Financial performance and CO
2-emissions: an
examination of their relationship and the moderating
effect of environmental score, emission regulations and the
public’s awareness of climate change.
Abstract
This study investigates the relationship between CO2-emissions of firms, their financial performance and whether environmental performance, emission regulations and awareness of climate change have any influence
on the relationship. We examine a sample of 906 firms from 42 countries for the period 2010-2019. Using an OLS regression with year, industry, and country fixed effects, we found that CO2-emissions have a negative
influence on a firm’s financial performance. In addition, environmental performance seems to positively moderate this effect. Also, we find mixed results for the public awareness of climate change. Finally, emission
regulation does not seem to significantly influence this relationship. We conclude that firms that lower their CO2-emissions and, simultaneously, increase their environmental performance, experience higher financial
performance.
Master’s Thesis International Financial Management - EBM022A20
Author: Jasper Kruizenga Student number: S2968355
Introduction
The effects of climate change become more and more severe and visible. The recent wildfires in Australia, heavy rain fall in Spain, the abnormal temperature rises in the Arctic region of Siberia, or the overall rise of the sea level, are just a few of many examples. The consequences have an impact on the way we live, but also on the way we do business. In order to diminish these consequences as much as possible, The Paris Climate Agreement was enacted on the 12th of December, 2015. Its aim is to retain the global rise in temperature below 2.0C. The Climate Accountability Institute stated in a press release that the twenty biggest fossil fuel companies in the world are responsible for 35% of the global CO2 emission. A recent study by the Potsdam Institute for Climate Impact Research (Glanemann et al., 2020) shows that the efforts that are needed to achieve the reduction in emission, will be economically favorable. It is argued that the costs of adhering to the Paris Agreement outweigh the costs of the adverse effects of doing nothing.
In this paper, we study the relationship between the CO2-emissions1 of firms and their financial performance. Our data covers 906 firms from 42 countries around the world. This international sample allows us to investigate several country-related variables that could be of influence on the relationship. That is, our model includes several moderating effects are investigated, which are the impact of a firm’s overall environmental score, the public awareness of climate change in the country where the company resides in, and whether or not a country has any policy directed towards reducing emissions.
We investigate the relationship between CO2-emissions and financial performance for the period 2010-2019. In addition to the variables used to examine the moderating effects
discussed earlier, we included some control variables to reduce omitted variable bias as much as possible. However, there are more sources of endogeneity which can create bias in our results. With regard to autocorrelation and reversed causality issues of our independent variables, we use lagged versions of all our time-variant variables. Also, we control for year-, industry-, and country fixed effects, to reduce bias from time-invariant omitted variables.
The results show that a negative relationship between CO2-emissions and financial performance exists, meaning that high emissions harm the financial performance of firms. Which is likely the result of investors punishing firms for their polluting behavior. We also found that environmental performance of firms positively moderates this effect, this is likely the result of firms emphasizing their pro-environmental behavior to the market, and therefore getting awarded accordingly. Furthermore, country-level emission regulations do not seem to have any influence on the relationship. Also, conflicting results are found with regard to public awareness. Namely, the awareness of “global warming” of country’s inhabitants seems to have a positive moderating effect on the relationship between CO2-emissions and firm financial performance. Whereas the awareness of “climate change” does not seem to have a significant effect.
This paper complements previous findings in several ways. Alvarez (2012) found that, although firms significantly decreased their CO2-emissions in the period 2006-2008, no effect of this reduction on firm performance was measured in the subsequent three years2. Our analysis of a longer time period contradicts these findings. In addition, this paper tries to complement Claver et al. (2007) and Gonenc and Scholtens (2017) by analyzing the environmental performance of firms from a wide array of industries. Finally, Choi et al. (2020) investigated the effect of unusual temperature rises on the attitudes of investors in cities
towards highly polluting firms. They found that these high-emission firms underperform due to the increasing awareness of climate change. We add to these findings and build on the novel notion by using these attitudes of countries’ inhabitants as a variable and see the influence on the relationship of CO2-emitting firms and their financial performance. As Climate Change is a phenomenon that has an impact on everyone in the world, including all businesses, we want to emphasize the importance of people’s awareness of the phenomenon.
The rest of this study is structured as follows. Firstly, we will discuss findings from previous literature and construct our hypotheses. Secondly, we will discuss where we extracted our data from and how we constructed our variables. Then, we will discuss our research design. Subsequently, the regression results will be discussed. Then we will discuss our conclusion. Finally, we will elaborate on managerial implications and the limitations of our research.
Background and Hypotheses
Friedman (1970) reasoned that firms have a sole purpose of maximizing value for their shareholder, as “social responsibilities” are the concern of individuals and not business entities. Essentially this would mean that businesses mainly need to focus on increasing their financial performance. Nowadays, the consensus shifted away from this idea, as businesses seem to be more involved in these “social responsibilities”, which do not harm financial performance and might even increase them (Margolis et al., 2009)3. Moreover, as part of these social concerns, the environment has attained increased attention. Scholars proclaim that businesses need to prioritize reducing emissions in order to reduce climate change, especially in heavily industrialized countries (Krausmann et al., 2018; Lee et al., 2015; Krausmann et al., 2020; Kagawa et al., 2015).
Many studies have been focusing on how the stock market reacts to global warming. Climate change will have an impact on all investments in every sector of the economy (Ritchie and Dowlatabadi, 2015). Choi et al. (2020) found that the rise of temperature in a city causes high-carbon-emission firms’ stock to underperform in the period that temperatures are abnormally high, and interestingly, this effect is not reversed in the long-term. This finding suggests that climate change itself, when physically experienced, could harm the financial performance of firms.4 In addition, mutual funds also seem to react to noticeable climate related phenomena. They seem to decrease their weights of stocks of companies that are located in an area where there has been a climate related calamity. Interestingly, this shift away from these stocks does not seem to be due to the plummeting of the stock prices, but rather to the fact that these firms operate in an area that is liable to climate change related calamities (Alok et al., 2020). Investors, as highlighted by Ritchie and Dowlatabadi (2015), seem to be more prone towards re-investing in different, less polluting sectors than investing in emission reducing technologies. Nonetheless, they also note that investors’ ability to overcome this impact of climate change and greenhouse gases on their investments is very limited. In a similar vein, Krueger et al. (2020) conducted a survey under investors and found that investors do believe that risks related to climate change have financial repercussions for firms in their portfolios. It is even stated that investors believe that these financial consequences are already being experienced.
