. Similarly to CSR, CER is often measured in scores. In the case that CSR is measured in ESG scores based on the environmental, social and corporate governance performance of a company, CER is measured on a score regarding (natural) resource use, emission and waste (reduction) score and environmental innovation & opportunity score of a company. Numerous academics among Dimson et al. (2015) and Ferrell, Liang and Renneboog (2016) recall the shortcoming of the use of static and delimited measures such as CSR and CER performance scores produced by for instance; Thomson Reuters ESG Scores and MSCI KLD ESG Rating. The scores are aggregated in up to one-hundred elements containing of reported environmental numbers (e.g. environmental R&D expenses) and dummies, thus it is difficult to fully reflect the magnitude of direct relation between environmental measurement and the dependent variables investigated. Using incremental values creates the benefits to investigate a statistically direct relation between a financial performance measure and an environmental performance measure. Therefore, this study investigate the relation between companies’ energy use and reported CO₂, NOx and SOx emissions which I recall as CER in this study. A larger involvement of firms in CER indicates a reduction in energy use or CO₂, NOx and SOx emissions.

Unlike the most other literature, this study uses a sample of European companies instead of US based or Asian companies which could result in different outcomes. My preliminarily sample contains of 429 companies of the MSCI (Morgan Stanley Capital International) Europe Index. The MSCI Europe represents the performance of large- and mid-cap equities across 15 developed countries in Europe. This index covers roughly 85% of the free float-adjusted market capitalization in each country.3_{This study uses a 15-year sample period from 2003 till 2017 because Thomson}

Reuters claims to have decent ESG data coverage since 2002. After I delete all subsidiaries to constitute, this study maintains final sample of 423 unique RICs (Reuters Instrument Codes). To distinguish the firms between industries, this study use the TRES (Thomson Reuters Economic Sector) that categorizes the RIC’s in ten main different industry sectors.

(18)

18 The data is updated for each fiscal year, mainly reported in the last month of the year for most companies. There is some deviation for the report data, some companies reporting yearly in February or March.

In addition to the other variables, this study uses the Total Return Index (TRI) from Thomson Reuters Eikon. In line with Fama and French (1992), I collect monthly data points to generate yearly returns and yearly volatility. Since monthly return create more data point compared to yearly return they assure more accurate measurement than yearly returns. The TRI is already adjusted for dividend pay-outs, dividend repurchases, stock splits and currency fluctuations.

Furthermore, all accounting data such as total assets, total liabilities, EBIT, retained earnings among others are extracted from Thomson Reuters Eikon. All financial data in this study is collected in US Dollars to overcome problems with currency discrepancies.

4.2 Data description

(19)

19

Table 1: The left column of the left side of the table shows the country of headquarter with respectively the number of companies

that have their headquarter in the given country and the percentage of the total sample size in the right column of the left table. The left column Industry displays each Thomson Reuters Business Classification Economic Sector, with respectively the number of companies present in the Economic Sector and the percentage of the total sample size at the right side of the table.

Country of Headquarter Nr. % of total Industry Nr. % of total

United Kingdom 88 20,8% Financials 87 20,6%

France 69 16,3% Industrials 77 18,2%

Germany 62 14,7% Consumer Cyclicals 69 16,3%

Switzerland 39 9,2% Basic Materials 43 10,2%

Sweden 26 6,1% Consumer Non-Cyclicals 35 8,3%

Spain 21 5,0% Healthcare 32 7,6%

Netherlands 20 4,7% Telecommunications Services 21 5,0%

Italy 19 4,5% Utilities 21 5,0%

Denmark 16 3,8% Energy 20 4,7%

Finland 13 3,1% Technology 18 4,3%

Republic of Ireland 11 2,6% Grand Total 423 100,0%

Belgium 10 2,4% Norway 10 2,4% Luxembourg 6 1,4% Austria 5 1,2% Portugal 3 0,7% Isle of Man 1 0,2% Jersey 1 0,2% Mexico 1 0,2% South Africa 1 0,2%

United States of America 1 0,2%

Grand Total 423 100,0%

(20)

20

Table 2: This table shows an overview of the descriptive statistics of the sample used in this study. The left column displays the

variables of the descriptive statistical for each row. The column Mean describes the mean value of the variable. The column Median describes the median value of the variable. The Maximum and Minimum column describe respectively the maximum and minimum value of the variable. The Std. Dev. column describes the standard deviation of the variable. The Observations column describes the number of observations of per variable in the sample used in this study.

