Introduction

This research introduces ‘carbon beta’, a market-based measure of climate transition risk determined by a stock’s return sensitivity to a pollutive-minus-clean (PMC) portfolio. Unlike traditional metrics based solely on current emissions, carbon beta reflects investor expectations regarding future transition risks. It incorporates factors such as technology, management quality, innovation capacity, and financial health that affect a company’s ability to manage the shift to a low-carbon economy. For example, while Tesla, Inc. and W&T Offshore, Inc. have similar emission levels near the 80th percentile, their carbon betas diverge significantly, with Tesla in the bottom 5% and W&T Offshore in the top 5%.

Data and methodology

The study utilizes a comprehensive dataset merging US stock market data from the Center for Research in Security Prices (CRSP) with financial statements from S&P Capital IQ Compustat and greenhouse gas emissions data from S&P Trucost. The final dataset contains more than 780,000 monthly return observations for more than 7,100 unique firms, spanning the period from January 2007 to December 2020. Accounting variables are winsorized at 1% and 99% cutoffs to mitigate the effects of outliers and potential data errors.

The pollutive-minus-clean portfolio

The researchers construct a PMC portfolio as an observable proxy for carbon risk. This self-financing portfolio takes a long position in the most-polluting 30% of firms and a short position in the least-polluting 30% of firms, based on Scope 1 and 2 emissions. To ensure size neutrality, separate portfolios are formed for firms valued above and below the median NYSE firm. The PMC factor captures the return difference between polluting and cleaner firms, essentially representing the realised materialisation of climate risks.

Estimating carbon betas

Carbon betas are estimated by running time-series regressions of a firm’s daily stock returns on the PMC factor, while controlling for standard Fama-French factors (market, size, value) and the Carhart momentum factor. The estimations use a 36-month rolling window containing approximately 750 daily observations. The results show that Energy, Materials, and Utilities firms have the highest carbon betas, accounting for more than 70% of emissions. Conversely, IT, Financials, and Health Care firms exhibit the lowest carbon betas.

Summary of validation exercises

The study finds that carbon beta correlates strongly with alternative measures of climate risk, such as the Sautner et al. (2023) Climate Change Exposures (CCE) and MSCI Climate Value-at-Risk (CVaR). Regression results indicate that larger, more profitable, and more innovative firms face lower transition risk. In contrast, firms with higher capital intensity, greater leverage, and higher greenhouse gas emissions show higher carbon betas. A one-standard deviation increase in log emissions corresponds to a 0.27- to 0.51-standard deviation increase in carbon beta.

Capturing forward-looking aspects of carbon risks

Carbon beta is uniquely linked to green innovation. Within the Energy sector, a one-unit increase in ‘green share’ (the fraction of a firm’s patents that are green) reduces carbon beta by roughly 34% of a standard deviation. This suggests that markets incorporate green innovation into climate risk assessments, a pattern not observed when using simple emissions or carbon intensity metrics. Furthermore, carbon beta provides a complementary signal that overlaps with but extends beyond the inputs of existing models like MSCI’s CVaR.

Carbon beta and hedging of climate shocks

The research examines how stocks with high carbon betas perform during climate shocks. Using a Climate Policy Uncertainty (CPU) index based on Wall Street Journal articles, the study finds that a one-standard deviation increase in carbon beta corresponds to a 15-bps return decline per standard deviation shock to CPU. Similar negative patterns are observed during months with abnormally high temperatures or severe droughts, where a one-standard deviation increase in carbon beta is associated with roughly 32 bps lower monthly excess returns.

Pricing of carbon risk

The study explores whether transition risk is priced in financial markets. After correcting for errors-in-variables (EIV) bias and including industry fixed effects, the results suggest a small but significant risk premium. A standard deviation increase in carbon beta is associated with an additional 9.6 bps monthly return, or approximately 1.15% per annum. This suggests that investors demand higher expected returns for holding assets with high exposure to climate transition risk.

Conclusion

Carbon beta offers a transparent, accessible, and forward-looking measure of climate transition risk that can be applied to any asset with a sufficient return history. It enables investors to build climate-aware strategies and identify ‘climate hedge’ potential. Regulators and policymakers can also utilise carbon beta to flag firms with high transition risk and distinguish genuine green innovators from pollutive firms.