Paris, Gare de Lyon, 4am. Passengers with crumpled, sleepless faces stream out of the station before dawn, taxis rushing in to claim this unexpected windfall as the metro has been closed for hours. It is the morning of Monday 13 July 2026, and the great forest of Fontainebleau, south of Paris, has been ablaze since the previous day. High-speed trains linking Marseille and Provence to Paris have been diverted onto reduced-speed lines; a journey that usually takes around 3 hours took 7 this time. A lucrative night for Parisian taxis, a ruinous one for the rail operator: every stranded passenger will be fully refunded. The French railway company SNCF estimated the cost of climate-related disruption at €90 million for the first half of 2026.
The train may be the most decarbonised mode of long-distance transport, but it is also one of the most exposed to climate hazards, whether acute and sudden (wildfires) or chronic (rising temperatures). Under extreme heat, rails buckle, ageing trains' air-conditioning systems give out, and overhead lines expand and can even snap.
This summer, Eurostar amended its order with Alstom: the rolling stock ordered must now withstand temperatures of 55°C, up from the 45°C originally specified.
But a heat-resistant train still cannot run through wildfire, nor without power. The operator of the British electricity grid, the National Energy System Operator (NESO), warned at the height of the mid-July heatwave of a risk of running short of electricity. Demand was peaking, driven largely by air conditioning and ventilation. At the same moment, one of its key suppliers, the French utility EDF, had to scale back nuclear output because river water needed to cool its reactors had grown too warm. When a heatwave hits, the whole economy overheats.
Grasping physical risks and the financial risks running beneath them is therefore a far from straightforward exercise. What role should the financial sector play now that physical risk and resilience must be built into portfolios, loans and ESG risk management? And how can it cope when so much of the relevant information (a heat-resistance clause in a rolling-stock order, for instance) never appears in an annual report or a data feed? We put these questions to Dr Anna Moss, who joined Aberdeen six years ago after a decade of postdoctoral research on climate adaptation, providing advice to the Scottish Government on how to develop and monitor adaptation policy.
For Anna Moss, finance is following a pattern she first saw in public policy: adaptation lagging behind mitigation. But the frequency and magnitude of extreme climate events are changing. They can no longer be written off as isolated incidents, and so resilience needs a consistent, systematic place in investment analysis.
The usual starting point is climate scenario analysis. It remains useful for a top-down view across sectors, sub-sectors and the wider economy. But it has well-known blind spots for physical risk including:
Discounting: physical losses mostly materialise decades ahead, and conventional methodologies heavily discount distant events, so large and highly likely losses end up understated.
Aggregation: at portfolio level, transition risks tend to outweigh physical risks, so physical impacts rarely come through clearly. Transition risk tends to hit issuers in the same industry in a similar way. Physical risk, by contrast, depends on location and varies widely from one company to another, so it gets diluted when aggregated.
The deeper problem is that physical risk is local. A hazard strikes a given plant, mine or railway line, not a sector average. Aberdeen is therefore starting to take a bottom-up approach that looks at individual asset locations, not only for real estate and infrastructure but also for the listed companies it invests in.
In practice, physical-risk analysis is often constrained by data gaps and differing methodologies. Hazard maps, portfolio holdings, the geolocated assets of each company, and the indirect links between companies through supply chains are frequently incomplete, inconsistent, or difficult to link. Even the largest banks are still working out how to make these layers talk to each other.
Much of the financial impact is indirect, which is exactly what the train example shows. Aberdeen's research team has produced several internal deep dives this year to show investment desks where physical risk is already landing on the bottom line:
European droughts and food security: chronic warming combined with El Niño1, wildfires, low water levels on the Rhine and the Danube disrupting commodity transport, and fertiliser supply restricted by tensions around the Strait of Hormuz.
Heatwaves and power grids: cooling demand pushes electricity consumption up while heat reduces generation and network efficiency, from nuclear plants limited by warm river water to overheating grid infrastructure.
Flooding and commodities: in Chile, which supplies around a quarter of the world's copper, flood events linked to El Niño and increasingly frequent rainfall can shut mining operations down.
The common message, says Anna Moss, is that these pressures are no longer one-off exceptions. Companies need resilient sourcing, flexible supply chains, efficient water use and climate-resilient production.
Knowing where a company is exposed is only half the picture. The real question is what the company is doing about it. But, whereas emissions give a clear metric, targets and milestones, and an end point for mitigation, adaptation has no finish line: it is an iterative process, so it has to be assessed through processes rather than a single number.