Variable Mean Median Maximum Minimum Std. Dev. Observations

Return On Assets (in %) 6.13 4.95 29.78 -2.90 5.68 5193

Return On Equity (in %) 17.40 14.80 103.59 -22.18 16.05 5530

Altman Z-score 2.10 1.93 6.13 -0.09 1.12 4868

Tobin’s Q ratio 0.79 1.15 30.34 -27.68 5.86 6022

Excess returns (in %) 13.60 14.70 324.25 -202.72 33.95 5858

Volatility (in Std. Dev.) 26.13 22.50 235.72 3.22 14.83 5857

Energy use (in Gigajoules) 5.00E+07 3.33E+06 2.90E+09 2 1.84E+08 3643

CO₂emission (in tonnes) 6.67E+06 2.96E+05 2.39E+08 10 2.18E+07 4135

NOx emission (in tonnes) 3.31E+04 1944 1.09E+06 0 9.39E+04 1452

SOx emission (in tonnes) 2.89E+04 1394.5 6.63E+05 0 7.74E+04 1370

LogSize (Logaritm of total assets in USD)

23.65 23.40 28.93 16.57 1.84 6026

LogCapEx (Logaritm of capital expenditures in USD)

19.89 19.98 24.42 10.87 1.76 3868

Leverage ratio 0.19 0.17 2.67 0 0.16 6025

R&D intensity ratio 0.14 0.03 92.30 -0.02 2.27 1677

(21)

21

Table 3: The table below shows the correlation matrix. The correlation between the variables for each column of the first row is

and of each row of the first column is displayed in the matrix. The correlation between the same variables always equals to unity for indisputable reasons.

Variable ROA ROE Tobin’ s Q Altma n Z-score Excess Retur n Volatil ity Energ y Use CO₂ NOX SOX LogSi ze LogCa pEx Lever age R&D Intens ity ROA 1.00 ROE 0.74 1.00 Tobin’s Q 0.74 0.34 1.00 Altman Z-score 0.02 -0.04 0.07 1.00 Excess Return 0.01 -0.01 0.01 -0.05 1.00 Volatility -0.22 -0.17 -0.17 0.00 -0.02 1.00 Energy Use -0.12 -0.13 -0.04 0.01 -0.03 0.07 1.00 CO₂ -0.10 -0.11 -0.05 0.01 -0.04 0.07 0.92 1.00 NOX -0.04 -0.07 0.05 0.00 -0.03 0.00 0.39 0.38 1.00 SOX -0.04 -0.07 0.03 0.02 -0.03 0.05 0.48 0.47 0.94 1.00 LogSize -0.33 -0.13 -0.46 -0.04 -0.08 -0.02 0.34 0.36 0.21 0.20 1.00 LogCapEx -0.28 -0.12 -0.36 -0.02 -0.08 0.03 0.39 0.41 0.29 0.28 0.91 1.00 Leverage -0.23 0.06 -0.46 -0.02 0.03 0.00 -0.04 -0.03 0.00 0.01 0.00 0.01 1.00 R&D Intensity 0.41 0.39 0.24 -0.01 -0.05 -0.16 -0.26 -0.26 -0.14 -0.15 -0.10 -0.26 -0.12 1.00

(22)

22

5. Results

5.1 Results

This section shows the main findings of the study regarding the results of statistical tests and regressions performed in the statistical software program EViews 10. Furthermore, I elaborate on the implications of the findings, and discuss on the similarities and differences of existing literature on CER and CFP. Lastly, this section states the limitations of this study.

This study investigates the relation between CFP and CER. Using a panel dataset, reported energy use, CO₂, NOx and SOx emission values are regressed on the CFP measures as is shown in Appendix D. I find a positive significant relation between CO₂ emission and both ROA, ROE and the Altman Z-score at a 99% confidence level. However, I mainly find insignificant t-statistics of the CER coefficients for the majority of the tests. This results suggests that an increase in CO2

emission increases ROA, ROE and Altman Z-score which is contradicting the hypothesis this study states. The possibility for this result could be that companies that are increasing their performance face a higher CO2 emission due to increases in their production levels Wang, Li and Gao (2014).

Since the probability of the findings in this result are mainly non-significant I do not find evidence for a relation between CER and CFP and firm risk using this model. In economic terms, this result is very hard interpretable since the values of the reported for energy use and emitted CO₂, NOX and

SOX are ten-thousand- to ten-millions-fold of the CFP measures. Thus, the main result of this study

reports increases or reductions of energy use and the emitted CO₂, NOX and SOX in percentages as

the explanatory variables.