Past studies have tried to shed light on the relationship between emission of carbon dioxide and the financial performance of firms, but often in the context of a particular country or sector. Some studies tried to find a direct relationship between CO2-emissions and financial performance of firms. Alvarez (2012) found a negative effect in 2007 and no significant effect
in 2008-2010. Trinks et al. (2020) investigated whether the carbon efficiency of firms influences their financial performance. They find that firms which are carbon efficient, experience higher financial performance than similar firms with higher emissions. This effect likely stems from efficiency benefits of their operations and that these firms are less liable to carbon emission regulations.5
Others found that firms that do not invest much in environmental innovations and do not invest in lessening emissions, are being “punished” by the market, and their firm value is harmed as well. Whereas firms that actively invest in efforts to reduce their emissions, benefit financially. Therefore, investing in research and development (R&D) efforts can be financially favorable and successfully reduces emissions (Lee et al., 2015; Cole et al., 2013). Lioui and Sharma (2012) even stipulate that, for environmental corporate social responsibility to be beneficial to firms, a firm has to have R&D activities, that will become more efficient because of the environmentally responsible actions.
As for environmental performance of firms, a significant amount of research has been conducted on the matter. Claver et al. (2007) stipulated that agricultural businesses that have a proactive attitude towards environmentally friendly initiatives could benefit from enhanced economic performance. In addition, Gonenc and Scholtens (2017) investigated the relationship between financial performance and environmental performance of fossil fuel firms. It is found that firms with higher financial performance (measured as Tobin’s Q and return on equity), generally have higher environmental scores. However, this relationship between Tobin’s Q and environmental performance is negative for fossil fuel-related firms, and it is even found that better financial performance implies higher emissions. Although Scholtens (2008) concludes
that it is much more common for financial performance to come before social performance, it can differ between aspects of CSR. Therefore, one should mind that studies (Waddock and Graves, 1997; Margolis et al., 2009) have found indication that the relationship between social performance and financial performance could be reversed as well, that is, high financial performance can lead to better social performance. To tackle this reversed causality issue, Gonenc and Scholten (2017) use two- and three-stage least squares estimations. However, Lee et al. (2015) use lagged variables in their regression to overcome reverse causality bias. We follow this example.
In a similar vein, Fonseka et al. (2019) investigated the effect of environmental information disclosure on the cost of equity capital for Chinese energy firms. They found that for companies that produce “clean” energy, their cost of capital decreases with increasing disclosure of environmental information. However, contradictory to Gonenc and Scholtens (2017), Fonseka et al. (2019) found that for companies that still use fossil fuels to produce energy, the cost of capital increases, which has an impact on their financial performance. Furthermore, Kolk et al. (2001) investigated the environmental reporting efforts of the Fortune Global 250 and whether the country of origin and industry they operate in has an influence. Although they found that regulations have some influence on the environmental reporting efforts of firms, they stipulate that the interest of the public in environment related issues seems to be of greater importance.
With regards to emission regulations and policies, it seems that the most viable way to reduce the overall emission of carbon dioxide is through government actions6. Tvinnereim et al. (2017) show that the general consensus of the population in Norway prefers government policies over individual actions. Although these results are specific to Norway, it does give an
insight in the attitude of inhabitants of a Western democracy. Furthermore, as accords like the Paris Climate Agreement are not legally enforceable, countries themselves need to implement regulations and policies that restrict CO2-emission. It is shown by Lee et al. (2015), that its predecessor, the Kyoto Protocol, which was legally binding, had a positive influence on environmental investments and thus in the long-run enhanced firm performance. It also required companies to disclose their CO2-emission, which in turn resulted in additional positive environmental actions (Lee et al., 2015). Rozenberg et al. (2020) investigated the effects of different policy options to reduce CO2-emission as much as possible. They found that in the long-run all policy instruments investigated (i.e. carbon tax or carbon price levied through a market) ultimately lead to a transition to clean capital and that the maximum emission value, as established by the Paris Agreement7, will not be exceeded.
As mentioned before, public awareness seems to play a major role in the directions taken by companies in addressing climate related issues (Kolk et al, 2001). Furthermore, empirical research shows that the development of awareness depends on several factors. Vu et al. (2019) conducted an analysis on how the media in 45 countries around the world portrait climate change in the news and found that this often plays an important role in the public gaining awareness of climate change. It seems that the way climate change is framed by media varies significantly and is influenced by local political, socioeconomic and environmental factors. Furthermore, Sampei and Aoyagi-Usui (2009) showed that the mass-media helped increase the people’s awareness of climate change. To support their argument, they use the example of the Japanese government. In their attempt to combat climate change, they needed the media to raise and maintain the public’s awareness. Moreover, Choi et al. (2020) observed
that a rise in local temperatures prompts peoples’ attention towards climate change, concluding that once people physically experience the effects of climate change, their awareness rises.
Based on the wide array of literature discussed, we constructed several hypotheses are constructed and discussed:
Firstly, as Fonseka et al. (2019) and Krueger et al. (2020) both emphasized that the general focus of providers of capital shifts towards more environmental friendly firms. One could derive from this that firms focusing on reducing CO2-emission as much as possible, results in easier access to capital, which in turn can induce investment and increase financial performance. Also, as Lee et al. (2015) found, the market punishes those firms that do not engage in environmental friendly practices. This could entail that companies that keep emitting high levels of CO2 will underperform. Therefore, it is hypothesized that a firm’s CO2-emissions have a negative influence on its financial performance (H1).
environmental score has a positive moderating effect on the relationship between CO2-emissions and financial performance (H2).
