Does Climate Change Really Affect Our Businesses?
An Empirical Study of Carbon Dioxide Emissions and
Corporate Financial Performance
Name: Biya Du, Student number: S2647435University of Groningen, Faculty of Business and Economics, International Financial Management
Supervisor: Prof. Dr. Bert Scholtens Date of the submission: 12th of June.
Abstract
This paper investigates how firms’ carbon dioxide (CO2) emissions relate to their corporate financial performance. The European Union Emissions Trading System (EU ETS) establishes a CO2 trading policy to limit corporate CO2 emission in response to climate change. I analyse 924 companies’ CO2 emissions from the EU ETS regulated industries and non-regulated industries and their effects on four different corporate financial indicators (return on equity, stock return, volatility of stock price and Tobin’s q) in several advanced economies (the EU, the US, Canada and Australia) in the period of 2005 to 2015 ( the EU ETS Phases I-III). I apply a pooled ordinary least square (OLS) model to the aforementioned four economic areas and to each EU ETS phase. This study complements and extends the work of previous researchers. I find that an increase of CO2 emissions in the EU and Australia has a weak negative impact on return on equity and Tobin’s q, possibly due to the implementation of emission policies. These policies encourage companies to invest in low-carbon technologies while maintaining high-efficiency operations. I also find that CO2 emission has little impact on US and Canadian firms’ financial performance. This is mainly because the US and Canada do not have an influential emission control policy. Furthermore, the result shows in the EU ETS Phase I and Phase II CO2 emissions have no impact on firms’ financial performance. This is possibly due to the excessive emission allowance allocation. This result confirms the previous research. The conclusion is that policies implemented by the government should give corporations incentives to reduce CO2 emissions. Future research can identify the types of firms that tend to reduce CO2 emissions by testing whether the firms’ financial performance will have influence on their CO2 emissions.
Keywords: CO2 emissions, the EU ETS, corporate financial performance.
JEL classification: G30, Q56, Q58
2
1
Introduction
With the signing of the Kyoto Protocol, nations committed to work together to reduce greenhouse gas emissions in response to climate change. The consequences of climate change are a threat to human health. Manmade CO2 emissions contribute significantly to climate change. To limit CO2 emissions, nations impose various policies and regulations on their industries. For example, the European Union enforces the EU ETS (introduced on January 1st, 2005). The approach they use is a “cap-and-trade” mechanism, so companies can cut emissions in a cost-effective way (European Union, 2013). Australia has its own CO2 emission trading scheme and the trading regulation has been changed multiple times. The US has a state and a local policy. The US Climate Change Science Program (introduced in June 2008) is a climate change investigation program in conjunction with cabinet departments and federal agencies. But it is argued that states have limited resources to contribute to climate issues and no power to comply with an international agreement. Thus, companies have little incentive to reduce their emissions. Despite being an advanced economy, Canada withdrew its participation from the Kyoto Protocol in 2011 to avoid heavy penalties for not achieving the emission reduction goal. Early in 2007, the Prime Minister of Canada, Stephen Joseph Harper, opposed the fine on the binding targets and suggested that Canada would not comply unless the targets also apply to China and India (Flannery, 2013). Following the logic of Harper, it seems that enforcing the reduction of carbon emissions in a country has a negative effect on the total production in that country. This, in turn, will reduce the competitiveness of that country compared to others. Is this really true? In this respect, it is of interest to examine how changes to companies’ CO2 emissions impacts their financial outcomes.
3
benefit from the trading system in two ways. First, companies have opportunities to gain additional incomes. The idea behind the EU ETS is to create a reward system. Firms that reduce CO2 emissions can sell their unused EUAs to other firms. Second, there is an indirect benefit. Companies disclose the reduction of CO2 emissions in their annual reports as an environmental contribution (Gallego-Alvarez, Segura, & Martinez-Ferrero, 2014). Companies can improve their reputations with shareholders and the government. The condition of this policy is that EUAs retain value. A surplus of EUAs in the market will not motivate firms to take actions to decrease their CO2 emissions. This research provides an empirical study of the degree of effectiveness regarding the EU ETS development phases.
4
The remainder of this paper is organized as follows: In the Section 2, I provide a brief review of the literature and present the hypotheses. Section 3 describes the data. The research model and the methodology are formulated in Section 4. Section 5 demonstrates the results. Finally, Section 6 discusses the conclusion and identifies limitations and recommendations for the future research.
2
Literature review
This section reviews recent studies that relate to carbon emission and financial performance topic and, based on this review, a hypothesis is developed. There is an extensive body of literature that studies the relationship between CO2 emissions and firms’ behaviours. Martin, Muuls, & Wagner (2016) focused on the EU carbon market. They studied the impact of carbon trading prices on firms’ carbon emission management and financial outcomes under the EU ETS in the last ten years. Gallego-Alvarez, Segura, & Martinez-Ferrero (2014) conducted a global inquiry into the impact of various levels of carbon emissions on corporate operational and financial performance. Several other researchers have studied the link between firms’ carbon emissions and financial performance. Some significant studies are those of Busch & Hoffmann (2011); Hart & Ahuja (1996); Wang, Li, & Gao (2014); Oestreich & Tsiakas (2015); and Xu, Zeng, & Tam (2012). Research mainly differs by methodology, data period and geographic scope. This research follows the approach of Gallego-Alvarez, Segura, & Martinez-Ferrero (2014), who study the variation of carbon emission and firms’ financial and operational performance. I extend their work by focusing on firms’financial performance, including three more financial indicators and concentrating on the period of the EU ETS. Similar to Gallego-Alvarez, Segura, & Martinez-Ferrero (2014), most of the aformentioned researchers only use one indicator for corporate financial performance. I consider this to be insufficient to determine a firm’s overall financial performance. Waddock & Graves (1997) suggest that firms’ financial performance has various dimensions. Thus, I include four financial indicators to illustrate different dimensions of firms’ financial performance. I also use up to date data, investigating a period of eleven years (from 2005 to 2015).
5
2.1 Theoretical perspectives
The existence of a carbon trading mechanism strengthens the relation between carbon emissions and financial performance. A firm’s value is automatically linked with its CO2 emission by putting a price on the emission permit. It cannot be neglected that a firm’s value is possibly dependent on its CO2 emissions. According to the European Union (2013) report, the EU ETS is “the world’s biggest emissions trading market, accounting for over three quarters of international carbon trading”. Besides the EU, many other geographic areas have their own CO2 emission policies. For example, the carbon pricing scheme in Australia, the clean power plan in the US, the climate action tracker in Japan and the emission trading scheme in South Korea. With global trading, markets are closely connected with each other. The CO2 emission trading policy in one market can impact industrial production in another. Therefore, I take the sample companies from different geographic areas, namely the EU, US, Canada, and Australia. There are limited studies regarding CO2 emissions and corporate financial performance conducted in an international scope. Gallego-Alvarez, Segura, & Martinez-Ferrero (2014) include 21 countries worldwide but they expressed that uncontrollable factors (such as culture differences and economic development) may impair the credibility of their results. To mitigate the influence of culture and economy factors, I choose culturally close and economically advanced countries. Several scholars provide their perspectives on how the EU ETS, by pricing CO2 emissions permits, influences firm’s value. Scholtens & van der Groot (2014) demonstrate a significant impact of the EU ETS on the value of firms. However, their conclusion can only apply to the firms whose productions relied heavily on CO2 emissions. I will include all industries to test whether this conclusion still holds. The reason for including companies from all industries is the intertemporal relationship among companies in different industries. A study from Boersen & Scholtens (2014) shows the impact of EUAs future prices in the European electricity market. The electricity market is one of the fundamental commodity markets. The change of the electricity price directly influences the cost of production in many other sectors. Since the price of EUAs has a significant effect on the electricity market, many other industries will also be affected by the price of EUAs. Therefore, even companies that do not belong to the EU ETS are influenced by it. This research, unlike many EU ETS studies, includes companies that belong to the EU ETS-regulated industries and non-regulated industries.