Table 4 here below, displays the relation of 2nd dimension percentage change in the CER measures on the dependent CFP measures controlling for firm characteristics such as the size, capital expenditures, leverage ratio and R&D intensity ratio. Appendix E shows the full table including coefficients of control variables and their significance and confidence levels. The control variables are generally consistent with expectations, firm size being negative related with return based financial performance measures due to growth potential, highly leveraged firms face higher volatility of market value and higher risk of bankruptcy.

(23)

23 the replacement cost of the assets. Panel E and F, respectively shows the results of the estimation of excess returns and volatility with the energy use and emissions differences in percentages compared to last’ years reported energy use and emissions.

Table 4: This table shows the results of the Ordinary Least Squares regression of the period 2003-2017 with the percentage change

of the independent variables; energy use, CO₂ (carbon dioxide) emission, NOx (nitrogen oxide) emission and SOx (sulphur oxide) emission and the dependent variables in panel A till panel F; ROA, ROE, Altman Z-score, Tobin’s Q, excess return and volatility. The control variables include the logarithm of the total assets as size of the company, the logarithm of capital expenditures, the leverage ratio and R&D intensity ratio. The OLS model uses both cross-sectional and period fixed effects as indicated. ***, ** and * display the probability respectively at a 99%, 95% and 90% confidence level.

Panel A: ROA dependent variable

Independent (1) (2) (3) (4)

∆Energy use 0.0017**

∆CO2 emission -0.0003

∆NOX emission -0.0010

∆SOX emission -0.0001

Controls Yes Yes Yes Yes

Industry fixed effect Yes Yes Yes Yes

Time fixed effects Yes Yes Yes Yes

𝑹𝟐_{𝑨𝒅𝒋𝒖𝒔𝒕𝒆𝒅} _0.74 _0.73 _0.74 _0.76

N observations 952 1012 535 545

Panel B: ROE dependent variable

Independent (1) (2) (3) (4)

∆Energy use 0.0030*

∆NOX emission 0.0020

𝑹𝟐_{𝑨𝒅𝒋𝒖𝒔𝒕𝒆𝒅} _0.66 _0.65 _0.58 _0.59

N observations 987 1038 550 562

Panel C: Altman Z-score dependent variable

Independent (1) (2) (3) (4)

∆Energy use 1.82E-05

∆NOX emission -0.0004

∆SOX emission -6.14E-06

(24)

24

N observations 966 1014 528 539

Panel D: Tobin’s Q dependent variable

Independent (1) (2) (3) (4)

∆Energy use 0.0058

∆CO2 emission 0.0055

Time fixed effects No No No No

𝑹𝟐_{𝑨𝒅𝒋𝒖𝒔𝒕𝒆𝒅} _0.08 _0.08 _0.02 _0.02

N observations 1002 1051 559 571

Panel E: Excess Return dependent variable

Independent (1) (2) (3) (4)

∆SOX emission 0.0102

𝑹𝟐_{𝑨𝒅𝒋𝒖𝒔𝒕𝒆𝒅} _0.33 _0.31 _0.31 _0.34

N observations 1002 1049 559 571

Panel F: Volatility dependent variable

Independent (1) (2) (3) (4)

∆SOX emission 0.0010

𝑹𝟐_{𝑨𝒅𝒋𝒖𝒔𝒕𝒆𝒅} _0.56 _0.55 _0.58 _0.61

N observations 1002 1049 559 571

(25)

(26)

26 first hypothesis that this study states, a positive effect of CER on financial performance measures ROA, ROE, Tobin’s Q and excess return, as the second hypothesis, a positive relation between CER and firm risk measures Altman Z-score and volatility are rejected.