Thirdly, as governments are progressively implementing policies and regulations managing CO2-emission of companies, this will likely have an impact on the relationship between financial performance and their carbon emission. Also, as emphasized by Lee et al. (2015), the policies will enhance environmentally friendly investments and can eventually lead to enhanced firm performance. Firms that are more environmentally friendly are less susceptible for emission regulations (Trinks et al., 2020). What is more, Rozenberg et al. (2020) stipulate that an optimal carbon price can result in achieving the emission goal8 as quickly as needed. This could increase costs for companies that have high-levels of CO2-emission and decrease their cash flows resulting in lower financial performance. Moreover, this could induce a shift towards less carbon-intensive capital, which in the long-run could result in higher financial performance of low CO2-emitting firms. In conclusion, regulations with regards to emission reduction, increases the costs for heavy polluting firms, which reduces their financial performance. Therefore, we hypothesize that operating in a country that has implemented a Carbon Tax negatively moderates this relationship between CO2-emissions and financial performance (H3).
Finally, still a significant amount of people exist that still underestimate the adverse effects of climate change (Choi et al., 2020), or consider it an exogenous phenomenon (Van Wijnbergen en Willems, 2015). As awareness seems so be a key factor in overcoming the adverse effects of climate change, and increases environmental efforts of firms (Kolk et al., 2001; Sampei and Aoyagi-Usui, 2009), it is important to take this into consideration in our analysis. Intuitively, it makes sense to assume a negative relationship exists between CO2-emissions of firms operating in a particular country and the awareness of climate change
inhabitants within that country. As the interest of the public in climate change increases, it is likely that there will be more emphasis put on reducing global warming, resulting in even more pressure on heavier polluting companies with higher CO2-emissions. Following this reasoning, it is hypothesized that the degree of climate change awareness in a country negatively moderates the relationship between CO2-emissions and financial performance (H4).
Data and Variables
This study investigates the relationship between CO2-emissions and the financial performance of a sample of firms from a wide array of industries for the period 2010-2019. This time span is primarily based on the availability of CO2-emission data. In comparison to previous research (Claver et al., 2007; Cole et al., 2013; Lee et al., 2015) we use a highly international sample, comprising of 42 countries and a recent period. Also, we take into consideration the general public’s attitudes towards climate change, which has not been researched extensively to date but is believed to be of great importance (Kolk et al. 2001; Sampei and Aoyagi-Usui, 2009).
Following Gonenc and Scholtens (2017), this paper draws its firm-level data on firm financials from the Thomson Reuter’s EIKON database. In addition, Thomson Reuters also provides access to the Asset 4 Environmental Social Governance (ESG) database. This database contains historical data on four pillars, which are corporate governance, social, economy, and environment. In this study, we use CO2 Equivalents Emission Total and the
With regards to the country-level data, we consulted two databases. In order to see whether a country has implemented a carbon tax or a related carbon regulating policy instrument, this study accessed the World Bank’s Carbon Pricing Dashboard. Previous literature has used either data from the World Bank or reports as a source (Cole et al, 2013; Rozenberg et al, 2018; Vu et al., 2019), which acknowledges the reliability of this data source. This dashboard provides up-to-date data on which countries around the world have implemented or are planning on implementing regulations regarding carbon pricing. It consists data ranging from 1990 till 2020.
Following Choi et al. (2020), in order to measure the general awareness of a country’s residents of climate change, and determine its moderating effect, the Google Search Volume Index is consulted. This database contains search volume data from 2004 till present, for every country in the world. Google provides this data via Google Trends, and allows one to search any word or topic regarding any subject. Google’s algorithms translate, group and sort the search data, which allows one to investigate the search history of a certain topic geographically ranging from global data up till city-level. The available data is normalized, anonymized, sorted in categories based on topics and grouped. For inclusion purposes we both extracted search volume data on the topic “global warming” as well as “climate change”.
Due to the scarce complete data of CO2-emissions for each firm for every year in a 2010-20199 time period, a sample of 10,290 firm-year observations from 42 countries where the headquarters are located, is collected. After deleting financial firms from our sample10, we were left with an unbalanced panel of 9,060 firm-year observations.
9 As the EIKON database only provided fiscal year data time series data, it is assumed for simplicity reasons that all fiscal year periods are similar across our sample of firms from different countries.
Main variables
We follow Gonenc and Scholtens (2017) in the use of the firm-level variables. The dependent variable that we use is Tobin’s Q. This proxy for financial performance is often used in academic literature and defined as the ratio of (book value of total assets + market value of common equity − book value of common equity) to book value of total assets. As the market value of equity is part of Tobin’s Q, it provides us with a good representation of long-term effects on financial performance, since stock market valuations are often future-oriented (Trinks et al., 2020). Also, as Tobin’s Q is based on the valuations of the market, it is much less liable to manipulation by managers or accounting regulations (Scholtens, 2008).
The main independent variable is the CO2-emission of firms. Using the Asset 4 ESG database this study uses the “CO2 Equivalents Emission Total” (CO2EET) as a proxy for firm CO2-emission. The Asset4 ESG database defines this variable as the “Total CO2 and CO2 equivalents emission in tonnes”. In order to be able to infer some meaningful insights, considering most absolute emissions amounts are very large, we take the natural log of CO2-emissions.
In addition, this database also contains an “Environmental Pillar Score”, which is an average score of all indicators that have an influence on how a firm impacts living and non-living natural systems. It reveals how a firm is able to use best management practices to mitigate environmental risks to create long-term shareholder value. This score is used as another firm-level independent variable, which will be a proxy for the environmental performance of the firms in our sample.
Finally, our last independent variable will be the climate change awareness country-level variable. As opposed to Choi et al. (2020), who only use “global warming” as a proxy for awareness, we include search data for both the topic “global warming” and “climate change”, as these terms are often used interchangeably. This variable will also be ranging from 0 to 100, 0 being no searching popularity in a country regarding climate change at all and 100 being the most popular searching topic in the particular country. This country-level variable will be used as a proxy for the awareness of a country’s citizens.
Control Variables
To be able to give some economical meaning to our analysis and overcome the issue of omitted variable bias, this study also includes a wide array of control variables. Control variables included in this study are size, leverage, innovativeness, capital intensiveness, energy intensiveness, dividends, and sales growth.