6
over ten years. In Phase I, due to the excessive amount of EUAs allocated to companies, the EU ETS was not effective. This implies that a firm’s value is not linked to its CO2 emission management. In Phase II, Martin, Muuls, & Wagner (2016) conclude that the EU ETS was still ineffective. The reason for this is that an economic recession happened to occur in the same period, and production largely decreased. Cooper (2010) also found the emission decline in the same time together with a production decline. In Phase III (from 2013 to 2021, I take data up to 2015), the trading mechanism shifts from free EUA allocation to auctioning. The changing of the mechanism possibly encourages firms to reduce their CO2 emission and sell their EUAs to the highest bidder. The effect of this phase has not been widely tested. This research will re-examine the conclusion of Martin, Muuls & Wanger (2016) on Phase I and Phase II, and test the CO2 emission and financial performance link in Phase III.
Further, several researchers provide a resources based view (RBV) and stakeholder theory to explain the association between CO2 emissions and firms’ financial performances. According to the RBV, the effective and efficient management of carbon emissions can be classified as a resource or capability of the firm. Resources and capabilities are defined by being valuable, rare, inimitable, and non-substitutable. The “development of low-carbon technologies will make it cheaper to reduce carbon emission”, says Martin, Muuls & Wagner (2016). Firms can build up competitive advantage by investing in low-carbon technology, which in return provides economic benefits. In stakeholder theory, companies should act in stakeholders’ interests to become successful. Stakeholders include customers, suppliers, employees and many other communities. By reducing CO2 emissions, firms can gain social recognition. The benefit of having brand recognition is that firms have the opportunity to expand their market share and charge premiums on their products (Luo, Lan, & Tang, 2012).
7
2.2 Results from empirical evidences
The arguments around the topic of the relationship between CO2 emissions and financial performance point in conflicting directions. Some researchers find positive relationship and others find no relationship. However, most of them show a negative association between emissions and financial performance.
A group of researchers (Veith, Werner, & Zimmermann, 2009; Wang, Li, & Gao, 2014; Xu, Zeng, & Tam, 2012) pose the empirical results with the positive relationship, indicating that the more emissions a company produces, the more financial benefits the company will obtain. Xu, Zeng & Tam (2012) use a regression model to study manufacturing industries in Australia. The increasing market demand leads to increased production that results in an increased amount of emissions. Simultaneously, the firms’ profits increase. Using entity fixed effect, Wally & Whithead (1994) suggest that firm’s poor financial performance is due to management drawing resources away to invest in emission-reducing technology. Oestreich and Tsiakas (2015) use a T-GARCH (1.1) model, investing the carbon emission and stock return under the EU ETS. They suggest that, during Phases I and II, firms’ emissions outputs are positively correlated with their financial performance. There are two reasons for this. First, the free allocation of carbon emission allowances brings free cash flow. Veith, Werner, & Zimmermann (2009) also use T-GARCH (1.1) model and confirm the positive relationship in Phase I. Second, Oestreich & Tsiakas (2015) state that “high carbon emissions have higher exposure to carbon risk and exhibit higher expected returns”. However, the study of Oestreich & Tsiakas (2015) is biased. The conclusion they draw is based on the price of EUAs. EUA prices cannot represent the amount of CO2 emissions. This research will study the exact relationship between carbon emissions and firms’ performance.
8
They argue that the price on carbon emissions should be closely related to the oil price because the oil price usually determines the cost of productions and cost of goods. Since oil price is currently low, the level of emissions will not present any relationship with financial performance.
Most researchers managed to ground an empirical result demonstrating a negative relationship between carbon emissions and firms’ financial performance. Gallego-Alvarez, Segura, & Martinez-Ferrero (2014) use time fixed effect to study 89 companies from 21 countries in the period of 2007 to 2014. They found that the reduction of carbon emissions leads to the increase of firms’ financial performance. Luo, Lan & Tang (2012) use event study method to investigate firms’ stock returns with the disclosure of CO2 emissions in China. Stock prices drop when the disclosed CO2 is high. They conclude that social pressure plays an important role whereas the size of the company is a determinant variable. Luo, Lan & Tang (2012) suggest that reputation cost will be more significant for large companies compared to medium and small companies. Moreover, for big companies with good reputations, it is easier to transfer the cost of investing in the environment to the end customers. In other words, the investment that a company spends on building a positive image among stakeholders in the society, stakeholders are willing to pay a premium for this company’s final products. Therefore decreasing CO2 emissions negatively impacts firms’ financial performance.
The EU ETS was established to respond to climate change. The EU ETS, by providing a platform for emission allowance trading, links firms’ carbon emissions with their financial performance. The literature does not offer a unanimous result regarding this relationship due to the usage of different samples across different time periods. From the literature, the degree of effectiveness of the EU ETS varies in its three phases. Many researchers use RBV and stakeholder theory to support the relationship between CO2 emissions and firms’ financial performance. I will also use these two theories to develop hypotheses.
2.3 Hypothesis development
9
(1998) investigate firms’ profitability and their environmental investments. They find that with better environmental performance, firms are more profitable. Combining with stakeholder theory, Gallego-Alvarez, Segura, & Martinez-Ferrero (2014) suggest that stakeholders react positively to firms’ ethical behaviour. Luo, Lan, & Tang (2012) conduct research in the Chinese market regarding the listed firms’ stock prices and their CO2 emission disclosure. They confirm the stakeholder view suggestsing that better CO2 management increases a firm’s share price. The establishment of the EU ETS further raises companies’ awareness of reducing CO2 emissions. Martin, Muuls, & Wagner (2016) present a graph showing that from 2005 there is a steady increase of new low-carbon technology patents. The carbon trading scheme also provides firm economic benefits to emit less CO2 and additional income can be earned by selling the allowances (European Union, 2013). In light of the theories and the EU ETS, the first hypothesis is:
H1: The relationship between carbon emission and corporate financial performance follows a
negative association.
This study is aware of the differences between the three phases of the EU ETS. According to Martin, Muuls, & Wagner (2016), the EU ETS Phase I and Phase II, firms’ CO2 emission behaviors are not affected by the trading scheme due to the over-allocation of EUAs to companies. This research is interested in how the carbon emssions associate with firms’ financial performance during these periods. Will firms emit more than usual? The fact that emission permits were allocated and have little value on the market (Anderson & Di Maria, 2010). Luo, Lan & Tang (2012) suggest firms’ environmental behaviours are related to their size. Although large companies are more likely to be exposed to media and news when they behave unethically, they will change only if the reputation costs exceed their production costs. This research uses a sample of listed companies, the majority of which qualify as “large corporate”. I suspect that firms take advantage of the excessive EUA allocation to emit more with little cost at the early stage of the EU ETS. Thus, this research poses the following hypothesis:
H2a: In Phases I and II of the EU ETS, carbon emissions and corporate financial performance
follow a positive association.
