Measuring climate risk in European markets

The challenges of measuring climate risk

As climate change progresses, investors are expected to factor climate-related risks into firm valuations. However, documenting the pricing of climate risks (climate risk premia) is challenging, as the green finance literature shows conflicting results. The lack of standardized metrics for assessing firms' exposure to climate risks and the difficulty in measuring climate risks create obstacles to climate risk-informed investment decisions. Consequently, investors might struggle to identify climate-risky investments or may underestimate these risks, leading to mispricing. Such mispricing could affect the financial sector and hinder its ability to transmit climate policies effectively.

Climate regulation in the European Union

The European Union (EU) has played a leading role in global climate regulation, engaging in key international agreements like the UN Climate Convention, the Kyoto Protocol, and the Paris Agreement. Beyond these commitments, the EU has developed a distinct “self-policy”, setting ambitious goals to reduce greenhouse gas emissions by 55% by 2030 and achieve climate neutrality by 2050. This approach is supported by the European Climate Law, the “Fit for 55” package, and the Just Transition Mechanism, which collectively aim to make the EU the first net-zero emissions continent.

Our approach to studying the pricing of climate risks in European markets

The EU’s adoption of a self-climate regulation and its ambitious decarbonization roadmap create room for research on exploring the relationship between Europe, specifically, and climate change risks.

Hence, this paper studies whether climate risks, physical and transition, are priced in European markets documenting to which extent these are perceived as financial risks and whether investors consider some firms or activities as more exposed to climate risks than others. To achieve a thorough analysis, it is crucial to have comprehensive and reliable measures of these risks.

Previous studies often focus on specific aspects of physical and transition risks¹, leaving a comprehensive understanding of their overall impact largely unexplored. This broader perspective becomes essential for making informed climate investment decisions and managing climate risks effectively. Through text analysis, we therefore propose novel climate risk vocabularies that improve upon existing literature, providing sophisticated climate indicators that uncover new financial implications of climate change.

Building climate risk vocabularies

Our approach starts by compiling authoritative texts on climate change to create two key documents: the Physical Risk Document (PRD) and the Transition Risk Document (TRD). These documents are processed using the term frequency-inverse document frequency (tf-idf) method. This method allows us to build climate risk vocabularies where terms are ranked from the most to least representative of climate risks.

Figure 1 shows the most relevant terms of the global climate risk vocabularies as word clouds, where each term size is proportional to its term relevance. The physical risk vocabulary includes multiple dimensions of physical risk such as extreme and chronic hazards, whereas the transition risk vocabulary includes various aspects of this climate risk such as technological advances and environmental policies. Terms like ‘ecosystems’, ‘sea level rise’, and ‘precipitation’ are representative of the physical risk topic, while terms like ‘hydrofluorocarbon’ (HFC), ‘bioenergy’, and ‘greenhouse gas’ (GHG) are representative of the transition risk topic.

Figure 1. The most relevant terms of physical and transition climate risk vocabularies, illustrated as word clouds

Measuring climate risk in daily news content

Next, we gather news articles from Reuters News starting in 2005 and apply a similar tf-idf method. We then calculate the “cosine similarity” between these news and the vocabularies, which allows us to measure how closely the daily news content relates to physical and transition risks. This method generates time series data, reflecting the daily focus on these risks (concern series).

The final outputs are the Physical Risk Index (PRI) and the Transition Risk Index (TRI) which are calculated as residuals to Autoregressive processes of order 1 of the concern series and capture unexpected changes or shocks in climate risk concerns, offering insights into how these risks are perceived over time.

Figure 2 shows the scatter plots of daily physical and transition media concerns, along with major PRI and TRI topics (vertical bars). PRI captures both acute physical risks such as floods, or extreme weather events, and chronic risks such as permafrost thawing, droughts, and sea level rise, as well as governments and institutions’ calls for climate adaptation actions and other adverse impacts of physical risk on, e.g. the ecosystem such as biodiversity loss.

Figure 2. Scatter plots of daily climate risk concerns in the media

News on regulations and measures to curb GHG emissions generate spikes in the TRI, e.g. news regarding the EU carbon reform deal or the Montreal Protocol, as well as news concerning the costs associated with the transition or the advances of technological innovation and renewable energies to reach, e.g. net-zero emissions targets.

Assessing climate risk premiums in European stock returns

Standard asset pricing theories suggest that assets serving as hedges against future market changes typically have higher prices and lower expected returns. Given that climate change risks affect future investment and consumption opportunities, investors may prefer holding assets that perform well under increasing climate risks, even if this means accepting lower returns.

To assess climate risk premia in European stock returns, we use a portfolio sorting approach. We perform time-series regressions of equity returns on climate risk indices, controlling for standard risk factors. Transition and physical risk betas indicate firm-level exposure to climate risks, allowing us to sort companies from low (negative) to high (positive) exposure and create portfolios. Low beta firms underperform when climate risks rise, while high beta firms perform well. A low-minus-high beta portfolio should generate positive excess returns if a climate risk premium exists, reflecting investors' compensation for bearing climate risks.

We also perform firm-level and sectoral-level analyses to investigate which information investors use to detect climate-risky stocks. Firms are sorted by GHG emissions, environmental scores, and ESG scores, while sectors are grouped by industry classification, allowing us to examine equity sensitivity to climate risks at both levels.

Finally, we examine country-level stock-market sensitivities to climate risks across Europe to identify potential regional differences in the stock market’s response to climate risks. This helps us gain insights into the significance of sub-national factors in shaping investment strategies and policy design.

Key findings

Climate-related risks, transition and physical, are increasingly being priced in European financial markets.

Economically significant risk premia have emerged for both transition and physical risks, particularly after 2015. Investors are increasingly demanding compensation for exposure to these risks, as evidenced by the higher returns required for holding stocks negatively associated with climate risks.

Green and brown stocks respond differently to heightened climate risks depending on the climate risk type, physical or transition.

Rises in transition risk typically increase the return of green stocks, whereas rises in physical risk typically decrease the return of brown stocks.

Investors screen climate-risky stocks by relying on firm-level or sectoral information differently, depending on the type of climate risk—physical or transition.

Firm-level information appears to be used by investors as a gauge for firms' exposure to climate risks, particularly for transition risk and since 2015, while sector classification is more effective for identifying exposure to physical risks.

Country-level stock markets are increasingly connected with climate risks.

European stock markets and climate risk connection have intensified post-2015, yet with some heterogeneity emphasising the need for national/sub-national considerations in investment and climate policies.

Implications for investors and policymakers

The implications of these findings are significant for both investors and policymakers.

For investors, the study highlights the importance of integrating climate risk considerations into their investment strategies.

For policymakers, the research underscores the need to ensure that financial markets accurately reflect the risks associated with climate change. Policymakers may need to consider additional measures, such as mandatory climate risk disclosures, to enhance market transparency and ensure that investors have the information they need to make informed decisions.

Our novel climate risk vocabularies and indicators find several applications in identifying, measuring, and studying climate risks.

View the publicly accessible data in this study on the Economic Policy Uncertainty website page:
Climate Risk Indexes

Read the full paper: Transition versus physical climate risk pricing in European financial markets: a text-based approach

¹ Physical risks arise from climate change-related hazards, such as floods or rising sea levels, while transition risks stem from the shift toward a greener economy, driven by technological advances, climate policies, and changing public preferences.

Lavinia Rognone is a lecturer in Sustainable Finance.