Cet article appartient au dossier : ESCP Europe Applied Research Papers 13.

Investor sentiment

Studying the Impact of Greenhouse Gas Emissions on Company Energy Valuation in Europe

L’étude du sentiment des investisseurs concernant les émissions de Gaz à effet de serre (GES) de compagnies énergétiques, à travers l’analyse de deux panels d’entreprises européennes, des entreprises d’énergies renouvelables et des entreprises d’énergies fossiles, entre 2013 et 2017, permet d'évaluer si une plus faible intensité d’émission de GES est valorisée par les investisseurs.

Tableaux 1 et 2


Pour en savoir plus

  • Tableaux 3 et 4

    Tableaux 3 et 4

  • Tab5


  • Tab6


Revue de l'article

Cet article est extrait de
Banque & Stratégie n°386

ESCP Europe Applied Research Papers 13

This article studies the relationship between energy companies’ green house gas (GHG) emissions and their relative valuation on the stock market in Europe. Indeed, when it comes to compare renewable energies (RE) companies to fossil fuel energy (FFE) companies, it appears that investors have mixed feelings. On one hand, institutional investors show growing concern over the green impact of their investments. There is a clear will to at least divest from fossil fuels, if not to promote sustainable companies and renewable energy. On the other hand, the risk-return profile of renewable energy (RE) companies differ from that of fossil fuel energy (FFE) companies, the first being riskier than the second because of unproven technologies and business models taking time to reach maturity. If indeed investors consider RE companies as being riskier, then those stocks might be less appealing on the market than those of FFE companies, because of investors risk aversion. In short, investors have to face a trade-off between sustainability and returns in the domain of energy companies in Europe. This article tries to shed a light on investor sentiment in the energy sector over the past years in Europe.

In the sector of energy companies, are investors sensitive to greener companies?

According to the Inventory of U.S. Greenhouse Gas Emissions, manmade carbon dioxide (CO2) emissions represented 82% of greenhouse gas emissions in 2017 in the US, while global energy production represented 72% of CO2 emissions. Therefore, energy production can be considered as the major source of greenhouse gas (GHG) emissions around the world. To limit global warming, a reduction of GHG emissions should therefore be linked to energy production.

Institutional investors hold a specific position to drive a change towards adoption of clean energy. They are the drivers of capital flows, and their investments could definitely limit global warming. This view is supported by Masini & Menichetti (2013), developing the idea that renewable energy technologies and business models not being mature, investors are driven away. This lack of funding limits innovation and improvements in renewable energy efficiency over traditional (i.e. fossil fuel) sources.

However, the sector has evolved over the past years. Business models have matured, and renewable energies are reaching grid parity, meaning that their price (LCOE or Levelized Cost of Electricity) is equal or lower than that of conventional energy sources. Renewable energies are therefore getting competitive and subsidies might be gradually removed by governments. A Deloitte Insights report (Rithu et al., 2018) even states that “having only recently been recognized as a ‘mainstream’ energy source, renewable energy is now rapidly becoming a preferred one. A powerful combination of enabling trends and demand trends – evident in multiple developed and developing nations globally – is helping solar and wind compete on par with conventional sources and win”.

Our study builds on the work of Henriques and Sadorsky (2018), who have investigated the possibility for institutional investors to divest from fossil fuels. They compared three portfolios of US stocks that allow for short selling: one containing fossil fuel companies and utilities, the second one containing clean energy companies and no fossil fuel companies nor utilities, and the third one (control portfolio) containing no fossil fuel companies, no clean energy companies, and no utilities. The authors have found that the clean energy portfolio offers the best risk-adjusted returns. They argue that it seems possible to divest from fossil fuel energy companies without losing performance. They also establish that investors are willing to pay higher fees to divest from fossil fuel energy companies towards clean energy companies, transaction costs being included. This points towards an evolution of the clean energy sector, which can be seen as more attractive to investors, both financially and morally. Indeed, the latest research in socially responsible investments indicates that renewable energy (RE) companies should be over-valued because of high investor demand.

In order to test investor’s commitment on environmental goals in the energy sector, we chose to have a look at the relative valuation of renewable energy (RE) and fossil fuel energy (FFE) companies on the European stock market. If RE companies trade at a premium relative to FFE companies, it might indicate that they are experiencing higher demand on the market, just because of their environmental sustainability. If, on the contrary, there is no such premium, we could infer that investors do not take sustainability into account when making their investment decisions.

This paper investigates two questions:

  • On European capital markets, is there a significant relative valuation difference between renewable energy (RE) companies and fossil fuel energy (FFE) companies?
  • If there is a significant difference in company valuation, to what extent does this difference come from the degree of environmental sustainability of a company?

A comparison of two samples of European energy companies, 2013-2017

This study uses data from Bloomberg for two samples of European renewable energy (RE) and fossil fuel energy (FFE) companies. Our sample period runs for 5 years (2013 to 2017) to try to assess the current maturity of the RE sectors while retaining the longest possible sample period.