10
auctioning. Learning from Phases I and II by gradually declining the total amount of allowances available on the market, firms that invest in low-carbon technology can benefit from selling their allowance at the highest bidding price (European Union, 2013). Therefore, this study grants the following hypothesis:
H2b: In EU ETS Phase III, carbon emissions and corporate financial performance follow a
negative association
In comparison with the research of Oestreich and Tsiakas (2015), we are examining the exact relationship between the amount of carbon emissions and firms’ financial performance under the EU ETS. I include more financial indicators than firms’ stock return. Comparing Gallego-Alvarez, Segura, & Martinez-Ferrero (2014)’s research, I choose the sample from culturally close and economically advanced areas to mitigate unobservable effects such as culture and economic development when doing a cross-border study. Moreover, this research brings the data as close to the present as possible. In the next section, I explain the sample and the research model in detail.
3
Data and model
3.1 Data
In this section, I will provide descriptions of the sample. This includes the number of the companies, the criteria for selecting a company, time period of the data, and the sample distribution in the countries/regions. Furthermore, I will explain the formulation of variables (dependent, independent and control variables). This includes the sources of data collection, the processing and calculation of secondary data.
11
and developed countries. As for the sample, the companies are selected based on following criteria:
1. Companies are listed on the stock exchanges of the selected countries from 2005 to 2015;
2. Companies have CO2 emissions data and required accounting data available from the period 2005 to 2015;
3. Financial sectors are excluded.
Similar to Gallego-Alvarez, Segura, & Martinez-Ferrero (2014), with the aim of investigating the impact of CO2 emissions on financial performance, this study proposes the following relationship:
Financial Performance = 𝑓 (CO2 emission, control variable) (1)
I will illustrate each component of model (1) in detail in the coming parts and provide the expectations on how CO2 emissions, together with the control variables, affect firms’ financial performance.
3.1 Dependent variables
This study uses four financial indicators to assess a firm’s financial performances. They represent different aspects of a firm’s financial behaviour. Unlike many other researchers who draw conclusions from only one or two indicators (see Table 1), I consider this to be insufficient. The four financial indicators are firms’ excessive stock returns (ESR), return on equity (ROE), volatility of the stock price (Vp), and Tobin’s q.
Table 1 Financial Performance Proxies
Authors Corporate Financial Performance
Hamilton (1996) Excessive Stock Return
Busch and Hoffmann (2011) ROE and Tobin’s q
Luo, Lan and Tang (2012) Stock Return and Stock Price Volatility
Gallego-Alvarez, Segura, & Martinez-Ferrero (2014)
ROE
Oestreich and Tsiakas (2015) Stock Return and Stock Price Volatility
12
Besides the researchers present in the table, many other researchers such as Wang, Li, & Gao ( 2014) and Kumar, Managi, & Matsuda (2011) also used Stock Return as an indicator. Stock Return is the return on firms’ stock price. Stock price is determined by firms’ income, cash flow and any other value change. I use excessive stock return (ESR) for our research. It is scaled as:
ESRit = (Stock Price i,t – Stock Price i,t-1 ) / Stock Price t-1 – Country’s Risk Free Rate t (2)
Stock Price i,t: firm i average stock price in year t;
Stock Price i,t-1: firm i average stock price in year t-1;
Country’s Risk Free Rate t: the 10-year government bond interest rate in the country where the
firm locates in year t.
ESR is different than stock return. It is a measure that is widely used to indicate value added by firms’ performance. This measure is in line with our goal. This study is testing whether reducing carbon emissions will add value to firms’ performance. The annual stock price of an individual company is collected from Datastream. This digital database is also the resource for the data of the interest rate of the10-year government bond which I use as the risk free rate.
Firm’s ROE is the second financial indicator. Researchers such as Gallego-Alvarez, Segura, & Martinez-Ferrero (2014); Xu, Zeng, & Tam (2012); and King & Lenox (2001) explain that ROE is an indicator of firms’ profitability in the fiscal year. I retrieve firms’ ROE value directly from the digital database Datastream. I expect that reducing carbon emissions will have positive effects on firms’ ROE.
I also include the volatility of a firm’s stock price (Vp). Yearly Vp is the standard deviation of the daily stock price. I take this data directly from the Datastream database. Price volatility usually relates to the amount of uncertainty of the price changes. The Vp value indicates the amount of change to a firm’s stock price. It is appropriate to use this indicator to assess firms’ performance. Luo, Lan & Tang (2012) suggest that every time a firm is required to disclose its CO2 pollution index, their stock price drops. I expect the spread of stock price is greater when the amount of emissions is higher.
13
value should indicate a firm’s growth potential. Many firms can largely reduce carbon emission output results via innovative technology or optimized production method (Martin, Muuls & Wagner, 2016). The new invention and optimization will naturally influence the market value of the company. Therefore, I expect that, by decreasing carbon emissions, firms will increase their Tobin’s q value. I use Datastream to obtain the data of firms’ annual total market value and firm’s annual total book value.
3.2 Independent variable
I scale the independent variable, CO2 emissions, as a ratio and calculate it as follows:
CO2 emission ratioit = (Direct CO2 emissionit + Indirect CO2 emissionit)/Total salesit (3)
Direct CO2 emissionit:: firm i its direct CO2 emission in tons in year t, this value is taken directly from the Datastream database.
Indirect CO2 emissionit:: firm i its indirect CO2 emission in tons in year t, this value is taken directly from Datastream database.
Total salesit:: The amount of sales that firm i in year t has achieved in the currency where the
firm is located.
Ameer & Othman (2011), Busch & Hoffmann (2011), Trumpp & Guenther (2015) and Chapple, Clarkson & Gold (2013) all use this ratio as a proxy for CO2 emissions. This ratio takes firms’ characteristics into consideration. This study is using a sample that measures whether, across multiple industries, the amount of CO2 emissions and total sales are closely related. Thus scaling CO2 emissions is appropriate. Datastream is the database I use to collect each firm’s direct and indirect CO2 emissions and the firm’s total sales.
3.3 Control variables
14
activities. For example. Malina et al.,(2012) suggest that the EU ETS has had an impact on the US aviation industry. As for many other industries, since markets are globally connected, including market return of EUAs is appropriate. The market return of EUAs is calculated as follows:
Market return of EUA’s future pricet = (Future price of EUAt – Future price of EUAt-1)/ Future price of EUAt-1
(4)
Future price of EUAt: the average price of EUAin year t;
Future price of EUAt-1: the average price of EUA in year t-1.
Market return of EUA’s future price is a market level data. It changes over time. I received the EUA price by requesting data from EEX (www.eex.com). The data are daily EUA future prices. I first annualise the data and then calculate the yearly return. This study expects the higher the EUAs market return, the better the firm’s financial performance will be.
This research also controls for firm size. Following Gallego-Alvarez, Segura, & Martinez-Ferrero (2014), I use a natural log of the firm’s total assets. They find the larger the firm size, the less likely it is to reduce its carbon emissions. This paper expects that the larger the firm, the better its financial performance will be. The total asset data can be diretly obtained from Datastream.
Busch & Hoffmann (2011) determined that when a company is in financial distress, carbon emission management will not be their management’s first priority. Financial distress will affect a firm’s productivity, and the reduction in production will directly reduce the amount of carbon emissions. Waddock & Graves (1997)’s research also suggests that risk tolerance will influence the attitude of the whole company towards pollution, waste reduction and investment in emissions control. The Altman Z-Score is adopted to measure the degree to which a firm is in financial distress. Altman Z-Score is an aggregated value that predicts the probability of a firm to go into bankruptcy and corporate distress status (Altman, 1968). I use the following method that proposed by Altman (1968) to calculate this value:
Altman Z-Scoreit = 1.2 X1 + 1.4 X2+ 3.3 X3 + 0.6 X4 + 0.999 X5 (5)
X1 = Working capitalit / Total assetsit;
X2 = Retained earningsit / Total assetsit;
15
X4 = Market value of equityit / Total liabilitiesit ;
X5 = Total salesit / Total assetsit. Where i is the firm; t is the year.