To build a common language, Aberdeen has chosen the IIGCC's Climate Resilience Investment Framework (CRIF), launched in June 2025, as its foundation. One reason is that it deliberately mirrors the alignment and maturity scale language of the Net Zero Investment Framework (NZIF), on which Aberdeen already bases an internal transition framework assessing both quantitative data and the credibility of company strategies. The resilience scale runs from bottom to top:
Not aligned: no commitment and no action on physical climate risk.
Committed: the company has an ambition to address physical risk and governance structures to implement it.
Aligning (planning): an adaptation and resilience plan is in place, feeds into business continuity planning and is disclosed.
Aligning (implementing): capital is allocated to back the plan, with internal mechanisms to deliver the strategy.
Aligned: resilience is embedded in strategy, with processes to understand the company's own physical risk, adequate capex, clear oversight and proper disclosure. Even here, it is an ongoing process rather than an end point.
Aberdeen translates each level into concrete questions about measures, governance and capex. The same question set serves bottom-up research and, where public disclosures fall short, can inform climate engagement with the companies judged most at risk.
This process-based approach also eases part of the data problem. Pinpointing a company's exact physical risk requires asset-level location data that is often missing. Assessing whether it has a credible resilience strategy, covering its most material hazards, is a more tractable question.
The process remains research-intensive, so Aberdeen focuses it where it matters most: products and mandates with a climate focus, such as a climate transition bond fund with a dedicated adaptation pillar, and holdings with the highest theoretical physical risk. Screening by data providers can flag whether companies at least mention an adaptation strategy, but judging its credibility still requires analysts.
The same framework also supports the opportunity side. Aberdeen distinguishes three adaptation opportunity sets, set out in a paper on investing in adaptation published in June:
| Opportunity set | Who they are | Examples | Investment character |
|---|---|---|---|
| Adaptation solutions | Companies whose products or services directly reduce exposure or vulnerability to physical hazards | Water efficiency, drought- and heat-tolerant products, resilient materials, risk analytics, insurance | Clear but small universe; often a minor share of a larger company's revenue |
| Resilience builders | Providers of core infrastructure and essential services that underpin system-wide resilience, especially in exposed regions | Utilities, transport, healthcare, food, communications | Largest opportunity set; closely tied to the adaptation finance gap, notably in emerging markets |
| Adaptation leaders | Companies in any sector that manage their own physical risk particularly well | Credible resilience strategies backed by capex, strong operational resilience, robust supply chains | Quality and downside protection rather than growth: less downtime and smaller financial impacts than peers |
The resilience builders category illustrates why the credibility assessment matters. Holding a monopoly on water or power supply in a region is not enough. A company must also show that its own infrastructure and practice are resilient before it can claim to build resilience for others.
The last two categories share a practical advantage. Rather than hunting for niche companies outside the usual universe, investors can start from holdings their analysts already know well, and ask which of them are operating in a resilient way.
For mitigation, the Science Based Targets initiative (SBTi) has given investors a widely recognised benchmark of credibility. Nothing comparable yet exists for corporate resilience. Each investor assesses companies on its own, and companies may face the same questions many times over. Until a shared standard emerges, Anna Moss expects the work to remain bottom-up and labour-intensive. Common frameworks such as the CRIF at least give investors a shared vocabulary for saying why a company is judged aligned.
Market data reflects the same gap. Major ESG data providers now offer granular climate value-at-risk models, multiple scenarios and large asset-level location databases. At company level, however, robust resilience assessments remain scarce, and country-level indices only go so far.
For financial institutions, the challenge is therefore as much about data management as about climate science. Hazard scenarios, geolocated assets, holdings, supply-chain links, public datasets and qualitative strategy assessments all need to sit in one place and feed a methodology that can evolve as frameworks mature. Investors who start building resilience into their analysis today, even step by step, will be better placed to catch the train before it leaves.
IPE, IIGCC launches climate resilience framework consultation
University of Notre Dame, ND-GAIN Country Index
WeeFin, Physical Risks and Public Data: What to Take Away from the FIR x WeeFin Event
Resilience: the next phase of climate strategy, Anna Moss and Scott McQuiston, 23 June 2026
Interview with Anna Moss, Aberdeen, 16 September 2026
1 El Niño is a periodic warming of surface waters in the equatorial Pacific that disrupts rainfall and temperature patterns worldwide.