The third hypothesis of this study states that the positive effect CER and CFP is larger for energy- and pollution-intensive companies then for low energy- and pollution-intensive companies. I compare companies in energy intensive Basic Materials TRES and low energy-intensive Financials TRES separately in appendix F and G and compare them both with the full sample excluding the sectors itself in appendix H and I. As a result of a decreasing size of the subsamples, R&D intensity is removed as discussed in the methodology section. Results of Basic Materials TRES in appendix F in panel A show larger significance levels compared to the sample excluding companies of the Basic Materials TRES displayed in appendix F especially for CO₂ and SOx emission coefficients shown in Panel A model (2) and (4). Nevertheless, the positive coefficient indicates that an increase of CO₂ and SOx emission increases ROA by respectively 0.0372 % and 0.0215 % compared to insignificant negative coefficients of 0.0023 % and 0.0001 % for ROA in other industries displayed in panel A of appendix H. Furthermore, I find a negative coefficient of 0.0009 % that is significant at a 95 % confidence level between CO2 and the Altman Z-score for companies in different sectors

than Basic Materials TRES in appendix H. This result indicates that the score increases if CO2

emission decreases. In addition, variances for Basic Materials TRES in Appendix F are mainly well explained ranging from Adjusted R-squared values of 0.88 to 0.40, except for the R-squared of Tobin’s Q, that ranges from 0.10 to 0.08.

Appendix G and I show the results for Financials TRES and the full sample excluding this sector. This model is unable to test for NOX and SOX emission due to the limited size of this subsample,

therefore only the results of energy use and CO2 emission on CFP measures are displayed. Panel

D of appendix G shows a significant relation between CO2 emission and Tobin’s Q at a 90 %

confidence level. The coefficient indicates that a percent increase in CO2 emission decreases

Tobin’s Q ratio with 0.0163. The economic influence of this results is neglectable. Comparing this result to the sample excluding Financials TRES, no huge differences are observed. I reject the third hypothesis since this study do not find significant results for both energy- and pollution intensive sectors as for low energy- and pollution-intensive sectors. I suggest that the differences in the findings with existing literature of this study regarding energy intensity of sectors by the different approach this study uses. Konar and Cohen (2001) find that energy intensive companies experience larger losses of market value due to environmental lawsuits while Cai, Cui and Jo (2016) find that CER initiatives reduce risk for U.S. energy intensive manufacturing firms while CER engagement increases risk for firms in low energy intensive service industries.

(27)

27 benefits of CER are widely investigated and mainly recognized. However, excessive investment to increase CER beyond of environmental risk exposure does not increase CFP (Fernando, Sharfman and Uysal, 2017). Moreover, strong significant evidence is not obvious in literature regarding the relation between CER and CFP. Friede, Busch and Bassen (2015) indeed show that the majority (58,7 %) of the studies included in their meta-analysis find a positive relation compared to 4,3 % that find a negative relation between CER and CFP. Nevertheless, this indicates that that a large part of the published studies do not find evidence for a relation between CER and CFP. Furthermore, this study uses reported values of emitted Carbon Dioxide, Nitrogen Oxide, Sulphur Oxide and energy use instead of generalized accumulated CER scores or effects of very poor or excellent environmental performance, which could explain deviations in the results compared to existing literature regarding CER. Indirect relations of CER such as general company reputation, branding, public image and perceived risk are not included in this research, whether the positive impact of CER on CFP can be found in these indirect performance measures should be studies in future research. As mentioned before, I speculate that the information of reported energy use, CO₂, NOX, SOX emission is not included in stock prices of financials markets. I suggest future research

to examine an event study that investigated abnormal returns regarding energy use and emission reporting of firms. Next to that, this study suggests that future research investigates whether CER engagement increases image, branding and company reputation.

This study has a few limitations, reported energy use that Thomson Reuters Eikon provides does not specifies the amount of renewable energy use and energy use that is extracted from natural resources. Next to that, to investigate the effect of CER on CFP, I would prefer to data on the costs and revenues of CER project is absent. Therefore, this study uses reported energy use, CO2, NOX

and SOX emissions to examine the effect on CFP. However, these CER measures are 4 of the 114

metrics used by Thomson Reuters to aggregate a companies’ CER score. Hence, energy use and emission do not fully cover CER and omits potential relations between CER and CFP that are not found in this research. Besides that, I suspect errors in the data since I observe extreme increases and reduction of the CER measures used in this study that are illogical after investigation. Next to that, the reporting bias might influence results since companies do not have report poor energy use and emission levels since they are not obligated to do so

5.2 Robustness and sensitivity

(28)

(29)

