This article was edited on June 6, 2021, to correct a few errors: the data sample was from 2010 to 2016; brown firms did not outperform green firms; and a bullet point was added about the strong contrast between brown and green stock performance.

New research shows that, since 2010, “green” stocks – those of companies with a low carbon footprint – have outperformed “brown” stocks. That may have been caused by increased demand from investors pursuing an environmental, social and governance (ESG) mandate, which means the effect is temporary and brown stocks now have higher expected returns.

The evidence of rising temperatures and the renewed policy efforts to curb carbon dioxide (CO2) emissions raises the question of whether those emissions represent a material risk for investors that is reflected in the cross-section of stock returns. This important question has led to a heightened interest in the subject from researchers and a flurry of new papers.

Maximilian Görgen, Andrea Jacob, Martin Nerlinger, Ryan Riordan, Martin Rohleder and Marco Wilkens contribute to the literature on the impact of carbon risk on asset pricing with their June 2019 study, “Carbon Risk.” They quantified carbon risk via a “Brown-Minus-Green factor” (BMG) derived from 1,600 firms with data from four major ESG databases. This factor allowed the estimation of carbon risk, using what they defined as “carbon beta,” which is the sensitivity of stock prices to changes in carbon-related emissions. Their Brown-Green Score (BGS) is a composite measure of three indicators designed to separately capture the sensitivity to carbon risk of firms’ “value chains” (e.g., current emissions), “public perception” (e.g., response to perceived emissions) and “adaptability” (e.g., mitigation strategies). Due to limited data availability, a problem with all ESG-related research, their data sample covered the relatively short period 2010 to 2016. Following is a summary of their findings:

  • Firms are becoming greener – mostly driven by green firms becoming significantly greener than brown firms. For instance, green firms reduced their average carbon intensity by roughly 16% annually versus roughly 2% annually for brown firms.
  • The BMG factor significantly increases the explanatory power of common asset pricing models, suggesting it is important in explaining variation in global equity prices.
  • Firms performed worse if they surprised markets by becoming browner (their BGS score rises) compared to the previous year.
  • Firms investing in innovation and clean technology, proxied by R&D expenditures, had lower carbon betas, while firms with dirty or “stranded” assets, proxied by property, plant and equipment assets, had higher carbon betas.
  • The carbon risk of the financial industry is strongly related to the carbon risk of the domestic firms they are likely to finance.
  • There was a strong contrast in the performance of the brown and the green portfolio over time. While the cumulative return of BMG was slightly positive in the period from 2010 to the end of 2012, the effect reversed in the following period and over the full period brown firms performed worse than green firms on average during our sample period.
  • The cumulative difference in returns between brown and green firms was roughly 14%, with green firms outperforming. This is consistent with increased investor focus on tackling climate change, which has led to increased demand for the stocks of green firms and reduced demand for the stocks of brown firms – the increased demand explains why green firms outperformed brown firms, which is inconsistent with economic theory.
  • Investors can achieve comparable expected returns and Sharpe ratios for their portfolios with similar exposures to other systematic risk (such as beta, size and value) factors or to specific industries while reducing carbon beta via a “best-in-class” approach – demonstrating that investors can achieve their sustainable investing goals without sacrificing returns by tilting their portfolios to companies with good scores but with similar exposure to other common factors.
  • Carbon betas are high and positive in countries like South Africa, Brazil and Canada, which means those countries will likely be negatively affected if the world speeds up the transition to a low-carbon economy. In contrast, average carbon betas are negative in European countries and Japan. On an industry level, tech firms have carbon betas near zero on average, while basic material and energy firms have the highest positive carbon betas, as expected. There are, however, significant differences in carbon betas within industries, suggesting that carbon risk is not simply a proxy for the risk associated with certain industries.