Careening, cascading risks of abrupt climate change

Why glaciologists and financial regulators should talk

Ice core storage

Glaciologists know that the earth's climate has shifted abruptly in the past. Ice cores drawn from thousands of feet down show very substantial increases in temperature—18ºF or more—over the course of a human lifetime. The scientists keep explaining to the rest of us that initial perturbations, spurred on by positive feedback loops that amplify and accelerate climate conditions, can push climate systems past tipping points and into nonlinear rates of change. We don't know exactly what part of the hockey stick we're in now, but we do know things are shifting quickly.

Last month was the 11th-consecutive month of record-breaking global warmth, , monthly global sea surface temperatures have been the warmest on record for 13 months in a row, sea ice coverage is low at both poles, and Brazil is experiencing devastating flooding that has driven half a million people from their homes. And that's just April 2024.

We also know the rate of change we are experiencing now could itself change suddenly. (The Ice at the End of the World, by Jon Gestern, does a terrific job of telling the story of how those ice cores came to be extracted and what they are showing us.) 

The best way to plan for abrupt changes that society has never experienced before is to construct hypothetical situations that incorporate extreme conditions and then test our systems against them. For an example of a first, cautious step along this path, take a look at the scenario planning the Federal Reserve asked large bank holding companies to do last year. The Fed should both be applauded for this initial effort and pushed to do more as soon as possible: America urgently needs its federal leadership to insist on better data and more rigorous and imaginative planning efforts, right now.

Last week, the Federal Reserve issued a guarded summary of climate scenario planning by six large bank holding companies: Goldman Sachs, Citi, Bank of America, JPMorgan Chase, Morgan Stanley, and Wells Fargo. ("Pilot Climate Scenario Analysis Exercise, Summary of Participants’ Risk-Management Practices and Estimates," May 2024.)

Like a wall of glass brick, the document is both substantial and opaque, letting readers see the movement of shapes beyond but little else. The Fed is constrained in both what it can ask of these banks and what it can report. It made clear several times that this scenario exercise was designed only to explore the banks' capacity to take physical climate change risks into account. The report seems to be saying, "Hey, we're just getting our feet wet! We're just learning. This is a multi-year effort." 

In this exercise, the banks were asked to look at their portfolios of commercial and residential real estate and translate the estimated climate shocks caused by both a severe hurricane in the Northeast in 2050 (all the banks considered this) and a different risk (which the banks chose individually based on their particular circumstances) into default and loss probabilities. They were asked to assume both ongoing insurance and insurance vanishing. And they emerged with average numbers that won't jolt any particular homeowner: Residential real estate—about half a million loans across five banks—would go down in value by about 0.1% in the Northeast on average in the event a serious storm arrived and there was no insurance to act as a backstop.

This is not a dramatic finding. It's very much an average, and it's based on a narrowly-defined hypothetical.

The key message from this summary was not the numbers the banks reported but the difficulty they had going through the exercise. Most were relying on third-party vendors, and it was extremely challenging to evaluate the models the vendors were using to give advice: "The "challenges [associated with reviewing the vendors' climate models] included limited data, lack of back-testing capabilities, nonlinear risks, scenario horizon, heavy reliance on judgment, limited reliability of model output, and time constraints." In other words, they were faced with black boxes they didn't understand.

Crucially, the banks were missing a boatload of data they'd like to have. Data about building characteristics, like property location, construction materials, and other things. Data about where insurance is offered, how much it costs, and how quickly insurance rates are going up. These data gaps are like crevasses in the Greenland ice sheet: unexpected, dangerous, and deep.

Here are the bright spots: The summary does show that major banks are interested in staffing up internally to assess physical risks "in order to reduce reliance on third-party vendors." It seems that the "indirect impacts" of climate change (effects on the local economies within which they operate as well as on the banks' facilities themselves) are seen as important by these major banks. At least some of these banks are interested in investing in better climate and exposure data and generally getting better at modeling climate risks. 

Our fractured, full-of-holes financial regulatory system has so far allowed banks to mostly avoid grappling directly with the possibility of swift, unprecedented physical changes being caused by climate change. Things are moving very slowly: this Fed report is coming out fully three years after President Biden issued an executive order calling for assessment of the physical risks of climate change "to the financial stability of the Federal Government and the stability of the U.S. financial system." The Fed report is a good first step in along these lines. 

It is also a call to action. It shows that there are data gaps that must be filled right away. It shows that scenario modeling is still in its early stages and needs to advance quickly. We should be throwing hypotheticals involving multiple, interconnected climate/economic challenges at our financial system and seeing where our weaknesses are. It shows that consultants are heavily relied on and greater in-house capacity is needed, right away, inside both government and financial institutions. If this is a multi-year effort, the government needs to step up the pace. 

Meanwhile, the ice scientists keep churning out reports of their own, attempting once again to communicate that it is undeniable that rapid, destabilizing change has happened before—at a time when C02 levels were climbing 100 times more slowly than they are now.

No doubt the politicization of GHG emission-reduction-reporting has made everyone wary, but physical climate change risk planning has to happen. No one can deny that these changes are accelerating, and everyone has to plan for them. Right now.

Perhaps some glaciologists should be brought on staff at Treasury.