First of all, this study controls for size, measured as the natural log of Assets in USD (Gonenc and Scholtens, 2017; Alok et al., 2020; Trinks et al., 2020). In order to be able to compare the differences in size between firms from different countries, the assets have to be converted to USD. Size is likely to have a negative influence on Tobin’s Q, as it is likely that larger firms, i.e. firms with more tangible assets, will have more polluting assets.Also, larger firms are often subject of higher scrutiny with regards to adhering to (emission) regulations. This could lead to higher costs which could affect financial performance (Cole et al., 2013).
CO2-emissions is. However, intuitively it could be that leverage is negatively correlated with the CO2-emission of firms as a higher leverage often means that businesses take on more debt to finance their investments, which should be directed towards less polluting assets, as investors seems to value those companies higher (Fonseka et al., 2019; Krueger et al., 2020).
A proxy frequently used in academic literature for innovativeness is the ratio of research and development (R&D) expenditures to total assets. The general focus of firms shifts towards more environmentally sustainable business practices. This could also mean that R&D expenditures are channeled towards CO2-emission-reduction practices. Resulting in a positive effect on the relationship between CO2-emissions and financial performance (Cole et al., 2013; Lioui and Sharma, 2012). Also, R&D expenditures often increase the (carbon) efficient use of assets, which can result in better financial performance (Trinks et al., 2020).
As for the capital intensity, it is argued in the literature that it likely has a negative influence on CO2-emission of firms (Cole et al., 2013). The reasoning is that the higher a firm’s capital intensity, the higher its dependence on machinery, which often cause CO2-emission. This would subsequently lead to lower financial performance, as heavy polluters are likely to be negatively valued by the market. Capital intensity is computed as the total amount of fixed assets (Plant, Property, and Equipment (Trinks et al., 2020)) over Total Assets.
Furthermore, as the energy sector is one of the most polluting industries there is (Shive and Forster, 2020), it is likely that this will be reflected in firms’ total CO2-emissions. Following the same reasoning as with capital intensiveness, operating in the energy sector is negatively influencing Tobin’s Q, because of high CO2-emissions. Therefore, following Lee et al. (2015), we constructed a dummy variable for firms operating in the energy sector11, being 1 if a firm operates in the energy sector and 0 otherwise.
In addition, we control for dividends, by constructing a dummy variable which attained a value of 1 if a firm paid out dividends to its shareholders in a particular year, and 0 otherwise. Dividends are likely to be positively related to financial performance, as firms often distribute dividends to its investors if they realize excess returns resulting from high financial performance. Investors often expect increasing dividend payments if Tobin’s Q is higher than average (Lang and Litzenberger, 1989).
Finally, we control for sales growth, computed as the ratio of (Total Revenuet – Total Revenuet-1) / Total Revenuet-1. Sales growth is an alternative measure for firm performance (Alok et al., 2020), and is therefore assumed to be negatively related to CO2-emissions, as indicated earlier. This would also mean that it is positively related with Tobin’s Q, as higher sales growth is likely to result in a higher valuation of a company, as the company is clearly able to benefit from additional sales opportunities.
Methodology
Our study aims to investigate whether a relationship exists between the CO2-emissions of firms and their financial performance. We would also like to identify whether this relationship, if it exists, is moderated by the environmental performance of a firm, whether the country of origin is regulating emission levels, and to what extent the residents of a country are aware of climate change. We include interaction terms of all of these variables with CO2-emission, to be able to investigate whether significant differences exists if a company scores high on the Environmental Pillar Score, operates in a country where a emission regulation is in effect or public awareness of global warming is relatively high.
Tobin’s Q,i,t = β1 CO2EET,i,t + β2 Environmental Pillar Score,i,t + β3 CO2EET,i,t ×
Environmental Pillar Score,i,t + β2 Emission regulation,c,t + β3 CO2EET,i,t × Emission
regulation,c,t + β4 Public awareness,c,t + β4 CO2EET,i,t × Public awareness,c,t + θ′Xi,t + δc
+ηj +φt + εi,t,
Where, θ′Xi,t represent the control variables, as described earlier. Furthermore, δc, ηj, and φt embody the country, industry and year fixed effects, respectively.
In comparison to previous studies (Claver et al., 2007; Cole et al., 2013; Lee et al., 2015), our analysis comprises a very international (i.e. 42 countries) sample from a recent time period (i.e. 2010-2019). Businesses from a wide array of countries and different industries are studied, as Global Warming is not just a phenomenon that has influence on a particular country or industry, but is of concern to everyone. In addition, the value in a particular year of many variables in our regression is likely influenced by the value of that variable in a previous year. Also, our dependent variable Tobin’s Q could drive several of our independent variables as well. For example, Waddock and Graves (1997) and Margolis et al. (2009) provided evidence that some studies find that CSR influences financial performance, whereas often financial performance comes first (Scholtens, 2008). This could be the case for most of our variables. Therefore, in order to tackle the issue of reversed causality and autocorrelation, we follow Lee et al. (2015) and use lagged variables for all time-variant variables in our regressions. In addition, we control for leverage, size, innovativeness, capital intensiveness, energy intensiveness, dividends, and sales growth.
Results
Our analysis aims to investigate whether a relationship exists between CO2-emission and financial performance. First, the descriptive statistics will be presented and an univariate analysis will be conducted. Thereafter, the outcomes of the regression analysis will be discussed .
In Table 1, the summary statistics for the countries in our sample can be found. As can be seen, of our international sample of 42 firms, most firms operate in Japan, The United Kingdom and The United States. Although the number of observations are relatively low, firms from Finland, Portugal, and Spain score the highest on environmental performance. This entails that companies in these countries, on average, are perceived to do more with regards to protecting the environment and orienting on the longer-term. In addition, it can be drawn from the table that in 27.8% of the firm-year observations a country had an emission policy in place. This means that in almost three-quarter of the observations in our sample period, countries did not have any regulations with regards to emission reduction.
