A senior financial expert questions climate risk

In a speech on May 11, U.S. Federal Reserve Governor Christopher Waller provocatively questioned the Fed’s focus on climate risk. “Climate change is real, but I do not believe it poses a serious risk to the safety and soundness of large banks or the financial stability of the United States,” he stated. Following in the tradition of other free thinkers such as Stuart Kirk, Michael Shellenberger, and Bjorn Lomborg, Governor Waller argued that physical risks “are not material enough to pose an outsized risk to the overall U.S. economy”, while “transition risks are generally neither near-term nor likely to be material given their slow-moving nature.” His bottom line:

“placing an outsized focus on climate-related risks is not needed, and the Federal Reserve should focus on more near-term and material risks.”

Waller’s assertion regarding the Fed’s rightful focus is reasonable and his query about climate risk’s near-term materiality is a critical question. At a time when markets face threats from many quarters – great power conflicts, geopolitical hardball over energy resources, inflation, massive government debt, commercial real estate crises, and bank implosions, to name several – why give climate change disproportionate weight? 

Climate risk is nearer and deeper than it seems

A skim through the recent climate science literature is a quick path to answers. Global leaders surveyed by the World Economic Forum see this and rank a host of climate-related and environmental risks at the top of their list of concerns for the coming decade. Yet most of what we hear about the climate crisis – even from experts in science, policy, business, and the media – doesn’t quite capture the urgency and scope. In fact, the set of ecological and societal phenomena labeled “climate change” encompasses a much broader range of potential hazards and shocks in shorter time frames than most of us tend to consider. What does a world with average global warming of 1.5˚C warming – predicted by the World Meteorological Organization to appear by 2027 – actually look like (to speak nothing of 2˚C or more)? It’s safe to assume it will scarcely resemble the present or recent past in many important respects. 

Most troubling about Waller’s statements is his professed certainty. He asserts climate risks will filter through the financial system in predictable, observable ways that regulators will see coming. And he suggests that we have a firm grasp on the nature of climate risks and that they are manageable. But we know from the recent descent of apocalyptic air quality across the Northeast U.S. from Canadian forest fires that this is not true. 

Climate scenarios offer a false sense of security

Perhaps Waller’s confidence derives from his reading of recent climate stress tests, such as those by the Bank of England (BoE) and European Central Bank (ECB), which found limited exposure of the participating financial institutions to climate impacts. These exercises, despite building largely upon Intergovernmental Panel on Climate Change (IPCC) scenarios, assessed potential climate losses well within the bounds of typical stress tests – hardly headline-grabbing material. (In defense of the BoE and ECB, the BoE noted the “substantial uncertainty around the true magnitude of these risks” of its 2021 CBES exercise, appropriately qualifying the findings, while the ECB found that both physical and transition climate risks are material to its asset portfolio.) 

Unfortunately, climate scenarios currently favored by practitioners (and the basis for BoE and ECB climate risk exercises), such as the CMIP6 suite used by the IPCC and those developed by the International Energy Agency (IEA) and Network for Greening the Financial System (NGFS), have given many observers a false sense of security that climate risks are under control. This is because such scenarios primarily adopt a long-term time horizon to 2050 or beyond, implying the near term is safe territory. Additionally, most of these scenarios collectively derive from integrated assessment models, or IAMs, that adhere to a set of structural constraints (e.g., excluding the financial sector and extreme, disruptive events) and other limiting assumptions. Such models and scenarios are typically focused on longer-term trajectories at the global scale of a clearly delineated set of causes and effects: (i) technological change and policy adoption, (ii) resultant global emissions, (iii) attendant climate effects, and (iv) impacts on assets. In the case of physical climate hazards such as extreme weather, the scenario or stress test typically examines the impact of a single event, such as a hurricane or flood, in isolation.

What’s missing from climate stress tests?

