Fannie Mae climate-risk analytics provider revealed

Eva Marie Uzcategui/Bloomberg

A major mortgage investor has been working to quantify the extent to which its holdings are exposed to potential damage from natural disasters using analytics from at least one private vendor.

That government-related investor, Fannie Mae, has been working with a set of climate risk analytics from Jupiter Intelligence to assess its portfolio of over 17 million single- and multifamily assets, the technology vendor revealed Tuesday.

That means Fannie has been looking at its climate risk exposures "today, in the past and every five years in the future," through technology made possible by the availability of natural-world data from sources like satellites and other forms of automation, said Rich Sorkin, CEO of Jupiter Intelligence.

"All of this, the simulations, the incorporation of sensor data, the inclusion of [artificial intelligence], is leveraging advancements in AI that came out of other sectors, and also the dramatically declining cost of cloud computing and just hardware in general," said Sorkin, of the advances in automation that allowed his six-year-old company to build its models.

The announcement comes as hurricane season, a peak period of climate risk between June and the end of November, gets underway. At the same time, some states with concentrated disaster risk, like Florida and California, have been struggling with insurability. Regulators, too of late, are increasingly focused on how disaster risk could affect entities they oversee.

Those factors have been fueling broad mortgage industry interest in assessments Jupiter offers, which can quantify risk by using different models for various perils. In particular, home lenders are concerned about the impact on home insurance, Sorkin said.

"Generally speaking, the mortgage market isn't going to work if there isn't insurance available," said Sorkin.

Fannie's regulator and conservator, the Federal Housing Finance Agency, for several years has had a natural disaster response team that coordinates with the government-sponsored enterprises and other stakeholders.

However, the FHFA has had work to do in managing and quantifying concerns caused by disasters or energy transitions, said Daniel Coates, deputy director for the division of research and statistics, during an online event late last year.

"The next step will be to figure out ways for these potential consequences to be considered by our regulated entities and us during the normal course of policymaking," he said at the event held by Ceres, a nonprofit that coordinates with capital markets experts on sustainability issues.

Using a score to size up the portfolio's disaster risk could help Fannie estimate and prevent potential losses, and ensure there aren't racial disparities in servicing outcomes. 

Fannie's equitable housing plan specifically calls for it to "provide climate analytics to empower communities with data, enabling them to make a stronger case for change and resources."   

While regulatory interest in developing climate risk models has been growing, scores may not immediately be as accurate in predicting the impact of disasters on mortgages as more established measures the industry uses to assess interest rate and credit concerns.

"Efforts to quantify climate risks are likely to fall short of the precision of existing models," Sean Becketti, a former Freddie Mac economist, wrote in a report written for Mortgage Bankers Association think tank Research Institute for Housing America in 2021.

Jupiter's climate-risk score suite has its limits but improves upon past models in this area that were more retrospective, Sorkin said.

"We did have a global financial crisis in 2008 that wasn't captured in the [traditional] models. I think a lot of what's going on in 2023, that's just 15 years later, isn't well captured in the [standard] economic models," he said. "Our models are directionally correct and quite useful in terms of understanding what's happening on an aggregate basis."

While natural disasters may not be entirely predictable, known experience with it can be quantified and projections based on global warming trends can be applied to it, said Sorkin.

"Certainly, the planet as a whole is indisputably getting warmer, the atmosphere itself as well as the oceans. There's incontrovertible data on that, and that's driving changes in the weather that are much better understood than the decision making of the vast majority of large corporations are taking it into account," he said.