29

6. Conclusion

This research investigates whether there is a relation between Corporate Environmental Responsibility and Corporate Financial Performance using a MSCI Europe sample from 2003 to 2017. In this respect, I test the relation between energy use, CO₂ emission, NOX emission, SOX

emission and review the effect of a reduction or increase of these measures on the financial performance and risk measures; ROA, ROE, Altman Z-score, Tobin’s Q, excess returns and volatility. Hereby, I attempt to answer the research question this study states; “Can corporate environmental responsibility increase corporate financial performance and reduce firm risk by reducing energy use and CO2, NOx, SOx emission?” To investigate the relation between the energy

use and emissions of companies this study uses Ordinary Least Squares estimation. This study does not find a relation between CFP measures and the levels of energy use and emissions. In addition, in a second test I mainly find insignificant relation between percentage increase/reduction of CER and CFP, however this study find a positive relation between energy use and emissions and the accounting based performance measures. The implication of this result is that an increase level of energy usage or pollution tends to increase financial performance but I speculate that the opposite is true. It is generally more likely that high performing companies face larger energy use and emitted CO₂, NOX emission and SOX due to business expansion with higher levels of production,

(30)

30

7. References

Altman, E. (1968). Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy. The Journal of Finance, 23(4), 589-609.

Berchicci, L., & King, A. (2007). 11 postcards from the edge: a review of the business and environment literature. The Academy of Management Annals, 1(1), 513-547.

Cai, L., Cui, J., & Jo, H. (2016). Corporate environmental responsibility and firm risk. Journal of Business Ethics, 139(3), 563-594.

Chava, S. (2014). Environmental externalities and cost of capital. Management Science, 60(9), 2223-2247.

Crane, A., Matten, D., & Moon, J. (2004). Stakeholders as citizens? Rethinking rights, participation, and democracy. Journal of Business Ethics, 53(1-2), 107-122.

Derwall, J., Guenster, N., Bauer, R., & Koedijk, K. (2005). The eco-efficiency premium puzzle. Financial Analysts Journal, 61(2), 51-63.

Dimson, E., Karakaş, O., & Li, X. (2015). Active ownership. The Review of Financial Studies, 28(12), 3225-3268.

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. (1992). The cross‐section of expected stock returns. the Journal of Finance, 47(2), 427-465.

Fernando, C. S., Sharfman, M. P., & Uysal, V. B. (2017). Corporate environmental policy and shareholder value: Following the smart money. Journal of Financial and Quantitative Analysis, 52(5), 2023-2051.

Ferrell, A., Liang, H., & Renneboog, L. (2016). Socially responsible firms. Journal of Financial Economics, 122(3), 585-606.

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.

(31)

31 Godfrey, P. C., Merrill, C. B., & Hansen, J. M. (2009). The relationship between corporate social responsibility and shareholder value: An empirical test of the risk management hypothesis. Strategic Management Journal, 30(4), 425-445.

Guenster, N., Bauer, R., Derwall, J., & Koedijk, K. (2011). The economic value of corporate eco‐ efficiency. European Financial Management, 17(4), 679-704.

Hart, S. L., & Ahuja, G. (1996). Does it pay to be green? An empirical examination of the relationship between emission reduction and firm performance. Business strategy and the Environment, 5(1), 30-37.

Heinkel, R., Kraus, A., & Zechner, J. (2001). The effect of green investment on corporate behavior. Journal of Financial and Quantitative Analysis, 36(4), 431-446.

Hu, J., Wang, S., & Xie, F. (2018). Environmental responsibility, market valuation, and firm characteristics: Evidence from China. Corporate Social Responsibility and Environmental Management.

Jo, H., Kim, H., & Park, K. (2015). Corporate environmental responsibility and firm performance in the financial services sector. Journal of Business Ethics, 131(2), 257-284.

Karpoff, J. M., Lott, Jr, J. R., & Wehrly, E. W. (2005). The reputational penalties for environmental violations: Empirical evidence. The Journal of Law and Economics, 48(2), 653-675.

King, A. A., & Lenox, M. J. (2001). Does it really pay to be green? An empirical study of firm environmental and financial performance: An empirical study of firm environmental and financial performance. Journal of Industrial Ecology, 5(1), 105-116.

Klassen, R. D., & McLaughlin, C. P. (1996). The impact of environmental management on firm performance. Management Science, 42(8), 1199-1214.

Konar, S., & Cohen, M. A. (2001). Does the market value environmental performance?. Review of Economics and Statistics, 83(2), 281-289.

Lee, K. H., & Min, B. (2015). Green R&D for eco-innovation and its impact on carbon emissions and firm performance. Journal of Cleaner Production, 108, 534-542.

McWilliams, A., & Siegel, D. (2000). Corporate social responsibility and financial performance: correlation or misspecification?. Strategic Management Journal, 21(5), 603-609.

Ng, A. C., & Rezaee, Z. (2015). Business sustainability performance and cost of equity capital. Journal of Corporate Finance, 34, 128-149.

