The Carbon Premium: How Climate Risk Is Repricing Markets

Key Takeaway

Many investors assume that ESG-friendly portfolios automatically sacrifice returns, or conversely, that climate risk is already fully priced. The evidence suggests neither view is correct. Bolton and Kacperczyk (2021) document a carbon premium of approximately 1-2% per year: firms with higher carbon emissions earn higher stock returns, consistent with investors demanding compensation for transition risk exposure. Yet this premium coexists with deep measurement problems. Berg, Koelbel, and Rigobon (2022) show that ESG ratings from different providers diverge so severely that using them as climate risk proxies is unreliable. Meanwhile, Hsu, Li, and Tsou (2023) find a distinct pollution premium driven by environmental regulatory risk. The practical challenge is to distinguish between transition risk, physical risk, and regulatory risk, each of which affects portfolio construction differently.

The Carbon Premium: Compensation for Transition Risk

The central question in climate finance is whether investors demand higher returns from carbon-intensive firms. If carbon emissions create financial risk through potential regulation, carbon taxes, or stranded assets, then rational investors should require compensation for bearing that risk. Bolton and Kacperczyk (2021) test this hypothesis systematically.

The Bolton-Kacperczyk Framework

Using comprehensive emissions data from the CDP (formerly Carbon Disclosure Project) matched to U.S. equity returns, Bolton and Kacperczyk examine whether carbon emissions predict cross-sectional stock returns after controlling for standard risk factors. They decompose emissions into three categories: Scope 1 (direct emissions from owned operations), Scope 2 (indirect emissions from purchased energy), and Scope 3 (all other indirect emissions across the value chain).

Their central finding is striking. Firms with higher total carbon emissions earn significantly higher stock returns, with the effect concentrated in Scope 1 (direct) emissions. The carbon premium is approximately 1-2% per year, robust to controlling for the Fama-French five factors, industry fixed effects, and a battery of firm-level characteristics.

Why Scope 1 Drives the Premium

The concentration of the premium in Scope 1 emissions is economically intuitive. Direct emissions are the most visible to regulators and the most likely target of carbon pricing mechanisms. A coal-fired power plant faces stranded asset risk from emissions regulations in a way that a software company with indirect Scope 3 supply chain emissions does not. The market appears to price this distinction correctly: the premium compensates for the most salient and measurable form of climate transition risk.

The Premium Has Grown Over Time

Bolton and Kacperczyk document that the carbon premium has strengthened over their sample period, consistent with growing investor awareness of climate transition risk. As climate policy has moved from theoretical possibility to concrete implementation (the EU Emissions Trading System, national carbon taxes, the Paris Agreement framework), the market has increasingly priced the risk that carbon-intensive business models face regulatory disruption.

This temporal pattern distinguishes the carbon premium from a spurious correlation. If the premium reflected an omitted risk factor unrelated to climate, there would be no reason for it to increase as climate policy became more salient.

Physical Risk versus Transition Risk

Giglio, Kelly, and Stroebel (2021) draw a critical distinction between two channels through which climate change affects asset prices.

Transition risk arises from the policy and technology shifts required to move toward a low-carbon economy. Carbon taxes, emissions regulations, shifts in consumer preferences, and technological disruption of fossil fuel industries all create transition risk. This is the risk that Bolton and Kacperczyk primarily measure.

Physical risk arises from the direct consequences of climate change: rising sea levels, extreme weather events, water scarcity, and ecosystem disruption. Physical risk operates on longer time horizons and affects asset values through damage to physical capital, supply chain disruptions, and changes in agricultural productivity.

The distinction matters for portfolio construction because the two risks have different geographic, sectoral, and temporal profiles. A real estate portfolio concentrated in coastal Florida faces physical risk that has little overlap with the transition risk facing a European utility company. Hedging one does not hedge the other.

Giglio, Kelly, and Stroebel note that physical risk is particularly difficult to price because the time horizons involved (decades to centuries) exceed the typical investment horizon and because the probability distributions of climate outcomes involve deep uncertainty rather than quantifiable risk. This creates the possibility that physical risk is systematically underpriced, a hypothesis with concerning implications for long-term investors.

The Pollution Premium: Regulatory Risk in Action

Hsu, Li, and Tsou (2023) identify a related but distinct premium. Firms with higher toxic emissions (measured by the EPA's Toxics Release Inventory) earn higher stock returns, and this pollution premium is concentrated in firms located in states with stricter environmental regulations.

Mechanism: Regulatory Exposure

The Hsu-Li-Tsou finding suggests that the pollution premium is driven by regulatory risk rather than a broad societal concern about environmental impact. Firms in heavily regulated states face higher compliance costs, greater litigation risk, and a higher probability of operational disruptions from regulatory enforcement. Investors demand compensation for these tangible financial risks.

The pollution premium is distinct from the Bolton-Kacperczyk carbon premium in two important ways. First, it is driven by toxic emissions (chemicals, heavy metals) rather than greenhouse gas emissions. Second, it is modulated by the local regulatory environment rather than global climate policy. This distinction matters because it implies that environmental risk premia are not monolithic; different types of environmental exposure create different types of financial risk.

The Regulatory Channel Creates Predictability

Hsu, Li, and Tsou show that changes in environmental regulation predict changes in the pollution premium. When states tighten environmental standards, the pollution premium for firms in those states increases. This regulatory predictability offers a potential edge for investors who can anticipate the direction of environmental policy.