What also stands out is that all firms in our sample are considerably large, with
20.236 being the smallest firm. As size is measured as the natural log of assets in USD, this is already a major company. Furthermore, firms do not seem to invest a lot in R&D efforts, averaging a 1.2% on their R&D expenditures in comparison to total assets. Moreover, firms in the sample are not really dependent on their fixed assets, as the average capital
Table 1
Sample countries and summary statistics. This table shows the number of observations and means for the main firm-level variables and all country-level variables used in the regressions. The firm-level data for all companies from the 42 countries below are obtained from the Thomson Reuters EIKON database. The country-level emission regulation data is acquired from the World Bank’s Carbon Pricing Dashboard and the country-level awareness data is extracted from the Google Search Volume Index. All obtained data is from the period 2010 to 2019. Tobin’s Q represents the ratio of (book value of total assets + market value of common equity − book value of common equity) to book value of total assets. CO2EET represents the natural log of “CO2-emission and equivalents
total” in tons. The Environmental Pillar Score is a score from 0 to 100, indicating the aggregate score on how a company impacts the environment. Emission
regulation is an indicator variable, being 1 if a country has an active law regulating CO2-emission in a particular year. Awareness Global Warming and Awareness Climate Change both measure the search interest in a country per year, ranging from 0 to 100, 100 being maximum interest.
Country N Tobin’s Q CO2 EET Environmental Pillar Score
Emission Regulation Awareness Global Warming Awareness Climate Change Australia 370 1.518 13.668 58.877 0.8 19.583 29.558 Austria 50 1.073 14.384 66.645 0 15.542 19.542 Belgium 90 1.480 12.298 64.530 0 18.658 31.583 Brazil 130 1.350 14.027 68.149 0 3.408 12.825 Canada 380 1.326 13.987 61.219 0.1 17.5 25.65 China 20 1.385 14.274 61.246 0 20.75 7.692 Colombia 10 1.691 15.883 60.136 0 10.783 36.367 Denmark 110 2.728 12.326 60.496 1 20.25 16.358 Finland 150 1.539 12.898 74.121 1 24.78 22.05 France 460 1.435 13.219 72.961 0.6 20.458 30.375 Germany 340 1.386 14.136 72.318 0 23.892 28.717 Greece 40 1.452 14.981 59.818 0 8.308 13.5 Hong Kong 60 1.212 14.891 60.895 0 23.275 22.875 Hungary 20 1.000 13.536 72.489 0 6.875 13.017 India 140 2.419 14.619 66.596 0 18.042 13.958 Ireland 90 2.045 13.123 61.532 1 16.783 19.725 Israel 10 1.462 11.610 46.382 0 19.542 13.1 Italy 140 1.102 14.290 73.315 0 22.458 17.458 Japan 1720 1.237 13.452 64.676 0 16.867 19.283 Jersey 10 1.712 12.398 59.335 0 14.408 3.05 Luxembourg 20 1.195 16.876 73.561 0 14.933 11.475 Malaysia 10 5.529 11.771 48.224 0 14.767 17.625 Mexico 50 1.929 14.790 71.517 0.6 18.542 28.408 Netherlands 170 1.618 13.494 70.278 0 23.975 25.85 New Zealand 40 1.476 11.964 51.089 1 17.9 22.725 Norway 80 1.622 12.440 67.593 1 13.217 24.95
Table 2
Descriptive statistics for all variables used in the analysis. Tobin’s Q is the ratio of (book value of total assets + market value of common equity − book value of common equity) to book value of total assets. CO2EET is the natural log of CO2-emission and equivalents total of firms in tons. Environmental Pillar Score is a score ranging from 0 to 100, 100 being the highest, reflecting the efforts a firm undertakes to protect the environment. Emission
regulation is a dummy variable being 1 if a country has an emission related policy in place in a certain year, and
0 otherwise. Awareness Global Warming represents the popularity of the search topic “Global Warming”, 100 being most popular and 0 being no popularity at all. Awareness Climate Change represents the popularity of the search topic “Climate Change”, 100 being most popular and 0 being no popularity at all. Leverage is the ratio of total debt over total assets. Innovativeness is the ratio of R&D expenditures to total assets. Capital Intensiveness is the ratio of total fixed assets (i.e. Plant, Property and Equipment) over total assets. Energy Intensive is a dummy variable being 1 if company operates in the energy sector, and 0 otherwise. Dividends is a dummy variable being 1 if a firm paid any dividends in a year, and 0 otherwise. Sales growth is the ratio of (Total Revenuet – Total Revenuet-1) / Total Revenuet-1. The sample period is 2010-2019.
Number of observations (N) = 9,060
Mean Median StdDev. Minimum Maximum
Tobin’s Q 1.630 1.323 0.915 0.689 5.707
CO2EET 13.471 13.482 2.267 7.561 18.459
Environmental Pillar score 64.630 67.257 19.086 0.331 98.726
Emission regulation 0.278 0 0.448 0 1
Awareness Global Warming 18.552 16.583 8.058 0 59.833 Awareness Climate Change 26.102 24.167 12.120 0 74.333
Leverage 0.259 0.248 0.149 0 0.702 Size 23.226 23.184 1.321 20.236 27.340 Innovativeness 0.012 0 0.024 0 0.123 Capital Intensiveness 0.306 0.265 0.224 0 0.872 Energy Intensiveness 0.075 0 0.263 0 1 Dividends 0.853 1 0.354 0 1 Sales growth 0.026 0.015 0.170 -0.876 5.840
In order to capture the effect CO2-emissions have on firm’s financial performance, we conducted an OLS-regression with lagged variables, of which the results are shown in Table 3 and 4. The use of lagged variables helps us control for autocorrelation and reversed
The adjusted-R2 of our main model is 0.471, which suggests that are model is properly able to explain the variation in Tobin’s Q.
In all our regressions that include CO2EET, we find a statistically significant negative effect on Tobin’s Q. This suggests that CO2-emissions lower the financial performance of firms. The coefficient of interest for hypothesis 1 is obtained from Table 4, that is, -0.103. As this coefficient is highly significant it suggests that there indeed exists a negative relationship between Tobin’s Q and CO2-emissions of firms. In line with Lee et al. (2015) and Krueger et al. (2020), it is likely the result of the stock market punishing firms for harming the
environment. Investors place less value on heavier polluting firms, which harms their financial performance. Therefore, we can accept the first hypothesis.
(Fonseka et al., 2019; Krueger et al., 2020), which are oftentimes companies with high environmental performance. These findings are in unison with hypothesis two.