(32)

32 Porter, M. E. (1990). The competitive advantage of nations. Competitive Intelligence Review, 1(1), 14-14.

Porter, M. E., & Van der Linde, C. (1995). Toward a new conception of the environment-competitiveness relationship. Journal of Economic Perspectives, 9(4), 97-118.

Russo, M. V., & Fouts, P. A. (1997). A resource-based perspective on corporate environmental performance and profitability. Academy of Management Journal, 40(3), 534-559.

Sharfman, M. P., & Fernando, C. S. (2008). Environmental risk management and the cost of capital. Strategic Management Journal, 29(6), 569-592.

(33)

33

8. Appendices

Appendix A: Table of descriptive statistics of percentage change variables

Table 5: This table shows an overview of the descriptive statistics of the sample used in this study. The left column displays the

percentage change measures of the independent CER measures. The column Mean describes the mean value of the variable. The column Median describes the median value of the variable. The Maximum and Minimum column describe respectively the maximum and minimum value of the variable. The Std. Dev. column describes the standard deviation of the variable. The Observations column describes the number of observations of per variable in the sample used in this study.

Variable Mean Median Maximum Minimum Std. Dev. Observations

∆Energy use (in %) 16,53 0,42 918,79 -69,71 106,87 3215

∆CO₂emission (in %) 8,03 -0,69 359,44 -67,73 50,53 3675

∆NOx emission (in %) 6,12 -1,77 417,10 -74,67 57,20 1272

(34)

34

Appendix B: Overview of definitions

Table 6: This table shows the description of definitions and variables used in this study.

Definition Thomson Reuters Code Explanation of the definition

Total Return TR.TotalReturn The total return incorporates the price change and any relevant dividends

for the specified period. Compounded daily return for the specified period is used to calculate Total Return and it's effectively the dividend reinvested Total Return methodology. The most recently completed trading day is set as the default period. The Dividend type used is the most widely reported Dividend for a market and it is either Gross or Net.

Country of Headquarters

TR.HeadquartersCountry Country of Headquarters, also known as Country of Domicile.

TRBC Economic Sector Name

TR.TRBCEconomicSector Primary Thomson Reuters Business Classification (TRBC) Economic

Sector Description. TRBC Classifies companies with increasing granularity by Economic Sector, Business Sector, Industry Group, Industry and Activity.

Company Common Name

TR.CommonName Where available provides the name of the organisation most commonly

used. Energy Use

Total

TR.EnergyUseTotal Total direct and indirect energy consumption in gigajoules.

CO2 Equivalent Emissions Total

TR.CO2EmissionTotal Total CO2 and CO2 equivalent emission in tonnes. The value is the sum of

Scope 1 and 2 emissions and excludes Scope 3 emissions.

NOx Emissions TR.NOxEmissions Total amount of NOx emissions emitted in tonnes.

SOx Emissions TR.SOxEmissions Total amount of SOx emissions emitted in tonnes.

EBIT TR.EBIT EBIT is computed as Total Revenues for the fiscal year minus Total

Operating Expenses plus Operating Interest Expense, Unusual Expense/Income [SUIE] and Non-Recurring Items, Supplemental, Total [SUIT] for the same period. This definition excludes non-operating income and expenses.

Tangible Book Value Per Share, Total Equity

TR.TangibleBVPSTotalEquity This is the Tangible Book Value as of the end of the fiscal period divided by Total Common Shares Outstanding for the same period. Tangible Book Value represents Total Equity adjusted for Net Intangibles and Net Goodwill.

Total Equity TR.TotalEquity Consists of the equity value of preferred shareholders, general and limited

partners, and common shareholders, but does not include minority shareholders' interest.

Total Liabilities TR.TotalLiabilities Represents the sum of: Total Current Liabilities, Total Long-Term Debt, Deferred Income Tax, Minority Interest and Other Liabilities, Total. Retained

Earnings (Accumulated Deficit)

TR.RetainedEarnings Represents residual earnings from operations, not distributed to

shareholders. It may represent accumulated deficit when a company incurs losses over time.

Total Assets, Reported

TR.TotalAssetsReported Represents the total assets of a company.

Long Term Debt TR.LTDebt Represents debt with maturities beyond one year. Long-Term Debt may

(35)

35

Revenue TR.Revenue Is used for industrial and utility companies. It consists of revenue from the

sale of merchandise, manufactured goods and services, and the distribution of regulated energy resources, depending on a specific company's industry.

