Introduction: the legibility problem

Adaptation and resilience are terms used extensively across sustainable finance regulation, physical risk management, and built environment practice. However, they are not applied consistently. This paper focuses on the boundary between adaptation and resilience, where fragmentation is sharpest. The consequences of this ambiguity are tangible: transactions often fail because counterparties cannot price risk due to conflicting definitions. Additionally, policy instruments may misfire if they do not align with the way financial institutions underwrite assets. In 2025, the Financial Stability Board highlighted systemic risks originating from the real estate sector due to increasing climate hazards, emphasising the urgent need for more effective adaptation finance to address these risks.

1. The anatomy of the disagreement

The disagreement between adaptation and resilience is structural. In systems ecology, resilience is associated with thresholds between alternative regimes and is viewed as a property of systems rather than individual assets. Conversely, in risk management, resilience is typically treated as recovery speed, which is an outcome measure rather than a threshold measure. Three authoritative regulatory frameworks currently produce incompatible boundary lines: the EU Taxonomy, the Climate Bonds Initiative’s (CBI/UNDRR) Resilience Taxonomy, and the United Nations Environment Programme Finance Initiative’s (UNEP FI) Adaptation Finance Taxonomy Playbook. While the EU Taxonomy focuses on an activity’s substantial contribution to reducing physical risks, the CBI/UNDRR distinguishes between activities that adapt and those that enable others to adapt. UNEP FI often uses the terms interchangeably, treating both as reducing vulnerability and strengthening capacity.

2. What the definitional difference means for financing instruments

Mapping how adaptation and resilience perform across financing dimensions reveals why conventional capital market instruments struggle to scale. Adaptation is exposure-driven and responds to a permanently shifted baseline of chronic changes in average conditions. Resilience is risk-driven and responds to the proximity of irreversible thresholds. Adaptation returns are characterised by balance sheet protection and reduced impairment risk for existing assets rather than new income generation. Resilience involves systemic avoided losses that are diffuse across actors and not attributable to a single investment. Consequently, while adaptation can often be attributed to a specific asset, resilience cannot be attributed below the system level. Standard instruments like project finance often lack the revenue waterfall required for these investments, making outcome reporting structurally difficult.

3. Two approaches and their limits

The taxonomy-led approach establishes precise boundary conditions through regulation and technical screening criteria. However, its limitation is structural: if resilience is a system property, classifying individual projects as resilient is a category error. This approach can also lead to maladaptation, where local protection transfers or amplifies risk elsewhere, such as a sea wall redirecting hydraulic stress to adjacent areas. The demand-led approach takes meaning from revealed preferences, such as insurance retreat or rental compression. However, these market signals are often distorted where public subsidies absorb losses, creating a mismatch between signal and value. Near-term growth in adaptation investment is likely to concentrate on technologies with growing market demand, such as heat management equipment, water recycling systems, and flood-resilient construction materials.

4. The missing institutional architecture

Neither the taxonomy-led nor the demand-led approach resolves the underlying institutional gaps preventing finance from operating at scale. A limiting case is the Carbon Tracker analysis of Norges Bank Investment Management (NBIM). Although NBIM is an analytically advanced asset owner with a long-term mandate, it has found that common models understate exposure by several multiples and cannot yet translate analysis into portfolio-wide capital allocation signals. Five conditions are absent: a defined reference condition, an outcome verification infrastructure, a mechanism for capturing diffuse benefits, a governance structure for long time horizons, and a valuation framework calibrated to where adaptation value sits. Conventional expected-loss models systematically undervalue adaptation because benefits concentrate in low-probability, high-consequence tail events, such as the 2024 Valencia floods or the 2025 Los Angeles wildfires.

5. Key questions and who must answer them

Several critical questions remain because the necessary institutional structures do not exist at scale. These include determining baseline climate scenarios, measuring successful adaptation, and capturing broader social and fiscal benefits. Institutional owners for these questions include regulators, standards bodies, and public finance authorities. Governments must decide which commitment mechanisms carry sufficient legal durability for long-lived assets. Appraisal bodies and financial regulators are also tasked with developing valuation models that account for catastrophic loss prevention rather than just average annual cost reduction. Furthermore, planning authorities must determine the geographic scale at which adaptation adequacy should be assessed to ensure individual actions are not rendered worthless by a lack of coordinated investment.

6. The built environment

The built environment is where definitional problems become operationally expensive, as assets are long-lived and geographically fixed. For buildings and infrastructure, ambiguity shapes planning consents, asset valuations, and insurance underwriting. A major challenge involves the retrofit and repricing of existing stock; most buildings that will constitute the 2050 built environment were designed under different physical assumptions. Building regulations often act as de facto adaptation taxonomies but are frequently calibrated against historical data. Additionally, “adaptation interdependence” means the benefit of individual action is limited if surrounding infrastructure is inadequate. For example, flood resilience for a single property is near-worthless if local drainage fails. From 30 April 2026, RICS members valuing commercial property worldwide will be expected to document climate-related factors, requiring evidenced asset-level substantiation of impacts on value.

Conclusion: towards a shared language

The constraint to scaling adaptation and resilience finance is primarily institutional. The most productive contribution the sector can make is the institutional work of agreeing on reference conditions, building outcome measurement infrastructure, and designing benefit-capture mechanisms. Without these structures, definitions remain illegible across counterparties, resulting in systematic underinvestment in assets most exposed to a shifting physical baseline. Bridging this gap requires governance capable of making credible commitments across the full asset life cycle and across the long time horizons that climate adaptation actually requires.