To determine whether emission regulations have any influence on the relationship between CO2-emissions and firm financial performance in our sample, we again have to examine multiple coefficients. As can be drawn from Table 3, the effect of emission regulations on Tobin’s Q is negative and statistically significant at the 5-percent level. This would likely be the result of an increase in costs for heavily polluting firms due to these emission reducing regulations (Lioui and Sharma, 2012; Rozenberg et al, 2020). However, in our regression in Table 4, both coefficients, that of the standalone effect of an Emission Regulation, as well as the interaction term between Emission Regulation and CO2EET are both statistically insignificant. This means that we failed to find any moderating effect of emission regulations on the relationship between CO2-emissions and firm’s financial
performance. Although the literature suggests that countries with emission reducing policies will harm the financial performance of heavy polluting firms (Rozenberg et al., 2020), our sample comprises of large firms who are likely highly multinational, and are thus able to circumvent possible regulations in countries where no carbon policies are active. Thus, we can conclude that hypothesis three is rejected.
firms. Resulting in environmentally friendly firms are rewarded proportionally and heavy polluters underperform (Choi et al., 2020). The net effect seems to be positive in our sample.
Now, let us consider the moderating effect of the awareness of “global warming”. In Table 4, it seems that the awareness of global warming itself has a statistically significant effect on Tobin’s Q (i.e. -0.037). This means that the public’s awareness of global warming harms the financial performance of firms in our sample. Which is the opposite effect as obtained from Table 3. This could be for the same reason as mentioned earlier, but now this net effect is negative in our sample. If one looks at the interaction term between CO2EET and awareness of global warming, it can be seen that the statistically significant coefficient is 0.003, which implies that in countries where the public’s awareness of global warming is higher, the negative effect of CO2-emissions on firms’ financial performance is less negative than in countries where there is lower awareness. This contradicts the findings of Choi et al. (2020) who found that higher awareness leads to underperformance of polluting stocks. This is also an opposite effect as hypothesized in hypothesis four, and provides us with grounds to reject it.
As for the control variables in our analysis, highly statistically significant effects are captured by size, innovativeness, dividends, and sales growth. Also, in some of our
regressions in Table 4, capital intensiveness has an effect on Tobin’s Q, which is statistically significant at the 10 percent level. Leverage and energy intensiveness do not have statistically significant effects in our regressions.
Size seems to be negatively related to Tobin’s Q, as also found by Gonenc and Scholtens (2017). Innovativeness strongly positively influences Tobin’s Q, which could be a result of innovations that generate long-term value for corporations (Cole et al., 2013) and, as emphasized by Lioui and Sharma (2012), R&D positively influenced by environmental performance, which results in additional financial performance. Capital intensiveness also seems to have a statistically significant impact on Tobin’s Q, which could stem from the fact that the majority of firms in our sample generate profits from their fixed assets. In addition, the Dividends Dummy is also positively related to Tobin’s Q, as shown by Lang and Litzenberger (1989), investors expect higher dividends with higher financial performance. Finally, Sales growth also has a positive effect on Tobin’s Q, which makes sense, as sales growth can be used as an alternative proxy for financial performance (Alok et al., 2020).
Table 3
This table represents the OLS-regression results of the relationship between CO2EET and Tobin’s Q, the effect of the Environmental Pillar Score on Tobin’s Q, the effect of the Emission regulation on Tobin’s Q, and the effect of the Awareness of Global Warming variables on Tobin’s Q, separately and taken together. Definitions of all variables are presented in Table 2. All time-variant variables are lagged. The regression controls for industry-, country-, and year fixed effects. The sample period ranges from 2010 to 2019. Robust standard errors are clustered at the firm-level and presented between brackets. Furthermore, *, **, and *** represent
significance levels at the 10%, 5%, and 1% levels, respectively.
Dependent variable: Tobin's Q (1) (2) (3) (4) (5) (6)
CO2EET -0.044*** -0.043**
(0.017) (0.017)
Environmental Pillar Score 0.002 0.002
(0.001) (0.001)
Emission Regulation -0.073** -0.060**
(0.030) (0.030)
Table 4
This table represents the OLS-regression results of the relationship between CO2EET and Tobin’s Q and the moderating effects of the Environmental Pillar Score, Emission Regulations, and the Awareness variables, by including interaction variables. Definitions of all variables are presented in Table 2. All time-variant variables are lagged. The regression controls for industry-, country-, and year fixed effects. The sample period ranges from 2010 to 2019. Robust standard errors are clustered at the firm-level and presented between brackets. Furthermore, *, **, and *** represent significance levels at the 10%, 5%, and 1% levels, respectively.