The accounting data (e.g., working capital, retained earnings) can be directly derived from Datastream. The larger the Altman Z-Score, the safer the company will be. Thus, I expect the higher the Z-Score, the better the financial performance.
I divide the industries into two groups according to whether the company belongs to a sector that is eligible for the EU ETS or not. Researchers such as Anderson & Di Maria (2010); Chapple, Clarkson, & Gold (2013); and Scholtens & van der Groot (2014) take their samples from the industries that belong to the EU ETS with significant result. Although this research studies all companies across all sectors, I am expecting the same conclusion. I use a dummy variable to separate the EU ETS company eligibility. A company that is eligible for the EU ETS is marked as 1, a non-eligible company is marked as 0. See Appendix B for the sample distribution in industries.
In the following part, this paper discusses concerns regarding the selected variables, and provide descriptive statistics of the data.
3.4 Potential shortcomings of the variables
16
is not presented in determining the control variables, and the same issue as Gallego-Alvarez, Segura, & Martinez-Ferrero (2014). The US, Australia, Canada and many European countries within the EU had its own environmental trading schemes. This will influence the accuracy of the statistic result of this research.
3.5 Descriptive statistics
First, this part provides a descriptive analysis of independent variables in Table 2. The descriptive analysis of dependent variable will be presented in Appendix C.
It can be seen that CO2 emission ratio in the EU area are relatively lower than US, Canada and Australia. The EU area is known as an area with low CO2 emission, as The World Bank shows that the EU area from 2006 to 2011 has average 5 CO2 emission metric tons per capita. This number is low in comparison with around 15 average CO2 emission metric tons per capita in the US, Canada and Australia (CO2 emission metric tons per capita can be found on The World Bank website: http://data.worldbank.org/). Furthermore, focus on the EU market, the EU ETS Table 2 Descriptive analysis of independent variable
Variable Mean Standard deviation Median Kurtosis Skewness
EU CO2 Emission Ratio 12.501% 18.853% 11.466% 0.305 1.712
US CO2 Emission Ratio 36.831% 38.440% 21.653% 0.901 0.979
Australia CO2 Emission Ratio 24.612% 21.113% 23.437% 0.326 0.198
Canada CO2 Emission Ratio 28.265% 35.192% 27.814% 0.144 1.176
Market Return of EUAs Futures -4.808% 25.824% 2.842% -0.183 -0.152
Phase I EU CO2 Emission Ratio 14.290% 23.814% 14.744% 0.266 1.606
Phase II EU CO2 Emission Ratio 17.316% 21.726% 16.037% 0.241 1.475
Phase III EU CO2 Emission Ratio 12.133% 19.101% 12.303% 0.224 1.454
EU Firm Size 10.558 4.704 11.445 -1.476 0.080
U.S. Firm Size 16.112 2.682 16.129 1.834 -0.194
Canada Firm Size 13.054 6.742 13.198 1.576 -0.093
Australia Firm Size 14.774 1.867 14.866 1.305 0.410
EU Altman Z-Score 2.380 4.022 1.284 2.981 0.955
U.S. Altman Z-Score 2.323 5.322 2.348 11.621 -1.307
Canada Altman Z-Score 3.521 4.542 3.397 14.871 4.568
17
three phases show a clear decrease of carbon emission ratio. I will present whether this decrease is significant in the result section, and test whether the carbon emission influences firm’s financial performances in each phase.
Second, I present the data for whether the company is belonging to an industry that is eligible for being regulated under the EU ETS. The eligible industries are presented in the Appendix B. The distribution of the eligible and non-eligible companies are relatively equal.
Table 3 Number of companies eligible for the EU ETS Dummy variable Eligibility Covered regions
EU Canada AUS US
Eligible 1 289 44 60 81
Ineligible 0 291 55 31 73
Third, this research tests the correlation coefficient between independent variables. In the Data section, I explain the concern of the potential shortcomings of these variables. Table 4 presents the correlation results. The market return on EUA, Altman Z-Score and eligibility are significantly correlated with variable emissions. However, the strength of the correlation by absolute value is weak (all < 60%). Thus, I cannot draw the conclusion that this correlation will cause multicollinearity.
Table 4 Correlation matrix of explanatory variables Correlation matrix of explanatory variables
Emission EUA Size z Eligibility
Emission 1
EUA 0.044*** 1
Size 0.003 0.003 1
Z-Score 0.047*** -0.008 0.134*** 1
eligibility 0.156*** 0.034** -0.022 -0.009 1
***. Correlation is significant at 0.01 level (2-tailed) ** Correlation is significant at the 0.05 level (2-tailed) * Correlation is significant at the 0.10 level (2-tailed)
Emission is CO2 the emission ratio, measured as the firm’s total emissions divided by its total sales; EUA is the
18
In the next section, this study demonstrates the analysis technique that will be used to test the hypotheses.
4
Methodology
To test the hypotheses, I build several regression models to estimate the panel data that I have collected. Panel data enable researchers to evaluate a company’s performance over a period of time. I follow the research approach of Gallego-Alvarez, Segura, Martinez-Ferrero (2014). They also use panel data, and suggest that panel data provide a greater degree of freedom and allow for missing values (it is called unbalanced panel data). Some researchers indicate that panel data are effective for specifications but that, since in the estimation procedure the unobservable heterogeneity in the error term is eliminated, they are potentially unreliable for prediction (Pindado and Requejo, 2012). Lee (2007) suggests that whether the result can be interpreted correctly depends on the proper specification of the effect that the researchers apply. The first hypothesis regards whether there is a negative relationship between carbon emission ratios and firms’ financial performances. This research will investigate this hypothesis in four different geographic areas. I will first run a Two-Sample t-test between the EU CO2 emission sample and the three other countries. I expect there to be a significant difference between the carbon emissions of the EU area on one hand, and the US, Canada, and Australia on the other. Subsequently, I will use a pooled OLS as a specification to run the proposed model and test the relationship in each region. Gallego-Alvarez, Segura, and Martinez-Ferrero (2014) first conduct a Hausman specification test to determine their model; a fixed-effect model is appropriate for their data. Eviews is the software that this study used to process data. This research models include data that only change according to years. The market return of the future price of EUAs varies across time series but not entities. Therefore, I would have difficulty running the fixed-effect model in Eviews, as the software would show a near singularity error. For this reason, I apply a pooled OLS. I will first present the research models below, and then explain the concerns regarding the pooled OLS method.
4.1 Empirical model
19
ROEit= β0 + β1 CO2 Emissionit + β2 EUAt + β3 Sizeit + β4 Z-Scoreit + β5 Eligibilityi + ε (6) ESRit= β0 + β1 CO2 Emissionit + β2 EUAit + β3 Sizeit + β4 Z-Scorei + β5 Eligibilityi + ε (7) Vpit= β0 + β1 CO2 Emissionit + β2 EUAit + β3 Sizeit + β4 Z-Scorei + β5 Eligibilityi + ε (8) qit= β0 + β1 CO2 Emissionit +β2 EUAit + β3 Sizeit + β4 Z-Scorei + β5 Eligibilityi + ε (9) ROE is Return on Equity; ESR is Excessive Stock Return; Vp is the volatility of the stock return; q refers to
Tobin’s q; Emission is the CO2 emission ratio, measured as the firm’s total emissions divided by its total sales;
EUA is the market return on the EUA future price; Size is the natural log value of a firm’s total assets; Z-Score is
the Altman Z-Score that is calculated in formula (5); Eligibility is a dummy variable – if the company is operating in an industry that is eligible for the EU ETS, then this is 1, while if the company is operating in an industry that
is not eligible for the EU ETS, then this is 0; ε is the error term of the function; i is the firm; t is the year.