The final sample for RE companies is composed of 10 companies, as exhibited in Table 1, with the following procedure. To construct this sample, we first took the Société Générale’s ERIX index, as of December 31, 2018. As described by Société Générale: “The ERIX index selects the largest [european] companies in the areas of renewable energy such as wind, solar, biomass and water energy. […] The index members are the 10 largest and most liquid stocks from the list of eligible companies.” From this list, we searched in Bloomberg for additional comparable RE companies in Europe, the final result ending in a list of 15 European companies. Then, we excluded the following: companies which had their IPO after January 2013 (Neoen, Scatec Solar, Senvion); companies that did not report information for one year or more within the period because of a merger or an acquisition (Siemens-Gamesa); companies that have been experiencing financial distress for over 4 years (Meyer Burger Technology). The final sample consists in 10 european RE companies, as presented in Table 1.

The sample for FFE companies is composed of the 14 companies listed in Table 2. Identification of these companies was based on the MSCI Europe Energy Index as of December 31, 2018. As quoted by MSCI, “The MSCI Europe Energy Index is designed to capture the large and mid-cap segments across 15 Developed Markets (DM) countries in Europe […] classified in the Energy sector.” From this list, we searched in Bloomberg for additional comparable FFE companies in Europe, the final result ending in a list of 18 European companies. Then, we reduced the sample with the following indications: Royal Dutch Shell being mentioned twice for their A and B shares, we counted this company only once; companies with a significant pro-environment image were removed to avoid confusion (Neste Corporation); companies whose indicators were not fit within the period (Surgutneftegas had a negative Enterprise Value from 2013 to 2016). The final sample consists in 14 european FFE companies, as presented in Table 2.

For both samples, yearly data was gathered directly from Bloomberg with the following variables: Tobin’s Q, GHG or CO2 emissions (according to what the company reports), Beta of the stock, Leverage ratio, Total Balance Sheet Assets, Sales. As regards hypothesis 1, we only needed the Tobin’s Q variable, which will be presented in the next paragraph. The remaining variables are being used later on as control variables in our regression, so we discuss them when testing hypothesis 2.

Tobin’s Q is defined as the ratio of the market value of a firm over the replacement value of its assets. If Tobin’s Q is equal to 1, it means the market value reflects only the recorded assets of a company. If Tobin’s Q is superior to 1, it means that there are unrecorded assets or investment opportunities that the market takes into account when pricing the company. Companies with a high growth potential should therefore have a high Tobin’s Q. In this thesis, Tobin’s Q ratio has been preferred as an indicator of firm relative valuation instead of Book-to-Market ratio. Indeed, Tobin’s Q is computed using the current replacement value of assets which we feel is more consistent to use than an accounting value.

Is there a stronger demand for the stocks of renewable energy companies?

The first hypothesis to be tested is the following:

  • Companies in the renewable energy (RE) sector have a higher Tobin’s Q than comparable companies in the fossil fuel energy (FFE) sector over the past 5 years.

Table 3 shows descriptive statistics for both samples on the 2013-2017 period. The RE sample is of 50 observations (10 companies over 5 years) and the FFE sample of 70 observations (14 companies over 5 years).

As can be seen in Table 3, RE companies show a higher Tobin’s Q ratio than FFE companies both in terms of mean and median. However, the median analysis only reveals a 0,04 difference between the two samples, which could mean that the high average Tobin’s Q in the RE companies might come from only a handful of very high values.

To make sure the difference of the means is significant, we first perform a Levene’s test on the two samples to assess potential differences in variance. This Levene’s test is run as an equivalent single factor ANOVA (analysis of variance). The ANOVA is run between two groups, the first group being composed of 50 measures (difference between a single company Tobin’s Q and the group mean in the RE sector); and the second group being composed of 70 measures (difference between a single company Tobin’s Q and the group mean in the FFE sector).

As can be seen in Table 4, the p-value found for this test is equal to 0,000 which means that the variances of our two samples are different at a very significant level. That difference being ascertained, we now perform Welch’s T-test to assess whether the mean difference of Tobin’s Q among the two samples is significant (see Table 5).

Table 5 shows that the p-value for Welch’s test is equal to 0,01 which is considered very significant. Therefore, we can conclude that the difference between the Tobin’s Q of both samples is statistically significant. Hypothesis 1 is therefore confirmed: over the period 2013-2017, renewable energy companies have on average a significantly higher Tobin’s Q ratio than fossil fuel energy companies. In other words, this validates the theory that the renewable energy (RE) sector has matured since 2014, and that investors have a higher demand for RE companies’ stocks, which drives to higher prices and higher valuations, comparatively to fossil fuel companies (FFE). We can infer that, within this sector, investors probably divest from FFE companies in order to translate their capital flows towards RE companies.

Do lower green house gas emissions explain the preference for renewable energy companies?