Dependent variable: Tobin's Q (1) (2) (3) (4)
CO2EET -0.108*** -0.034* -0.045** -0.103***
(0.034) (0.018) (0.020) (0.037)
Environmental Pillar Score -0.012* -0.011*
(0.006) (0.006)
CO2EET x Environmental Pillar Score 0.001** 0.001**
(0.000) (0.000)
Emission Regulation 0.411* 0.260
(0.224) (0.251) CO2EET x Emission Regulation -0.038** -0.027
(0.001) (0.002)
Awareness Climate Change 0.005*** 0.005**
(0.001) (0.002) Leverage -0.175 -0.189 -0.198 -0.205 -0.217 -0.184 (0.157) (0.158) (0.157) (0.157) (0.157) (0.157) Size -0.066*** -0.121*** -0.109*** -0.109*** -0.109*** -0.079*** (0.025) (0.024) (0.021) (0.021) (0.021) (0.028) Innovativeness 3.803*** 3.733*** 3.970*** 3.927*** 3.906*** 3.526** (1.425) (1.415) (1.433) (1.435) (1.435) (1.410) Capital Intensiveness 0.166 0.103 0.104 0.102 0.102 0.164 (0.107) (0.108) (0.108) (0.108) (0.108) (0.108) Energy Intensiveness 0.139 0.043 0.038 0.038 0.037 0.139 (0.154) (0.144) (0.142) (0.142) (0.142) (0.156) Dividends 0.247*** 0.249*** 0.249*** 0.250*** 0.248*** 0.244*** (0.050) (0.050) (0.050) (0.050) (0.050) (0.050) Sales Growth 0.359*** 0.375*** 0.368*** 0.369*** 0.364*** 0.355*** (0.090) (0.090) (0.092) (0.091) (0.090) (0.089)
Year fixed effects YES YES YES YES YES YES
Industry fixed effects YES YES YES YES YES YES
Country fixed effects YES YES YES YES YES YES
Constant 2.833*** 3.501*** 3.375*** 3.150*** 3.130*** 2.871*** (0.497) (0.503) (0.482) (0.480) (0.478) (0.525)
Adjusted-R2 0.464 0.463 0.462 0.463 0.463 0.466
(0.017) (0.019)
Awareness Global Warming -0.042*** -0.037**
(0.013) (0.015) CO2EET x Awareness Global Warming 0.003*** 0.003*** (0.001) (0.001)
Awareness Climate Change 0.036** 0.028*
(0.015) (0.016) CO2EET x Awareness Climate Change -0.002** -0.002
(0.001) (0.001) Leverage -0.126 -0.177 -0.205 -0.156 (0.157) (0.157) (0.156) (0.156) Size -0.082*** -0.064** -0.069*** -0.083*** (0.028) (0.025) (0.026) (0.028) Innovativeness 3.734*** 3.932*** 3.686*** 3.712*** (1.410) (1.429) (1.422) (1.413) Capital Intensiveness 0.167 0.184* 0.176* 0.189* (0.106) (0.108) (0.107) (0.107) Energy Intensiveness 0.151 0.142 0.141 0.152 (0.155) (0.158) (0.156) (0.159) Dividends 0.250*** 0.239*** 0.244*** 0.241*** (0.050) (0.050) (0.050) (0.050) Sales Growth 0.367*** 0.356*** 0.354*** 0.360*** (0.090) (0.091) (0.090) (0.090)
Year fixed effects YES YES YES YES
Industry fixed effects YES YES YES YES
Country fixed effects YES YES YES YES
Constant 3.923*** 2.705*** 2.702*** 3.731***
(0.678) (0.493) (0.524) (0.702)
Adjusted-R2 0.467 0.466 0.467 0.471
Conclusion
We investigate the relationship between CO2-emissions of firms and their financial performance for 906 firms from 42 countries over the period 2010-2019. As climate change is having an increasing impact on the lives of everyone on this planet, it is important to increase its attention and direct efforts to reduce the adverse effects. Some academics already agree that businesses need to prioritize emission reduction to combat climate change, especially in those countries where the industrialization is a major part of the economy (Krausmann et al., 2018; Lee et al., 2015; Krausmann et al., 2020; Kagawa et al., 2015). As a proxy for financial performance, we used Tobin’s Q ratio. In addition, this paper wanted to see whether the relationship is affected by several moderating effects. We wanted to investigate whether a firm’s environmental performance, measured as a firm’s Environmental Pillar Score, had any impact. Further, we investigated whether the existence of an emission regulation in a country is of influence, and, finally, whether the general awareness of a country’s inhabitants of global warming had an influence. Also, year-, industry-, and country fixed effects were included in our regression to overcome omitted variable bias.
our proxy for emission regulation did not have a significant moderating effect on the relationship between CO2-emissions and financial performance, as well.
Several hypothesis were tested in our analysis. The first hypothesis assumed that a negative relationship exists between CO2-emissions and firm’s financial performance. Indeed, we found evidence in favor of this hypothesis. We also found support for our second hypothesis, which stipulated that a positive moderating effect exists of environmental performance on the relationship between CO2-emissions and firm’s financial performance. However, we rejected our third hypothesis, which theorized that emission regulation is strengthening the negative effect of CO2-emissions of firms. We did not find any evidence in support of this hypothesis. Likewise, our fourth hypothesis, suggesting a positive moderating effect of awareness of climate change on the relationship between emissions and financial performance, is rejected as well. As we did not find any conclusive evidence in support of this hypothesis.
Discussion and managerial implications
(Kolk et al, 2001; Van Wijnbergen en Willems, 2015; Choi et al, 2020), is that this study uses the awareness of climate change as a moderating aspect.
From a managerial perspective, our findings can have an influence on potential business decisions of managers in the future. As we showed that there exists a negative relationship between CO2-emissions and financial performance, managers might want to direct more of their investments towards less emitting practices. Additionally, managers could focus on policies that would enhance their environmental performance, as this functions as a positive moderator in the negative relationship. Ultimately, this would also mean that emissions of CO2 would have to come down.
However, this study also has some limitations. First of all, this study uses only one variable to measure the financial performance of firms, namely, Tobin’s Q. Although this measure is widely used in the literature (e.g. Lee et al., 2015; Gonenc and Scholtens, 2017; Trinks et al., 2020) it should be noted that, as it is related to a company’s value, both the costs incurred by doing environmental responsible actions and the potential benefits derived from it, are included in this measure (Lioui and Sharma, 2012). As a result, we cannot fully confirm the relationship between CO2-emissions, environmental performance and financial performance in this study. More measure of financial performance should be added.
As mentioned by Gonenc and Scholtens (2017), both qualitative and quantitative measures of environmental performance should be used and compared between different firms and industries in order to properly assess a firm’s environmental performance. These aspects in relation to environmental performance will enhance the determination of the moderating effect environmental performance has on the relationship between CO2-emissions and financial performance.
Fourthly, the assumption was made that firms are operating in the country where their headquarters are located. In other words, the variables of awareness and regulations are attributed to that country. However, as the firms in our sample are substantially large, it is reasonable to assume they operate in multiple countries. Therefore, future research should try and find a way to be able to include regulation and awareness variables, based on the countries in which revenues are generated.
Finally, as we surprisingly found mixed results for the awareness of climate change, one could consider different proxies to take this into consideration. Also, as Google can be censored by governments, using the Google trends data has some drawbacks. Future research should either erase countries from the sample in which google is censored or find a different, more representative proxy for awareness of inhabitants. In addition, future research could also follow a similar approach as Choi et al. (2020) and investigate whether within larger countries, differences of awareness in cities or states, influence the relationship of CO2-emissions and financial performance.