This model will be applied to all of the selected areas (EU with/without the UK, the UK, the US, Canada, and Australia) individually to generate an overall view of the carbon and corporate financial relationship. I specifically exclude the UK from the EU sample because half of the EU data I have collected are from the UK (277 companies). The variable EUA will be included to test this relationship in non-EU areas because, as explained before, the market is globally connected, and the price settled in the EU will be channeled to other markets.
The following models are used to test hypothesis 2. In hypothesis 2a, I expect that from 2005 to 2007 and 2008 to 2012 (Phase I and Phase II of the EU ETS), the increase in CO2 emissions positively affected a firm’s financial performance. This is because the excessive allocation of EUAs induces firms to take advantage of creating more emissions, and therefore firms can use cheaper energy or increase the amount of production to receive their revenue. In models (10), (11), and (12), the left-hand side of the equation represents the financial indicators, and the right-hand side represents the independent and control variables. For each period of time, I will test whether the carbon emission ratio has a positive or negative effect on each individual financial indicator.
(ROE, ESR, Vp and q)2005-2007, t = β0 + β1 Emissionit + β2 EUAt + β3 Sizeit + β4 Z-Scoreit + β5 Eligibilityi + ε (10) (ROE, ESR, Vp and q)2008-2012, t = β0 + β1 Emissionit + β2 EUAt + β3 Sizeit + β4 Z-Scoreit + β5 Eligibilityi + ε (11) (ROE, ESR, Vp and q)2013-2015, t = β0 + β1 Emissionit + β2 EUAt + β3 Sizeit + β4 Z-Scoreit + β5 Eligibilityi + ε (12)
ROE is Return on Equity; ESR is Excessive Stock Return; Vp is the volatility of the stock return; q refers to
Tobin’s q; Emission is CO2 the emission ratio, measured as the firm’s total emissions divided by its total sales;
EUA is the market return on the EUA future price; Size is the natural log value of a firm’s total assets; Z-Score is
20
in an industry that is eligible for the EU ETS, then this is 1, while if the company is operating in an industry that is not eligible for the EU ETS, then this is 0; ε is the error term of the function; i is the firm; t is the year.
The concern with using pooled OLS is that the unobservable fixed effects cannot be captured. As previously explained, when using Eviews to implement the model, I receive a warning indicating a singular matrix error. This is because the control variable market return of EUAs future is a market level variable, and each year only has one value. I cannot exclude this value, as it is crucial to control the impact of EUAs on firms’ financial performance.
In the next section, I will demonstrate the results and provide interpretations in response to hypotheses. In addition, I will compare these results with those of other researchers’ to further explain the economic interpretation and contribution of this research.
5
Results
Using the pooled OLS analysis techniques, the first hypothesis regards whether the amount of CO2 emissions had an impact on a firm’s financial performance in the EU area, the US, Canada, and Australia from 2005 to 2015. This study expects a negative relationship: with fewer carbon emissions, firms’ ROE, ESR, and Tobin’s q will generate a higher value and a smaller Vp. Firms tend to invest in efficient, low-carbon technologies that bring them profitability.
I first run a Two-Sample t-test. The aim is to use the EU carbon emission ratio as a benchmark to test whether the emissions from the US, Canada, and Australia are significantly different from those of the EU area. In the Appendix B, I present the Two-Sample t-test result. There are significant differences between the US, Canada, and Australia’s CO2 emissions in comparison with those of the EU. Table 6 presents the result of the first hypothesis using the pooled OLS analysis technique with the EU sample.
Table 6 Results obtained by applying pooled OLS to the EU sample
Independent variables Dependent variables
ROE (EU) ESR (EU) Vp (EU) q (EU)
21
***. Correlation is significant at the 0.01 level ** Correlation is significant at the 0.05 level * Correlation is significant at the 0.10 level
ROE is Return on Equity; ESR is Excessive Stock Return; Vp is the volatility of the stock return; q refers to Tobin’s
q; Emission is the CO2 emission ratio, measured as the firm’s total emissions divided by its total sales; EUA is the
market return on the EUA future price; Size is the natural log value of a firm’s total assets; Z-Score is the Altman Z-Score that is calculated as formula (5); Eligibility is a dummy variable – if the company is operating in an industry that is eligible for the EU ETS, this is 1, while if the company is operating in an industry that is not eligible for the EU ETS, this is 0; i is the firm; t is the year.
Table 6 indicates that there is no significant relationship between carbon emission ratios and any of financial performance indicators. The economic interpretation is that the decrease in the emissions is due to the general movement (for example, new technology available in the market, new energy resources replacing conventional ones). Financial performance is more dependent on other factors than they are on carbon emissions in the EU. In response to the first hypothesis, I cannot reject the null hypothesis that there is no relationship between the carbon emissions and corporate financial performance in the EU area. This is in line with the work of King and Lenox (2001), who suggest that carbon reduction is a “fixed characteristic” for the companies whose business strategy includes environmental management. Only if firms intend to reduce their carbon emissions, every metric ton of the reduction may reflect on their financial performance. In this research, I use a sample of companies from all different sectors. Based on this statistic result I can conclude that for many sectors, carbon management is not a business priority. This explains why many researchers only focus on carbon output-intensive companies as their sample and can find a significant relationship between carbon reduction and financial performances: because carbon management is most relevant for carbon emission-intensive companies. However, for the health of the environment, every company is expected to commit itself to reducing its carbon emissions.
Another possible reason is that over half of the EU data are from the UK; therefore, the UK companies’ behavior can largely impact the rest of the EU countries. This may result in the result for the EU not being representative. Therefore, I will apply the same method on UK companies alone and on the rest of the EU companies. The result of Two-Sample t-test for the
Observations 576 577 577 577
R2 0.212 0.011 0.307 0.473
22
UK and the rest of the EU companies’ CO2 emissions can be found in the Appendix C. The purpose is to see if the CO2 emissions level of the UK company samples are different from EU companies’. The statistic result show a significant difference at 90% confidence level.
The UK companies are now separated from the total EU sample. I apply the same regression model, and organize the result in Table 7.
