Climate stress testing isn't crazy. Not stress testing for climate risks is.

Bloomberg News

A long time ago I used to be an environmental policy reporter. Most of my time was spent covering the Environmental Protection Agency and policies coming out of its hazardous waste division, but there were also lots of cases wending their way through the courts, bills making their way through Congress and important policy questions around how to limit humans' disruption of the natural environment — and how to limit the environment's disruption of people's lives. 

Before too long it became apparent to me that all of these policy debates about what "toxic" means or what makes a wetland a wetland are really centered around one thing: money. Hazardous waste cleanups are stupendously expensive, as are wastewater treatment plants, rainwater detention tunnels, air emissions scrubbers and all the other man-made things we require companies to build to undo their man-made damage to the environment. Every dollar that a company doesn't have to spend on those projects is a dollar they can use to do something profitable, so naturally wherever regulators draw a line in the sand has a big impact on a company's profitability — or even its survival. 

So it is little surprise that this dynamic — which is a tale as old as time — has spilled into the policy debate about how to address climate change generally, and into the Fed's nascent efforts to have banks account for and mitigate the risks that climate change poses to their balance sheets specifically.

The arguments against having the Fed get involved fall into a few categories: It goes beyond what the central bank ought to be doing; it's already covered in operational risk controls and the cause-and-effect relationship between climate stress and balance sheet losses is not yet well-established.

I'm going to focus my attention on the latter point, because to me it seems to be the best argument against having the Fed or any other regulator get involved in addressing climate change via loan loss provisions and capital requirements. 

Stress testing works like this: If, say, unemployment reaches X level and, say, GDP falls to X level, then banks can expect Y losses to their balance sheets. Making sure that Y loss result is still above the minimum necessary to keep the lights on is the name of the game, and a century or more of economic data backs up some of the test's reasonable assumptions about how X results in Y. Climate change, by contrast, lacks the robust data set necessary to draw similarly firm connections between X climate change events and Y balance sheet results.

Consumer banking

Big banks are breaking their promise to fund a green energy transition

May 26, 2023 10:00 AM

But some recent developments illustrate — at least to my satisfaction — that these risks are there whether we can anticipate them reliably or not. Arizona has recently issued a moratorium on new construction in the Phoenix area that requires well water, a move meant to reduce demand for an ever-dwindling critical resource in the Colorado River basin. What is more, private insurers like State Farm and Allstate have recently announced that they are no longer offering homeowners insurance in California because of increased loss risks due to climate change. Nine states are suing the Department of Homeland Security over rate hikes for national flood insurance premiums — hikes that are likely inevitable.

You'll notice that these developments all have one thing in common: They affect people's ability to buy, rent or build homes. And you'll also notice that they are coming at a time when there is a pronounced lack of housing nationwide and an even more pronounced lack of affordable housing. If climate change is making it harder to build and insure homes in risky places, those homes are going to have to be built some other way or built somewhere else because people still need shelter they can afford.

The Fed is not the lynchpin of averting a climate catastrophe — the world will not be saved with higher capital requirements or loan loss reserves alone. It is also true that if climate stress testing ever grows teeth, the connections between climate change and credit losses will need to be firm — or at least defensible. 

But the writing is also on the wall, and the time is now to strengthen the connections between a changing climate and economic growth. People can and will adapt, but that adaptation is going to cost money. The question is do you pay now, or pay more later.