While these approaches are insightful and instructive, their freestanding application is not tantamount to robust risk assessment. Let’s consider what such scenarios are missing. For example, the BoE and ECB climate stress tests did not pair climate impacts with financial downturns and limited their scope to certain portfolio assets. In other words, scenarios of this type may model a hurricane in Florida, but they don’t model a real estate crisis resulting from the collapse of the flood insurance market.

The BoE and ECB also examined physical and transition risk separately, without exploring how they might interact – for example, if vastly more renewable energy or electric cars on the grid make it more difficult to manage climate change-driven heat waves, or if climate disasters lead to government policies that rapidly accelerate the phaseout of fossil fuels, a possible foreshadowing of which we saw in the G7 ministers' communique this April. 

The limitations in mainstream climate models and scenarios coupled with their unquestioning adoption represent, in Erica Thompson’s words, “a failure of imagination”. She writes in Escape from Model Land, “In Integrated Assessment Models… there are no radical futures. Everything in these models looks like it does today, just with a bit more technology or a different price on carbon.” But we know the course of human events is messy and takes unexpected sharp turns.

The methods applied in climate risk assessments and scenarios driven by IAMs’ long-term projections have known blind spots and limitations recognized by leading economists, such as ignoring or underestimating volatility, short-term acute impacts, and secondary effects; the NGFS and the Financial Stability Board have recently acknowledged many of these shortcomings. Could an El Niño-fueled heat wave drive crop failures, spiking food prices? Could the Panama Canal, currently limiting cargo due to drought, have to close if the drought persists, bottling up six percent of global trade? An IAM-based scenario wouldn’t be able to answer these questions. Nevertheless, the findings of climate risk assessments based on such scenarios are often adopted without qualification.

Fundamentally, many of the models undergirding climate risk assessment ignore the potential for sudden disruptive changes of any kind, be they in human and societal behavior, politics, market shocks, financial panics, tipping points, extreme events, and other phenomena that might sharply depart from status quo. In focusing on longer-term greenhouse gas emissions trajectories at the global scale – without a doubt, quite important – IAMs look past this year’s possibility of a catastrophic heatwave (such as the one envisioned in the novel Ministry of the Future), a climate-driven famine due to crop failures, or supply chain shocks familiar from COVID-19 and the Russian invasion of Ukraine. 

Hidden green swans?

In ignoring precisely those factors that could be most financially destructive, narrowly-bounded climate scenarios that find limited potential losses are a tautology. The exclusion of certain tail risks from materiality assessments and scenarios due to perceived low probability or mismatched time horizon is commonplace. For example, in a disclosure published this month, a major international bank determined that the transition risk exposure of its fossil fuel loan book is not “quantitatively material” as the bank expects fossil fuel asset impairments to only begin to bite after 2030, beyond the maturity of current assets. 

Such calculations could be right. But could litigation attributing climate losses to fossil fuel companies create massive new liabilities, or natural disasters reshape public views of climate action, leading to much more aggressive government climate policy? The transition risk in these circumstances would look markedly different. As the seminal 2020 “green swan” report from the Bank of International Settlements noted, “Traditional backward-looking risk assessments and existing climate-economic models cannot anticipate accurately enough the form that climate-related risks will take. These include what we call "green swan" risks: potentially extremely financially disruptive events that could be behind the next systemic financial crisis.” 

Thinking more deeply about climate risk

Climate change has been described as a ‘wicked problem’ due to its complexity. This complexity in society and the natural world propels event cascades that can be anticipated to a degree but not perfectly forecast. In such contexts, robust and adaptive decision-making is needed requiring narrative-driven scenarios and “‘exploratory modelling’, in which simulations and data are used not to predict the future but to map large sets of assumptions onto their consequences, without necessarily privileging any one set of assumptions over another.” 

Conversations I have with my colleagues frequently probe these assumptions, many of which are so deeply embedded in our acceptance of the status quo as to go unnoticed. Regulators, risk managers, and investors would be well-served to think harder about the unknowns before professing confidence in the limited materiality to financial assets of a risk domain as large and complex as climate change. The stakes are too high and the warning signs too apparent to ignore.