Table 7 Results obtained by applying pooled OLS to the UK sample
Independent variables Dependent variables
ROE (UK) ESR (UK) Vp (UK) q (UK)
β0 0.064(3.95)*** 0.063(1.21) 24.682(16.66)*** 0.520(3.75)*** β1 (Emission) 0.008(1.76)* -0.078(-0.66) 0.004(0.53) 0.002(-0.31) β2 (EUA) 0.010(1.33) -0.033(-1.06) -2.547(-3.37)*** 0.114(1.50) β3 (Size) 0.004(0.19) 0.002(-0.11) 0.209(2.14)** 0.007(1.20)*** β 4 (Z-Score) 0.011(5.57)*** 0.001(1.24) 0.181(2.10)** 0.046(7.02)*** β5 (Eligibility) 0.001(-1.02) -0.003(-0.15) 0.053(0.12) -0.101(-1.43) Observations 276 276 276 276 R2 0.175 0.132 0.104 0.261 F-statistic 8.873 0.657 4.552 12.513
***. Correlation is significant at the 0.01 level ** Correlation is significant at the 0.05 level * Correlation is significant at the 0.10 level
ROE is Return on Equity; ESR is Excessive Stock Return; Vp is the volatility of the stock return; q refers to
Tobin’s q; Emission is the CO2 emission ratio, measured as the firm’s total emissions divided by its total sales;
EUA is the market return on the EUA future price; Size is the natural log value of a firm’s total assets; Z-Score is
the Altman Z-Score that is calculated in formula (5); Eligibility is a dummy variable – if the company is operating in an industry that is eligible for the EU ETS, this is 1, while if the company is operating in an industry that is not eligible for the EU ETS, this is 0.
23
on the website of the World Bank (data.worldbank.org/indicator/EN.ATM.CO2E.PC). UK CO2 emissions are above the average EU CO2 emissions per capita. The associations between carbon emissions and other dependent variables are not significant. The conclusion is that I cannot reject the null hypothesis. But this study can conclude that there is a positive relation between CO2 emissions and a firm’s ROE for UK sample.
I separate UK companies from the EU sample, and apply the research model to the rest of the EU sample. The results are presented in Table 8.
Table 8 Results obtained by applying pooled OLS to the EU sample excluding the UK
Independent variable Dependent variables
ROE (EU exl. UK) ESR(EUexl.UK) Vp (EU exl. UK) q (EU exl.UK)
β0 0.207(26.74)*** -0.052(3.24)*** 25.001(23.76)*** 2.024(26.16)*** β1 (Emission) -0.055(1.84)* 0.007(0.23) 0.004(-0.31) 0.002(-1.78)* β2 (EUA) 0.004(1.08) 0.016(2.37)** -0.164(-0.21) 0.065(1.27) β3 (Size) -0.031(-20.78)*** 0.002(-1.51) 0.303(10.16) -0.281(-22.07)*** β 4 (Z-Score) 0.001(2.78)*** 0.004(-1.13) 0.028(-1.11) 0.044(2.67)*** β5 (Eligibility) 0.003(-0.07) 0.008(-1.45) 01.291(3.10)** -0.013(-0.39) Observations 301 301 301 301 R2 0.181 0.014 0.013 0.201 F-statistic 87.231 6.778 2.219 98.227
***. Correlation is significant at the 0.01 level ** Correlation is significant at the 0.05 level * Correlation is significant at the 0.10 level
ROE is Return on Equity; ESR is Excessive Stock Return; Vp is the volatility of the stock return; q refers to
Tobin’s q; Emission is the CO2 emission ratio, measured as the firm’s total emissions divided by its total sales;
EUA is the market return on the EUA future price; Size is the natural log value of a firm’s total assets; Z-Score is
the Altman z-score that is calculated as formula (5); Eligibility is a dummy variable – if the company is operating in an industry that is eligible for the EU ETS, this is 1, while if the company is operating in an industry that is not eligible for the EU ETS, this is 0.
24
with Busch & Hoffmann (2011), who also find a significantly negative relationship between carbon emissions, and a firm’s ROE and Tobin’s q values. The difference is that Busch & Hoffmaan (2011) suggest an overall U-shaped relationship, in which only the companies with carbon awareness will generate a positive ROE and Tobin’s q value for every unit of emission reduction. This study partially rejects the null hypothesis, and concludes that CO2 emissions are negatively correlated with a firm’s ROE and Tobin’s q value in the EU (excluding the UK). Furthermore, I use the same method to test this relationship in the US, Canada, and Australia. The results are presented in the Appendix. For the US (see Appendix F), it is difficult to conclude the existence of the relationship between the amount of carbon emissions and the financial indicators. Still, the result suggests a solid relationship in terms of stock price volatility and carbon emissions (p-value <0.00). The interpretation is that for every unit increase of carbon ratio, there is an 8% rise in the stock price volatility. In the same model, the result also shows a significant negative relationship between the market return of EUA and Vp. It can be concluded that the outcomes for the other corporate financial indicators (ROE, ESR, and Tobin’s q) are probably due to the US’s lack of a centralized carbon emission policy. Instead, there is a decentralized way of managing emissions: each state independently decides whether to take measures to manage them. Christoph et al., (2015) confirmed this assumption by conducting a critical assessment of US climate policy. They pointed out the inefficiency of state-wide CO2 emission regulations. Therefore, there are no materialized effects of carbon emissions on firms’ performances. However, since Vp is a firm risk indicator, it does make sense to conclude that with the rising amount of CO2 emissions, firm will run more risks, which is reflected in the volatility of their stock prices.
This is a different situation for Canada (see Appendix G). Carbon emissions show a negative correlation with Canadian firms’ stock price volatility. This is due to the nature of the production sector: the increasing emissions indicate a higher amount of production (Schmalensee, Stoker, & Judson, 2001). Thus the rationale behind this negative significance is that fewer emissions of a firm signal a lower production, which implies a business downturn and suggests a higher risk of the firm in the market. The negative significant relationship of the variables Altman Z-Score and Vp supports this perception (smaller Z-Scores indicates a worse financial status).
25
variable is added. However this result cannot be used since the financial indicators are highly dependent on control variables. Interestingly, the results indicate that carbon emissions are significantly negatively correlated with firms’ Tobin’s q at a 90% confidence level. Tobin’s q is a proxy for firms’ potential growth; in this case, it indicates that the less emissions are generated, the higher the potential is for growth. This could potentially be explained by investments in low-carbon innovation with high efficiency, which brings firms expansion potential.
Summarizing the statistic results regarding the first hypothesis, the outcomes demonstrate that the link between carbon emissions and financial performance is weak. This conclusion is drawn based on the investigation of the relationship in different geographical areas. There are three reasons for this outcome. First, the benefit that can be gained from the drop in oil prices exceeds the penalty on carbon emissions. This is in line with Abrell, Ndoye & Zachmann (2011). The historically low oil price is extremely beneficial for European firms, as Europe is a large oil import region rather than a region where oil originates. Therefore, firms can gain from expanding production with the low costs, and can increase sales volume in order to remain price competitive. The second reason is the economic downturn worldwide. This is an endogenous factor that is linked with the firms’ performance. It means that firms have different priorities than putting effort into emission management in this unstable business environment. This has not only distorted the link between carbon emission abetment, but has also led to unreliable assessment of reducing CO2 emissions. To truly encourage firms to invest in low-carbon technology and to be environmentally friendly, regulations should be more strictly enforced. The restriction on carbon intensive-output firms is necessary; moreover, in response to the climate change, the ultimate goal of these regulations is achieving carbon abetment. Thus, policy makers and regulation enforcement should take into consideration the different types of businesses and should make use of their intertemporal relationship to facilitate an environmentally cautious business world.
26
negative relationship between carbon emissions and firms’ financial performance. This is because of the shift of the EU ETS trading mechanism from free allocation to auction. In the auction trading mechanism, firms are more motivated to reduce carbon and trade their allowances to the highest bidder. Two-Sample t-test is applied to test whether there is a significant decrease of CO2 emissions between each phase in the Appendix I. In the following part, I present the result of H2a in Tables 9 and 10.
