14 October 2026
Extreme heat can reduce a company’s profitability without causing any visible damage to its assets. A factory may remain operational but produce less. A power plant may have to reduce output. Workers may become less productive, while cooling and maintenance costs rise.
For investors, these disruptions present a valuation challenge. Unlike physical damage caused by floods or storms, the financial consequences of extreme temperatures can be gradual, difficult to isolate and largely invisible in traditional risk assessments.
Artificial intelligence could help bridge this gap. By combining climate data, operational information and financial modelling, AI systems may identify relationships that conventional analytical approaches struggle to capture.
However, as Mustafa Dah argues in his recent Enterprising Investor article, “How Investors Can Evaluate AI Heat-Risk Models for Company Valuations”, the usefulness of these models depends on their ability to demonstrate how physical climate exposure translates into measurable financial consequences.
The challenge is not simply to identify where companies face rising temperatures. It is to understand what those temperatures mean for production, costs, margins and ultimately company value.
The missing link between climate risk and financial performance
Financial institutions already rely on sophisticated tools to assess physical climate risks. Banks, insurers and asset managers use climate scenarios, geospatial information, catastrophe models and vulnerability assessments to estimate potential losses across portfolios - yet, the results can vary substantially between providers. The article refers to research comparing six firm-level physical climate-risk scores, which found considerable differences even between methodologies designed to measure similar risks.
A separate benchmarking exercise conducted by the Global Association of Risk Professionals for the UK’s Climate Financial Risk Forum compared thirteen vendors assessing the same property portfolio. The exercise revealed substantial discrepancies in hazard and damage estimates, alongside limited standardisation of heat-related metrics.
These differences highlight a fundamental problem: models may not be measuring the same economic exposure.
Some approaches focus primarily on buildings and physical infrastructure, estimating factors such as air-conditioning expenses or potential equipment failures. However, a company’s most significant heat-related losses may arise elsewhere. Higher temperatures can reduce employee productivity, constrain industrial processes, increase energy consumption or disrupt supply chains without causing physical damage to the underlying assets. A model that accurately estimates property exposure may therefore overlook the mechanisms through which heat affects a company’s earnings. AI can potentially improve these assessments by integrating more detailed operational data. But greater computational sophistication does not automatically produce more reliable financial conclusions.
From temperature data to company valuation
Dah identifies five connected stages that a credible heat-risk assessment should address:
Hazard → Asset exposure → Operating sensitivity → Financial effect → Valuation
Each stage introduces potential sources of uncertainty.
Accurate temperature forecasts are of limited value if the model fails to identify the facilities exposed to extreme heat. Even when the relevant assets are correctly mapped, estimating the operational impact requires information about production processes, cooling systems, working conditions and the company’s capacity to adapt.
The next challenge is establishing causality. A decline in production during a heatwave does not necessarily mean that high temperatures caused the decline. Lower demand, maintenance activities, equipment failures or higher input costs may also affect performance.
Models must therefore distinguish heat-related effects from other factors influencing financial results. Only then can investors estimate the implications for revenue, operating expenses, profitability and cash flow. This distinction is essential because a precise-looking financial estimate is not necessarily an accurate one.
The final stage involves incorporating the expected effects into company valuations. Here, the article highlights another important risk: double counting. Expected heat-related losses should generally be reflected in projected cash flows rather than being included again through a higher discount rate. An additional risk premium may be justified for systematic risks not already captured in those forecasts, but investors must distinguish expected financial losses from residual uncertainty.
EDF: When rising temperatures reduce electricity production
The French energy company Électricité de France (EDF) provides a concrete illustration of how extreme heat can affect company value. Nuclear power plants depend on water for cooling. During periods of high temperatures and low river flows, environmental restrictions on cooling-water discharges may require operators to reduce electricity generation. According to figures cited in the article, EDF estimates that annual nuclear production losses related to environmental conditions could increase from an average of 0.3% between 2000 and 2020 to 1.5% by 2050.
