Introduction

The report examines how long-term climate stress tests and scenario analyses should account for changes in portfolio composition over extended horizons, often reaching 2050 or later. It identifies a key methodological issue: whether portfolios should be modelled as static or allowed to evolve as economies transition. The paper notes that most current exercises rely on static assumptions, despite their limitations, and aims to clarify available alternatives, their trade-offs, and appropriate use cases.

Static portfolio approach

The static portfolio approach assumes that portfolio composition remains unchanged throughout the stress-test horizon. This method is widely applied because it is simple to implement and aligns with traditional short-term stress-testing practices. Its primary value lies in illustrating the consequences of inaction, as outcomes are driven solely by exogenous shocks. However, the report highlights that this assumption is unrealistic over long periods and may obscure sensitivities related to institutional responses. Front-loading climate shocks can also distort risk dynamics, limiting realism for long-term analysis.

Macro adjustment approach

The macro adjustment approach allows portfolios to evolve in line with a non-shocked macroeconomic baseline, such as sectoral expansion or contraction. Adjustments may be based on generic macro trends or on company-level plans and targets. This approach aims to reflect changing capital needs before shocks occur and is less disruptive than other dynamic options. The report notes, however, that calibration is complex and that this method cannot capture heterogeneity in firm or financial institution behaviour. As a result, it may reflect broad market conditions while oversimplifying actual responses.

Ex ante adjustment approach

Under the ex ante approach, portfolio changes are governed by predefined calculation rules derived from forward-looking commitments, policies, or benchmark requirements. Examples include coal divestment strategies or alignment with EU Climate Benchmark decarbonisation rates of around 7% per year. Because rules are set before shocks occur, this approach is more structured and harder to manipulate. However, the report cautions that outcomes depend heavily on the credibility and consistency of targets across institutions, and that ex ante rules are not designed to respond to specific shock dynamics.

Ex post adjustment approach

The ex post approach permits portfolio adjustments during the stress test in response to realised losses or identified shocks. This introduces a behavioural element, allowing institutions to model strategic responses over time. The report notes that this approach can support internal strategy development and help analyse interactions between shocks and responses. However, it is difficult to constrain meaningfully and carries a high risk of gaming. Inconsistent responses across institutions may also complicate interpretation, particularly when assessing system-wide implications.

Simulation of losses using different approaches

A simplified simulation illustrates how stress-test results vary significantly depending on portfolio assumptions. Applying a high-carbon negative shock and a low-carbon positive shock in 2025 and 2030 shows that static portfolios generate higher losses as time horizons extend. Dynamic approaches, including macro adjustments, index-based rules, and strategic responses, reduce losses to varying degrees. The results demonstrate that portfolio composition assumptions materially influence outcomes, especially in long-term analyses.

Conclusion

The report concludes that no single portfolio approach is universally appropriate. Static portfolios are useful for highlighting inaction, while macro, ex ante, and ex post approaches better reflect transition dynamics, commitments, or strategic behaviour. The choice of method should be determined by the objective of the stress test. Dynamic approaches remain at an early stage and require further development, testing, and supervisory clarity.