Table 9 Results obtained by applying pooled OLS to the EU sample for the EU ETS Phase I Independent variable Dependent variable
ROE ESR Vp q β0 0.049(6.39)*** 0.043(3.85)*** 24.101(29.14)*** 0.242(3.16)*** β1 (Emission) 0.032(1.04) -0.191(-0.49) 0.003(0.04) 0.004(1.11) β2 (EUA) 0.004(0.25) 0.017(0.69) 1.085(0.53) 0.104(0.38) β3 (Size) 0.004(1.23) 0.003(-4.73)*** 0.200(3.08)*** 0.022(3.88)*** β 4 (Z-Score) 0.013(4.75)*** 0.004(0.27) 0.182(1.15) 0.103(5.23)*** β5 (Eligibility) 0.004(0.18) 0.003(-0.71) 0.041(0.06) -0.008(-0.16) Observations 440 440 440 438 R 0.032 0.028 0.010 0.053 F-statistic 5.751 5.774 2.646 10.807
***. Correlation is significant at the 0.01 level ** Correlation is significant at the 0.05 level * Correlation is significant at the 0.10 level
ROE is Return on Equity; ESR is Excessive Stock Return; Vp is the volatility of the stock return; q refers to
Tobin’s q; Emission is the CO2 emission ratio, measured as the firm’s total emissions divided by its total sales;
EUA is the market return on the EUA future price; Size is the natural log value of a firm’s total asset; Z-Score is
the Altman Z-Score that is calculated as formula (5); Eligibility is a dummy variable – if the company is operating in an industry that is eligible for the EU ETS, this is 1, while if the company is operating in an industry that is not eligible for the EU ETS, this is 0.
The results for Phase I do not show any significant relationship between carbon emissions and financial performance. In the following part, I will test this relationship in Phase II and provide an interpretation and response for hypothesis H2a.
Table 10 Results obtained by applying pooled OLS to the EU sample for the EU ETS Phase II Independent variable Dependent variable
27 β0 0.073(11.38)*** 0.072(8.45)*** 23.471(38.34)*** 0.433(7.63)*** β1 (Emission) 0.003(2.75)*** -0.083(0.19) 0.003(-0.20) 0.002(2.76)*** β2 (EUA) 0.044(3.36)*** 0.047(3.22)*** -0.454(-0.40) 0.404(3.87)*** β3 (Size) 0.001(-1.77)* -0.006(-7.38)*** 0.219(3.72)*** 0.003(0.48) β 4 (Z-Score) 0.013(5.49)*** 0.004(-1.96)** 0.131(0.94) 0.080(5.99)*** β5 (Eligibility) 0.001(0.48) 0.013(-0.01) 1.032(2.30)** 0.008(0.20) Observations 543 543 543 541 R2 0.033 0.054 0.021 0.047 F-statistic 10.504 14.966 6.160 16.894
***. Correlation is significant at the 0.01 level ** Correlation is significant at the 0.05 level * Correlation is significant at the 0.10 level
ROE is Return on Equity; ESR is Excessive Stock Return; Vp is the volatility of the stock return; q refers to
Tobin’s q; Emission is the CO2 emission ratio, measured as the firm’s total emissions divided by its total sales;
EUA is the market return on the EUA future price; Size is the natural log value of a firm’s total assets; Z-Score is
the Altman Z-Score that is calculated as formula (5); Eligibility is a dummy variable – if the company is operating in an industry that is eligible for the EU ETS, this is 1, while if the company is operating in an industry that is not eligible for the EU ETS, this is 0.
28
Table 11 provides the result regarding H2b, testing whether in Phase III the carbon emission ratio was negatively related to firms’ financial performance.
Table 11 Results obtained by applying pooled OLS on the EU for the EU ETS Phase III Independent variables Dependent variables
ROE ESR Vp q β0 0.04(7.63) 0.01(-0.29) 23.50(36.23)*** 0.17(3.23)*** β1 (Emission) -0.00(0.11) -0.06(-0.19) 0.01(0.61) 0.00(-0.50) β2 (EUA) 0.01(0.96) -0.17(-7.40)*** -0.46(-0.53) 0.03(0.46) β3 (Size) 0.00(4.43)*** 0.01(3.80)*** 0.02(3.69)*** 0.03(7.23)*** β 4 (Z-Score) 0.00(1.68)* 0.00(-0.45) 0.06(0.45) 0.03(2.48)** β5 (Eligibility) 0.00(-0.24) -0.01(-0.97) 0.53(0.98) 0.00(-0.02) Observations 539 539 539 539 R2 0.02 0.05 0.01 0.05 F-statistic 5.24 13.74 3.68 13.29
***. Correlation is significant at the 0.01 level ** Correlation is significant at the 0.05 level * Correlation is significant at the 0.10 level
ROE is Return on Equity; ESR is Excessive Stock Return; Vp is the volatility of the stock return; q refers to
Tobin’s q; Emission is the CO2 emission ratio, measured as the firm’s total emissions divided by its total sales;
EUA is the market return on the EUA future price; Size is the natural log value of a firm’s total assets; Z-Score is
the Altman Z-Score that is calculated in formula (5); Eligibility is a dummy variable – if the company is operating in an industry that is eligible for the EU ETS, this is 1, while if the company is operating in an industry that is not eligible for the EU ETS, this is 0.
29
Overall, I have seen the progression of the carbon trading method in the EU ETS, from free allowance allocation to auction, in which firms are required to purchase their allowances on the market for the amount they will use. The real environmental effect for all of the business sectors in the EU ETS Phase III are still unknown. This research cannot draw a simple, directive conclusion but as time goes on, I can expect a stronger negative relationship between carbon emissions and financial performance under the EU ETS for all business sectors.
6
Conclusion
This research analysed the relationship between carbon emissions and financial performance across the regions of the EU, the US, Canada, and Australia. Furthermore, this research investigated this relationship in the three EU ETS phases. This study investigated a total of 924 companies from the 2005 to 2015. I used different financial indicators as proxies for firms’ financial performance. Furthermore, I used CO2 emission ratios as the independent variable. The increase or decrease in the CO2 emission ratio reflects a firm’s CO2 emission management efficiency. Due to climate change, regional governments are implementing policies and regulations to restrain carbon emissions. CO2 emissions can affect a firm’s value if there is a price for carbon emission permits. Does climate change really affect our businesses? I cannot give a straightforward answer to this question. The answers are dependent on regions, periods, and the variables that measure a business.
30
performance. Furthermore, I could not find a significant relationship between carbon emisisons and financial performance in Phase III.
The results also contrast with many studies that have found a highly significant relationship between carbon emissions and financial performance. However, some of these studies, including Luo, Lan & Tang (2012), Trumpp & Guenther (2015), and Wang, Li & Gao, (2014), only investigated high-intensity carbon emission companies, for whose business carbon management matters the most. This study extended the samples to include many other sectors, and found a weak relationship. Our results indicate that, to reduce overall emissions, every company should improve its CO2 emission awareness and contribute to the reduction.
The limitations cannot be neglected. I will discuss three major limitations of this study in the following paragraph, and I expect that this will also raise the attention of future researchers when conducting similar studies.
First, regarding the significant results, there is one fundamental concern that this research cannot address and that is out of its scope of control: the reliability of the emissions value. The issue does not regard the accountability of the data source. Instead, the worry is that, due to the increasingly strict emission regulations as well as the high environmental standards developing in the advanced economy, the companies that are unable to cope with these standards move away from these regions and establish themselves in places that have lax environmental rules (Dowell, Hart, & Yeung, 2000), such as emerging and less developed regions, which continue to pollute and to destroy global environmental condition. In addition, large companies can reduce their amount of reported emissions by outsourcing (or establishing a new entity to work with) the carbon-intense part of the business, suggested by Wally & Whithead (1994), such as manufacturing, to other locations that are not under the control of any ETS. Thus, even though these firms seem to have fewer emissions in their records, the fewer emission cannot comply with the better financial performance. These concerns and questions require further and deeper research. To provide empirical answers may be highly time consuming, since researchers should investigate the whole supply and distribution chain of each firm. On the other hand, the results may provide a better insight to the policy makers whose only concern is the environmental issue in their own region.