These long-term averages, however, conceal the potentially concentrated impact of individual events. During the heatwave of late June 2026, EDF reported that approximately 5.5 gigawatts of nuclear generating capacity, equivalent to 8.7% of its installed French nuclear capacity, was temporarily unavailable. The financial implications extend beyond the volume of electricity not produced.
If generation declines during periods of strong demand, the company may lose valuable sales or incur additional costs to purchase replacement electricity. A credible heat-risk model must therefore connect local environmental conditions with individual plants’ operating characteristics, production capacity and financial performance.
It should also account for adaptation investments. In early 2026, EDF submitted a detailed adaptation plan covering planned expenditure through 2040 across its French nuclear, hydropower and island operations.
For investors, these measures introduce another valuation question: does the present value of the future cash flows protected by adaptation exceed the present value of the investment required? Answering it requires more than estimating future temperatures. Investors need to understand how individual facilities respond to changing environmental conditions and how proposed investments could alter those responses.
Why company disclosures matter
The article argues that better AI models will not be sufficient without more detailed company-level information. Investors already receive sensitivity analyses showing how changes in interest rates, exchange rates or commodity prices could affect financial performance.
Companies with significant exposure to extreme heat could provide comparable disclosures. These might identify the temperature thresholds that affect operations, the facilities and activities exposed, and the resulting consequences for production, revenue, operating expenses, capital expenditure and cash flow.
Companies could also disclose how planned adaptation measures are expected to reduce these effects. For example, a business might estimate its production and free cash flow under moderate, severe and extreme heat scenarios, both before and after adaptation investments. Investors could then incorporate these assumptions into discounted cash flow models, assigning probabilities to different scenarios while recognising the uncertainty surrounding them.
Such disclosures would not eliminate forecasting difficulties. They would, however, make the relationship between climate exposure and financial performance more transparent and provide AI models with more reliable company-specific inputs. Where historical information is available, investors could also compare projected sensitivities with actual operating outcomes during previous heat events.
What investors should ask AI providers
As climate-risk modelling becomes more sophisticated, investors will need to evaluate not only the results produced by AI systems but also the assumptions and methodologies behind them. Dah identifies three particularly important questions.
First, which assets and operational dependencies are included in the assessment, and which temperature thresholds affect their performance?
Second, how does the model distinguish the financial impact of heat from other factors influencing company results?
Third, how are expected heat-related losses incorporated into cash flow forecasts, and what uncertainty remains?
These questions help establish whether a model provides a credible valuation input or merely generates a sophisticated estimate of physical exposure. AI can support the process by analysing large datasets, estimating operating sensitivities, developing scenarios and performing financial calculations. Nevertheless, the choice of scenarios, probability assumptions and discount rates remains fundamental to the resulting valuation. Investors must be able to examine and challenge those assumptions rather than relying exclusively on the final number.
Better models require better financial reasoning
The growing availability of AI-powered climate-risk tools creates opportunities for investment professionals. More detailed data and advanced analytical techniques could improve the identification of vulnerabilities that traditional models overlook, particularly where physical risks affect productivity and operating efficiency rather than directly damaging assets. But the value of these tools ultimately depends on whether they improve financial decision-making.
A model that identifies rising temperatures without establishing their impact on a company’s cash flows offers limited support for valuation. Similarly, a model that generates precise estimates without transparent assumptions may introduce a false sense of certainty. The most useful AI applications will therefore be those that connect physical risks with operational performance, financial consequences and investment decisions through a clear and verifiable analytical process.
As climate-related risks become increasingly relevant to corporate performance, the competitive advantage will not necessarily belong to investors with the most sophisticated models, but to those best able to evaluate what those models actually measure. AI can help quantify the financial consequences of extreme heat. Determining whether those estimates are credible, material and appropriately reflected in company valuations remains a task requiring rigorous investment analysis.