Our second hypothesis to be tested is the following:

  • Green house gas (GHG) emissions have a negative impact on relative company valuations (measured by Tobin’s Q) in the energy sector over the past 5 years.

In this approach, both samples are grouped together. The aim is to assess the impact of company activity, with respect to environmental friendliness, on the difference in valuation. The fact that companies can be more or less environmental friendly is not assessed here by attributing dummy variables to each company according to their activity sector, but by evaluating green house gas (GHG) emissions intensity per company. If investors are taking environmental friendliness into account when making their investment decision, then the demand should be higher for companies with lower GHG emissions intensity.

To test this hypothesis, a linear multivariate regression is performed, with the dependent variable being the firm’s over- or under-valuation measured by Tobin’s Q. The independent variable is the firm’s emission intensity (GHG emissions)/Sales, per year). Emissions intensity indeed provides us with information regarding environmental sustainability of firms, and standardizing GHG emissions by sales allows us to better compare across firms.

Some companies do not report GHG emissions per year. Whenever possible, the following methodology was applied to approach the value of their GHG emissions intensity:

  • obtain energy production (usually in kWh) for each energy production technology (wind, solar, etc.) for the last year, from the annual report;
  • apply the matching ratio of gCO2-eq/kWh (GHG emissions per kWh expressed in equivalent of grams of CO2) obtained from the annual report of the company, or, if unavailable, from the IPCC report on climate change (2014);
  • compute the resulting GHG emissions intensity for the last year;
  • if GHG emissions are not available for previous years (2013 to 2016), the emissions intensity ratio for the last year is assumed to be equal to the older known emissions intensity ratio.

As the corresponding data was unavailable for Lukoil and Tatneft, those two companies were removed from the sample to test hypothesis 2.

Control variables in this regression are size, growth, unlevered beta and financial leverage. Those variables are often used as determinants of Tobin’s Q, because of their expected explanatory power. They help assess company risk and opportunities and might affect the dependent variable. Therefore, by adding those control variables, we can measure the marginal contribution of green house gas (GHG) emissions to the explanation of the Tobin’s Q of a given company.

Size is measured as the ln (natural logarithm) of balance sheet total assets, as is customary in papers considering determinants of Tobin’s Q. Growth is measured as growth of sales (current year’s sales minus previous year sales divided by previous year’s sales). Unlevered beta is calculated from Bloomberg’s raw equity beta by removing the effect of financial leverage.

As a summary, in order to test hypothesis 2, we perform a multi-regression analysis on the companies’ Tobin’s Q relative to their yearly GHG emissions intensity, while controlling for 4 typical explanatory variables of Tobin’s Q: size, leverage, growth, and unlevered beta.

As a result, green house gas emissions contribute to the difference in valuation

The sample used for the regression is composed of the reunion of the RE companies and FFE companies samples. There is a total of 110 observations since Lukoil and Tatneft have been removed from the sample (see the part on GHG emission estimates). Results of the multivariate linear regression are presented in Table 6.

As expected, GHG emissions intensity has a negative impact on Tobin’s Q ratio in the energy sector, with a high significance, at the 1% level. The analysis also shows that GHG emissions are the second-most important determinant of Tobin’s Q. Indeed, if we focus on the standardized beta coefficient, we can see that the most influential variables are, in order, Leverage, then GHG emissions intensity (ex aequo with Size), then beta and growth (though the last one is not significant).

Our regression indicates a high R2 value of 0,92, implicating that 92% of the variations of Tobin’s Q can be explained using the significant variables. Our analysis will focus most on GHG emissions, since the other variables are classic explanatory indicators of Tobin’s Q.

GHG emissions intensity is negatively correlated with Tobin’s Q ratio, as expected in hypothesis 2. This means that the relative over-valuation of RE companies over FFE companies (as noted in hypothesis 1) can be partially explained by their lower GHG emissions intensity. We can infer that investors allocate their funds in energy companies depending on how environmentally sustainable they are. Our results are consistent with hypothesis 1 and, as previously mentioned, with Henriques and Sadorsky (2018), namely, that the RE sector has matured since 2014 and that high demand for RE companies’ stocks drives prices higher.


This study shows that companies with a lower green house gas (GHG) emissions intensity, such as renewable energy companies, are valued by public investors with a premium over less environmentally friendly companies. These results are consistent with other works on energy transition, and the trend could strengthened over the next years. Indeed, various technologies are getting mature and reaching grid-parity, adoption is growing all over Europe and investors provide increased capital, which helps foster innovation and increase efficiency. However, an increase in the relative valuation of those companies would also mean that it costs more to invest in those, therefore reducing the potential for a future return. This could partly explain the lower returns observed on those renewable energy companies, as compared to fossil fuel companies. Further research might be necessary to assess the stability of those results over time and across countries. Finally, alternative measures of valuation might be considered, alongside Tobin’s Q.


Sommaire du dossier

ESCP Europe Applied Research Papers 13

Articles du(des) même(s) auteur(s)

Sur le même sujet