Working Capital TR.WorkingCapital This item is defined as the difference between Current Assets and Current

Liabilities for the fiscal period. Available for Industrial and Utility companies.

Research And Development

TR.ResearchAndDevelopment Represents expenses for research and development of new products and services by a company in order to obtain a competitive advantage. Market Value

for Company

TR.MarketCapDS Market Value of Company (MVC) is the consolidated market value of a

company displayed in local currency. MVC for companies with a single listed equity security is the share price multiplied by the number of ordinary shares in issue. However, for companies with more than one listed or unlisted equity security MVC represents: EQUITY A(MV) + EQUITY B(MV) + EQUITY C(MV) etc. Unlisted securities are valued using the prices of the associated listed security. Dual Listed Companies are consolidated in MVC. Dual Listed Companies occur when two companies operate as a single economic entity under binding agreements. For each company, and for each eligible security of the company, the MVC history is provided on the primary quote only. Secondary quotes have only current data, updated daily.

Capital Expenditures - Actual

TR.CapexActValue Capital Expenditure are the funds used by a company to acquire or upgrade

physical assets such as property, industrial buildings, or equipment or the amount used during a particular period to acquire or improve long term assets such as property, plant, or equipment.

Total Assets - Actual

TR.TotalAssetsActual Total Assets is anything tangible or intangible that is capable of being owned or controlled to produce value and that is held to have positive economic value is considered an asseSimply stated, assets represent ownership of value that can be converted into cash (although cash itself is also considered an asset).

Return On Assets - Actual

TR.ROAActValue Return On Assets is a profitability ratio and as such gauges the return on

investment of a company. Specifically, ROA measures a company’s operating efficiency regardless of its financial structure (in particular, without regard to the degree of leverage a company uses) and is calculated by dividing a company’s net income prior to financing costs by total assets. Return On

Equity - Actual

TR.ROEActValue Return On Equity is a profitability ratio calculated by dividing a company’s

net income by total equity of common shares. Net Working

Capital - Actual

TR.NWCActValue Net Working Capital is the operating liquidity available to a business. Along

(36)

36

Appendix C: Descriptive statistics of TRES

Table 7: This table shows an overview of the statistics of the CER measures used in this study for Basic Materials, Financials and

other Thomson Reuters Economic Sector (TRES). The left column displays the independent CER measures; Energy use (in Gigajoules), CO₂ (Carbon Dioxide) emission, NOx (Nitrogen Oxide) emission and SOx (Sulphur Oxide) emission all displayed in tonnes per sector group. The column Median describes the median value of the variable. The column Mean describes the mean value of the variable. The Maximum and Minimum column describe respectively the maximum and minimum value of the variable.

Sector name: Median Mean Minimum Maximum

Panel A: Basic Materials TRES

Energy use (in gigajoules) 7.95E+07 1.95E+08 1.99E+05 2.90E+09

CO2 emissions (in tonnes) 5.01E+06 1.82E+07 2.29E+04 2.39E+08

NOx emissions (in tonnes) 5556 2.49E+04 32 8.97E+05

SOx emissions (in tonnes) 4503 1.90E+04 1 2.82E+05

Panel B: Financials TRES

Energy use (in gigajoules) 7.59E+05 4.85E+06 2614 1.19E+09

CO2 emissions (in tonnes) 4.25E+04 1.86E+05 139 1.03E+06

NOx emissions (in tonnes) 83 87 0.92 357.49

SOx emissions (in tonnes) 31.76 55.76 0.25 180

Panel C: Other TRES

Energy use (in gigajoules) 4.35E+06 3.84E+07 2 2.26E+09

CO2 emissions (in tonnes) 4.10E+05 6.75E+06 113 1.93E+08

NOx emissions (in tonnes) 1349.5 3.74E+04 0 1.09E+06

(37)

37

Appendix D: Absolute values of CER on CFP

Table 8: This table shows the results of the Ordinary Least Squares regression of the period 2003-2017 with the independent

variables; energy use, CO₂ (carbon dioxide) emission (in gigajoules), NOx (nitrogen oxide) emission and SOx (sulphur oxide) emission (all in tonnes) and the dependent variables in panel A till panel F; ROA, ROE, Altman Z-score, Tobin’s Q, excess return and volatility. The control variables include the logarithm of the total assets as size of the company, the logarithm of capital expenditures, the leverage ratio and R&D intensity ratio. The OLS model uses both cross-sectional and period fixed effects as indicated. ***, ** and * display probability of significance at a 99%, 95% and 90% confidence level.