31
intentions behind disclosing their carbon performance. It is believed that for some companies it is only a routine to publish their data, whereas for many other companies, it is a part of their strategic plan. Thus, there was a bias in the sample selection procedure in this study. The consequence is that the significant results that were obtained in the statistics test may lose their representative meaning for the population.
The third limitation is the endogeneity of the crisis period. This limitation has been mentioned before. It is impossible to separate the effect of the economic downturn on business profitability as a whole worldwide, which indirectly distorted the relationship between firms’ financial performance and carbon emission management.
All the limitations discussed above indicate the need for further research. To further investigate the relationship between carbon emissions and firms’ financial performance, researchers could test the reversed effect: whether a firm’s financial performance impacts its carbon management. Then I could identify the characteristics of firms that impact their CO2 emissions most.
In all, our research has provided an overview of the relationship between CO2 emissions and firms’ financial performance across regions of the EU, the US, Canada, and Australia. It also investigates this relationship in the three phases of the EU ETS. I used a large sample, including a large variety of sectors, investigated four different financial indicators, and brought our data up to date. In future research, I will investigate the reverse relationship between carbon emissions and financial performance to identify which types of firms tend to manage their CO2 emissions.
References
Abrell, J., Ndoye, A. & Zachmann, G., 2011. Assessing the impact of the EU ETS using firm
level data.Working paper. [Online]
Available at: http://bruegel.org/2011/07/assessing-the-impact-of-the-eu-ets-using-firm-level-data/
[Accessed 6 May 2016].
Albertini, E., 2013. Does environmental management improve financial performance? A meta-analytical review. Organization and Environment, vol 26, pp. 431-457.
Altman, E. I., 1968. Financial ratios, discriminat analysis and the prediction of corporate bankruptcy. Journal of finance, vol 23,pp. 589-609.
32
Anderson, B. & Di Maria, C., 2010. Abatement and Allocation in the Pilot Phase of the EU ETS. Environmental resources economics, vol 48, pp. 83-103.
Boersen, A. & Scholtens, B., 2014. The relationship between European electricity markets and emission allowance futures prices in phase II of the EU emission trading scheme.
Energy,vol 74 pp. 585-594.
Busch, T. & Hoffmann, H. V., 2011. How hot is your bottom line? Linking carbon and financial performance. Business and society, vol 50, pp. 233-265.
Chapple, L., Clarkson, P. M. & Gold, D. L., 2013. The cost of carbon: capital market effects of the proposed emission trading scheme (ETS). Journal of accounting, finance and business
studies, vol 49, pp 1-33.
Christoph, B., Xaquín, G.-M., Mikel, G.-E. & Luis, R., 2015. US climate policy: A critical assessment of intensity standards (working paper). Oldenburg Discussion Papers in Economics V-384-15.
Cooper, R., 2010. Europe's emissions trading system. Harvard project on international
climate agreements discussion paper10-40.
Dowell, G., Hart, S. & Yeung, B., 2000. Do corporate global environmental standards create or destroy market value. Management science,vol 46, pp. 1059-1074.
European Union, 2013. The EU emissions trading system (EU ETS), s.l.: Publication office. Flannery, T., 2013. Why Canada failed on Kyoto and how to make amends. [Online] Available at:
https://www.thestar.com/news/insight/2009/11/22/why_canada_failed_on_kyoto_and_how_to _make_amends.html
[Accessed 14 May 2016].
Gallego-Alvarez, I., Segura, L. & Martinez-Ferrero, J., 2014. Garbon emission reduction: the impact on the financial and operational performance of international companies. Journal of
Cleaner Production,vol 103, pp. 149-159.
Hamilton, J. D., 1996. This is what happened to the oil price-macroeconomy relationship.
Journal of monetary economics, vol 38, pp 215-220.
Hart, S. L. & Ahuja, G., 1996. Does it pay to be green? Empirical examination of the relationship between emission reduction and firm performance. Business strategy and
environment,vol 5, pp. 30-37.
King, A. A. & Lenox, M. J., 2001. Does it really pay to be green? An empirical study of firm environmental and financial performance. Journal of industrial ecology, vol 5, pp. 105-116. Kumar, S., Managi, S. & Matsuda, A., 2011. Stock prices of celan energy firms, oil and carbon markets: a verctor autoregressive analysis. Energy economy, vol 34, pp. 215-226. Lee, Y.-J., 2007. Diagnostic testing for dynamic panel data models. The 14 conference on
33
Luo, L., Lan, Y.-C. & Tang, Q., 2012. Corporate incentives to disclose carbon information: evidence from the CDP global 500 report. Journal of international financial management &
accounting, vol 23, pp. 94-120.
Malina, R., McConnachie, D., Winchester, N., Wollersheim, C., Paltsev, S., Waitz I. A., 2012. The impact of the Euripean Union Emissions Trading Scheme on US aviation. Journal
of Air Transport Management, vol 19, pp. 36-41.
Martin, R., Muuls, M. & Wagner, U. J., 2016. The impact of the European Union emissions trading scheme on regulated firms: what is the evidence after ten years?. Review of
environmental economics and policy , vol 10, pp. 129-148.
Oestreich, M. & Tsiakas, I., 2015. Carbon emission and stock returns: Evidence from EU emission trading scheme. Journal of baning and finance, vol 58, pp. 294-308.
Peteraf, M. A. & Barney, J. B., 2003. Unraveling the resource-based tangle. Managerial and decision economics, vol 24, pp. 309-323.
Pindado, J., Requejo, I. & Torre, C., 2012. Do family firms use dividend policy as a governance mechanism? Evidence from the Euro zone. Corporate Governance: An
International Review, vol 20, pp 413-431.
Schmalensee, R., Stoker, T. M. & Judson, R. A., 2001. World carbon dioxide emissions: 1950-2050. The review of economics and statistics, pp. 15-27.
Scholtens , B. & van der Groot, F., 2014. Carbon prices and fimr's financial performance: an industry perspective. Carbon Management, vol 5, pp. 491-505.
Sharma, S. & Vredenburg, H., 1998. Proactive corporate environmental strategy and the development of competitively valueable organizational capabilities. Strategic management
journal, vol 19, pp 729-753
Trumpp, C. & Guenther, T., 2015. Too little or too much? Exploring U-shaped relationships between corporate envirnmental performance and corporate financial performance. Business
strategy and the environment.
Veith, S., Werner, J. & Zimmermann, J., 2009. Capital market response to emission rights returns:evidence from the European power sector. Energy Economics, vol 31, pp. 605-613. Waddock, S. A., & Graves, S. B. (1997, April). The corporate social performance-financial performance link. Strategic Management Journal, vol 18, 303-319.
Wally, N. & Whitehead, B., 1994. It's not easy being green. Havard business review, vol 72, pp. 46-53.
Wang, L., Li, S. & Gao, S., 2014. Do Greenhouse Gas Emissions Affect Financial
Performance? – an Empirical Examination of Australian Public Firms. Business strategy and
environment, vol 23, pp. 505-519.
