‘Toolkit is broken’: climate modelling is straying away from climate science

If climate scientists were to review how economists model climate risk, most would vehemently disagree with what they see. New research from the University of Exeter and Carbon Tracker Initiative has found evidence to that effect. Mainstream economic models have strayed away from climate science.

60 climate scientists have weighed in on economists’ climate modelling. Their judgement is dire – economists’ models are dangerously inadequate.

The report, published today, outlines how a ‘fundamental disconnect’ has emerged in climate models that inform long-term investment decisions.

Damage functions

Asset owners, pension funds in particular, look to economists’ models to estimate the impact of climate change on their portfolios. What, for instance, would equity valuations look like in 2050 if current warming levels inch closer to 2°C?

To answer these questions, the models take historical relationships – between climate metrics and economic impacts – and project them forward. Damage functions, in technical parlance. The math, the report finds, is not capturing the full range of climate damage.

“These estimates ignore several important factors, including direct effects of heatwaves, rising sea levels, damage from tropical cyclones, and potential climate tipping points”, the report reads.

Partly, the fault rests in the choice of metric. Mainstream damage functions tend to use average global temperature as a proxy for climate impact. It has its merits, scientists reckon, but this average-over-extremes approach has little to say on tail risk probabilities.

Jesse Abrams, lead author and senior impact fellow at the University of Exeter says flawed damage functions are more than just minor technical glitches.

Net Zero Investor’s Renewable Infrastructure Forum 2026: 18^th, March, The Chesterfield Mayfair

“For financial institutions and policymakers relying on these models, this isn't a technical problem – it’s a fundamental misreading of the risks we face, which current models miss entirely because they assume the future will behave like the past, even as we push the climate system into uncharted territory”, Abrams warns.

The scientific community seemingly takes tail risk probabilities more seriously than economists do. Riccardo Rebonato, a professor at the EDHEC climate institute is amongst those calling for change.

His modelling,probabilistic in its entirety, produces a loss range depending on climate damage variation. Tipping points crucially, can be accounted for if need be.

Disciplinary silos

Blame also rests in disciplinary boundaries being rigid. In what is characteristically an interdisciplinary line of enquiry, strict disciplinary silos have emerged over the years.

Not enough economists are collaborating with climate scientists and vice versa. When researchers from Exeter and Carbon Tracker launched their survey, they encountered some friction. Scientists were reluctant to comment beyond disciplinary boundaries.

“This reluctance to comment across disciplinary boundaries is an important finding itself”, the report reads.

“Until the gap between scientists and economists’ expectations of future climate damages is closed and government bodies act to ensure the integrity of advice upon which investment decisions are made, financial institutions will continue to chronically under-price climate risks - meaning that pension funds and taxpayers will remain dangerously exposed”, commented Mark Campanale, Carbon Tracker Initiative’s founder and chief executive.

Investors on board

Given the centrality of modelling to decisions, some investors are unsurprisingly pushing back.

Phoenix Group, for instance, is throwing its weight behind the report’s findings. “Phoenix supports the report’s call for a more robust and co-ordinated approach to climate risk modelling”, said the group’s head of sustainable investment research Hetal Patel.

“As one of the UK’s largest asset owners, we urge policymakers to act decisively on the systemic risks identified in this research and to set clear expectations for financial sector users”, Patel affirmed.

Another major asset owner joining the call is UK local government pension pool Brunel Pension Partnership.

“The toolkit is broken”, warns Faith Ward, the pool’s chief responsible investment officer.

Climate change, she says, is both immediate and systemic in its risk profile.

“Yet, when translated through risk tools that are provided to us as investors, damages are estimated in fractions of a percent. The consequence is a failure to equip pension funds with a fully informed range of climate outcomes that can help them to address a fundamental systemic risk to the fund’s financial performance”, she adds.