ESG Ratings: The Measurement Problem

Any attempt to construct a climate-risk-aware portfolio confronts a fundamental measurement challenge. Berg, Koelbel, and Rigobon (2022) document that ESG ratings from six major providers (MSCI, Sustainalytics, Moody's, S&P Global, Refinitiv, and KLD) show correlations as low as 0.38 to 0.71. By comparison, credit ratings from Moody's and S&P agree approximately 99% of the time.

Sources of Disagreement

Berg, Koelbel, and Rigobon decompose the disagreement into three components. Scope divergence accounts for different rating agencies including different ESG categories in their assessments. Measurement divergence arises from agencies using different indicators to measure the same concept. Weight divergence reflects different weights assigned to common categories.

Measurement divergence is the dominant source of disagreement, accounting for more than half of the total rating spread. This means that even when rating agencies look at the same ESG dimension, they reach different conclusions about how a firm performs on that dimension.

Implications for Climate Risk

The measurement problem is particularly acute for climate risk because carbon emissions, the most objective and quantifiable environmental metric, account for only a small fraction of the overall ESG score. An investor who uses an ESG score as a climate risk proxy is relying on a signal that is heavily contaminated by subjective assessments of governance quality, labor practices, and other dimensions that have little to do with climate exposure.

This explains why naive long-green/short-brown strategies based on ESG scores produce unstable returns. The signal-to-noise ratio is too low to generate consistent alpha, and the disagreement across providers means that different implementations of the same strategy can produce contradictory results.

The NGFS Scenario Framework

The Network for Greening the Financial System (NGFS), a consortium of central banks and financial supervisors, has developed a set of standardized climate scenarios that are increasingly used for financial stress testing and long-term risk assessment.

The NGFS scenarios span three broad pathways. Orderly transition scenarios assume that climate policies are introduced early and gradually tightened, allowing a smooth transition to a low-carbon economy. Disorderly transition scenarios assume delayed policy action followed by abrupt tightening, creating sharp adjustment costs. Hot house world scenarios assume limited or no policy action, resulting in severe physical risks from unmitigated climate change.

Each pathway generates distinct implications for asset prices. Orderly transitions favor firms that can adapt gradually and penalize those with rigid carbon-intensive business models. Disorderly transitions create tail risks for both carbon-intensive firms (through sudden policy shocks) and clean energy firms (through demand volatility). Hot house scenarios primarily affect assets exposed to physical risk, including real estate, agriculture, and infrastructure in vulnerable regions.

Financial institutions are increasingly required to conduct climate stress tests using NGFS scenarios. The ECB completed its first economy-wide climate stress test in 2022, and the Bank of England and Federal Reserve have followed with their own exercises. These regulatory stress tests are translating climate scenarios into concrete capital requirements, creating a new channel through which climate risk affects asset valuations.

Constructing a Carbon-Hedged Portfolio

The research literature points toward several practical approaches for investors who want to manage carbon exposure systematically.

The Carbon Factor: Long-Short Construction

Following the logic of Bolton and Kacperczyk, a carbon factor can be constructed as a long-short portfolio: long stocks with high carbon emissions (which earn the carbon premium) and short stocks with low carbon emissions. This factor behaves like other systematic risk factors in asset pricing; it earns a positive premium on average but introduces volatility linked to climate policy developments.

An investor who wants to harvest the carbon premium would tilt toward high-emission stocks, accepting the transition risk. An investor who wants to hedge carbon risk would take the opposite position, accepting lower expected returns in exchange for protection against regulatory shocks.

Carbon Intensity versus Absolute Emissions

Portfolio construction requires choosing between absolute emissions (total tons of CO2) and carbon intensity (emissions per unit of revenue or market capitalization). Bolton and Kacperczyk find that both measures predict returns, but carbon intensity is more useful for portfolio construction because it is size-neutral. A large company with high absolute emissions but low carbon intensity per dollar of revenue may face less transition risk than a small company with lower absolute emissions but high intensity.

The Hedging Challenge

Engle, Giglio, Kelly, Lee, and Stroebel (2020) develop a method for hedging climate change risk using textual analysis of climate news. They construct a climate news index from media coverage and build portfolios that hedge against innovations in this index. Their approach illustrates a key challenge: because climate risk unfolds over decades, traditional short-horizon hedging instruments are poorly suited to climate exposure. Longer-horizon approaches, including direct investment in climate-resilient assets and strategic sector allocation, may be more effective than derivative-based hedging for managing climate portfolio risk.

What Remains Unresolved

Several important questions remain open. First, will the carbon premium persist as climate risk becomes more widely recognized and priced? The logic of risk factor pricing suggests yes, as long as the underlying risk is real. But if carbon-intensive industries shrink and transition risk is fully priced in, the premium could diminish or reverse.

Second, how should investors weight physical risk against transition risk? Current pricing appears to reflect transition risk far more than physical risk, which may create long-term mispricing opportunities for investors with sufficiently long horizons.

Third, can ESG ratings be reformed to provide a reliable climate signal? The Berg-Koelbel-Rigobon findings suggest that the current ESG rating infrastructure is not fit for purpose as a climate risk measure. Direct emissions data, combined with forward-looking scenario analysis, may provide a more reliable foundation for climate-aware investing.

The evidence across these papers suggests that climate risk is real, priced, and growing in importance, but that the tools available for measuring and managing it remain imperfect. Investors who understand the distinction between transition risk, physical risk, and regulatory risk, and who rely on direct emissions data rather than composite ESG scores, are better positioned to navigate a financial system that is still in the early stages of pricing climate change.