Bibliography
Addoum, J. M., Ng, D. T., and Ortiz-Bobea, A. (2020). Temperature shocks and establishment sales. The Review of Financial Studies, 33(3), 1331-1366.
Alok, S., Kumar, N., and Wermers, R. (2020). Do fund managers misestimate climatic disaster risk. The Review of Financial Studies, 33(3), 1146-1183.
Alvarez, I. G. (2012). Impact of CO2 emission variation on firm performance. Business Strategy and the Environment, 21(7), 435-454.
Bos, K., and Gupta, J. (2019). Stranded assets and stranded resources: Implications for climate change mitigation and global sustainable development. Energy Research & Social Science, 56, 101215.
Claver, E., Lopez, M. D., Molina, J. F., and Tari, J. J. (2007). Environmental management and firm performance: A case study. Journal of Environmental Management, 84(4), 606-619.
Cole, M. A., Elliott, R. J., Okubo, T., and Zhou, Y. (2013). The carbon dioxide emissions of firms: A spatial analysis. Journal of Environmental Economics and Management, 65(2), 290-309.
Choi, D., Gao, Z., and Jiang, W. (2020). Attention to global warming. The Review of Financial Studies, 33(3), 1112-1145.
Fonseka, M., Rajapakse, T., and Tian, G. (2019). The effects of environmental information disclosure and energy types on the cost of equity: Evidence from the energy industry in china. Abacus, 55 (2), 362-410.
Friedman, M. (1970). The Social responsibility of business to increase its profits. New York Times Magazine, September 13, 32.
Glanemann, N., Willner, S.N. and Levermann, A (2020). Paris Climate Agreement passes the cost-benefit test. Nature Communications 11, 110.
Gonenc, H., and Scholtens, B. (2017). Environmental and financial performance of fossil fuel firms: A closer inspection of their interaction. Ecological Economics, 132, 307-328.
Hull, C. E., and Rothenberg, S. (2008). Firm performance: The interactions of corporate social performance with innovation and industry differentiation. Strategic Management Journal, 29(7), 781-789.
Kagawa, S., Suh, S., Hubacek, K., Wiedmann, T., Nansai, K., and Minx, J. (2015). CO2 emission clusters within global supply chain networks: Implications for climate change mitigation. Global Environmental Change, 35, 486-496.
Krausmann, F., Lauk, C., Haas, W., and Wiedenhofer, D. (2018). From resource extraction to outflows of wastes and emissions: The socioeconomic metabolism of the global economy, 1900–2015. Global Environmental Change, 52, 131-140.
Krausmann, F., Wiedenhofer, D., and Haberl, H. (2020). Growing stocks of buildings, infrastructures and machinery as key challenge for compliance with climate targets. Global Environmental Change, 61, 102034.
Krueger, P., Sautner, Z., and Starks, L. T. (2020). The importance of climate risks for institutional investors. The Review of Financial Studies, 33(3), 1067-1111.
Lang, L. H., & Litzenberger, R. H. (1989). Dividend announcements: Cash flow signalling vs. free cash flow hypothesis?. Journal of Financial Economics, 24(1), 181-191.
Lioui, A., & Sharma, Z. (2012). Environmental corporate social responsibility and financial performance: Disentangling direct and indirect effects. Ecological Economics, 78, 100-111.
Lee, K. H., Min, B., and Yook, K. H. (2015). The impacts of carbon (CO2) emissions and environmental research and development (R&D) investment on firm
performance. International Journal of Production Economics, 167, 1-11.
Margolis, J. D., Elfenbein, H. A., and Walsh, J. P. (2009). Does it Pay to Be Good...And Does it Matter? A Meta-Analysis of the Relationship between Corporate Social and Financial
Performance. Available at SSRN: https://ssrn.com/abstract=1866371.
Petersen, M.A., 2009. Estimating standard errors in finance panel data sets: comparing approaches. Review of Financial Studies 22, 435–480.
Ritchie, J., and Dowlatabadi, H. (2015). Divest from the carbon bubble? Reviewing the implications and limitations of fossil fuel divestment for institutional investors. The Review of Economics & Finance, 5(2), 59-80.
Rozenberg, J., Vogt-Schilb, A., and Hallegatte, S. (2020). Instrument choice and stranded assets in the transition to clean capital. Journal of Environmental Economics and Management, 100, 102183.
Sampei, Y., and Aoyagi-Usui, M. (2009). Mass-media coverage, its influence on public awareness of climate-change issues, and implications for Japan’s national campaign to reduce greenhouse gas emissions. Global Environmental Change, 19(2), 203-212.
Scholtens, B. (2008). A note on the interaction between corporate social responsibility and financial performance. Ecological Economics, 68(1-2), 46-55.
Shive, S. A., and Forster, M. M. (2020). Corporate governance and pollution externalities of public and private firms. The Review of Financial Studies, 33(3), 1296-1330.
Tvinnereim, E., Fløttum, K., Gjerstad, Ø., Johannesson, M. P., and Nordø, Å. D. (2017). Citizens’ preferences for tackling climate change. Quantitative and qualitative analyses of their freely formulated solutions. Global Environmental Change, 46, 34-41.
Van Wijnbergen, S., and Willems, T. (2015). Optimal learning on climate change: why climate skeptics should reduce emissions. Journal of Environmental Economics and Management, 70, 17-33.
Vu, H. T., Liu, Y., and Tran, D. V. (2019). Nationalizing a global phenomenon: A study of how the press in 45 countries and territories portrays climate change. Global Environmental Change, 58, 101942.
Appendix
Table 6
Correlation matrix. This table depicts the correlation between all variables used in the regression analysis.
Tobin’s Q CO2 EET Environmental Pillar score Emission regulation Awareness Global warming Awareness Climate change
Leverage Size Innovative-ness Capital Intensiveness Energy intensive Dividends Sales growth Tobin’s Q 1.0000 CO2 EET -0.2789 1.0000
Environmental Pillar score -0.0268 0.2057 1.0000
Emission regulation 0.1118 -0.2449 -0.0589 1.0000
Awareness Global warming 0.0683 -0.0674 0.0715 0.1490 1.0000
Awareness Climate change 0.1699 -0.0866 0.0401 0.3154 0.5471 1.0000