Panel A: ROA dependent variable

Independent (1) (2) (3) (4)

Energy use 4.57E-10

CO2 emission 2.02E-07*** NOX emission 2.11E-05 SOX emission 2.21E-05** Controls Logsize -2.969282*** -3.593544*** -3.063596*** -4.700548*** Logcapex 0.615758 0.744854* 0.196177 0.541143 Leverage -18.79835*** -18.42457*** -15.41617*** -11.99199*** R&D Intensity -46.91275*** -61.22222*** -51.98810*** -38.64663***

𝑹𝟐_{𝑨𝒅𝒋𝒖𝒔𝒕𝒆𝒅} _0.73 _0.73 _0.74 _0.76

N observations 1016 1068 579 583

Panel B: ROE dependent variable

Independent (1) (2) (3) (4)

Energy use 1.00E-09

CO2 emission 5.51E-07*** NOX emission 5.71E-05 SOX emission 7.14E-05** Controls Logsize -8.374429*** -12.19011*** -13.72553*** -16.64040*** Logcapex 1.404357 1.947465 1.070014 1.740500 Leverage -6.388569 -2.347583 16.96157 21.43230* R&D Intensity -83.86706*** -123.2096*** -128.8457*** -80.29878**

𝑹𝟐_{𝑨𝒅𝒋𝒖𝒔𝒕𝒆𝒅} _0.64 _0.63 _0.60 _0.62

N observations 1052 1097 594 600

Panel C: Altman Z-score dependent variable

Independent (1) (2) (3) (4)

Energy use -4.31E-12

CO2 emission 1.15E-08**

NOX emission 7.16E-07

(38)

38 Controls Logsize -0.282863*** -0.318298*** -0.544047*** -0.563608*** Logcapex -0.026173 0.011411 0.070008 0.083832* Leverage -3.444139*** -3.369486*** -3.07870*** -2.969721*** R&D Intensity -2.771091*** -5.803240*** -3.672662** -2.475676*

𝑹𝟐_{𝑨𝒅𝒋𝒖𝒔𝒕𝒆𝒅} _0.87 _0.88 _0.90 _0.90

N observations 1027 1068 571 575

Panel D: Tobin’s Q dependent variabl

Independent (1) (2) (3) (4)

Energy use 9.16E-09***

CO2 emission 1.11E-07 NOX emission 2.25E-06 SOX emission 1.91E-05 Controls Logsize -0.705789 -0.165522 -0.236025 -0.041496 Logcapex 0.266257 0.149314 0.600874 0.736761 Leverage -6.418499** -5.410152 -7.906050 -6.529613 R&D Intensity 4.557357 16.88366 44.69005 47.38870

Time fixed effects No No No No

𝑹𝟐_{𝑨𝒅𝒋𝒖𝒔𝒕𝒆𝒅} _0.09 _0.07 _0.02 _0.00

N observations 1068 1110 603 575

Panel E: Excess Return dependent variable

Independent (1) (2) (3) (4)

Energy use 2.37E-09

CO2 emission -1.67E-07 NOX emission -7.67E-05 SOX emission -3.97E-05 Controls Logsize -12.21639** -9.797576** -13.54731* -12.73866 Logcapex -15.19201*** -15.66879*** -15.19481*** -14.73894*** Leverage -9.837378 -16.86695 -13.38315 -1.605904 R&D Intensity -122.2557*** -183.3174** -177.5246** -243.2695**

𝑹𝟐_{𝑨𝒅𝒋𝒖𝒔𝒕𝒆𝒅} _0.34 _0.33 _0.32 _0.32

N observations 1068 1108 603 609

Panel F: Volatility dependent variable

Independent (1) (2) (3) (4)

Energy use 1.86E-10

(39)

39 NOX emission -1.59E-05 SOX emission -1.77E-07 Controls Logsize -2.003004 -1.023804 -8.044148** -8.290082** Logcapex -0.246654 -2.405355 1.014303 1.324757 Leverage 12.05435* 9.962438 20.05534*** 20.19025*** R&D Intensity 7.513766 -5.804259 -60.80986 -40.00429

𝑹𝟐_{𝑨𝒅𝒋𝒖𝒔𝒕𝒆𝒅} _0.56 _0.52 _0.58 _0.59

Corporate Environmental Responsibility and Corporate Financial Performance; a study on reported energy use and emissions