Transparency around climate disclosures continues to be an area of focus for those looking to manage risk.
In this three-part series we use ICE’s climate solutions to explore two dimensions of climate risk that are relevant when performing climate scenario analysis: physical climate risk (risk from physical hazards like floods and hurricanes) and transition climate risk (risk from the transition away from fossil fuels to a low-carbon economy).
In Part 1, we explore an approach that may be used to assess physical climate risk exposure that may also help investors understand U.S. institutions’ level of exposure in the mortgage market. In Part 2, we travel to Europe to study various options that may be used in stress testing, and finally in Part 3 we return to the U.S. for a look at the cross-sector challenges that lay ahead for achieving net zero goals.
The characteristics of the properties that underlie the mortgage loans on an issuer’s balance sheet can be used to understand physical climate risk exposure. In other words, across all mortgages that an institution owns, the extent to which the properties linked to these mortgages are susceptible to physical hazards like hurricanes, floods, and wildfires can help to shed light on the degree to which physical climate risk should be a consideration in risk management for issuers and investors. Mortgage loans are merely one of the puzzle pieces, private loans/debt may also play an important role.
Stratifying the loan retention rates can reveal where institutions are retaining credit and physical climate risks. Loan retention rate refers to the number of loans originated or purchased that remain on an issuer’s balance sheet, as a percentage of the total loans originated or acquired.
Our analysis of Home Mortgage Disclosure Act (HMDA) filings published by the Consumer Financial Protection Bureau for 2021 found that the average retention rate across all loans for depository institutions with large mortgage footprints (>10000 loans in 2021) is 34.1%, with higher retention rates of loans with greater climate-related financial risk exposure (Figure 1). For all other large mortgage originators (>10000 loans in 2021), the overall rates of balance sheet retention are more than three times lower (Figure 1), with negligible (<5%) differences in retention rate as a function of climate-related risk.
We focus here on the distinction between large depository institutions and other large mortgage originators primarily because of their different business models: generally, depository institutions take deposits, purchase assets of perceived good credit quality, and profit from the spreads. By contrast, non-depository institutions like mortgage companies profit from the production and short-term sale of mortgages rather than the aggregation of balance sheet assets.
FIGURE 1. Loan retention rates as a function of ICE Physical Climate Risk Score for (left) depository institutions with large mortgage footprints, and (right) other large mortgage originators. Colors represent average loan size. Source: ICE and HDMA Data (see footnote 1 for full citation).
Our analysis revealed some regional differences in the retention rate trends discussed above. In the U.S. South, where most of the ultra-high climate risk sits (Florida, Louisiana, Texas, and parts of South Carolina and Georgia) and the Northeast, retention rates increase as a function of climate risk for the depository institutions with large mortgage footprints (Figure 2). However, in the other highest physical risk area of the country – the wildfire-prone region of the U.S. West – we see much lower rates of loan retention (as much as 20-25% lower) than in the Northeast and South for these institutions.
FIGURE 2. Loan Retention rates as a function of ICE Physical Climate Risk Score for (left) large depository institutions and (right) other large mortgage originators, each broken out by geographic region. Colors represent average loan size. Source: ICE and HDMA Data (see footnote 1 for full citation).
In short, the financial community will have to deal with physical climate risk. This analysis scratches only the surface of what is possible with ICE Data. Digging into the granular details of individual institutions’ exposure to these risks across multiple dimensions and hazards is all possible today.
ICE’s climate physical risk data combines cutting-edge climate science, catastrophe modeling, and geospatial machine learning technology with ICE Data Services’ leading municipal reference data to offer a new tool to quantitatively measure climate risk in municipal bonds, mortgage-backed securities, and real estate.
1 Home Mortgage Disclosure Act (HMDA) [Dataset], Consumer Financial Protection Bureau, (HMDA Data)
2 The first type of financial institution as defined in Regulation C adopted by the Consumer Financial Protection Bureau to implement the Home Mortgage Disclosure Act. 12 C.F.R. 1003.2
3 The second type of financial institution as defined in Regulation C adopted by the Consumer Financial Protection Bureau to implement the Home Mortgage Disclosure Act. 12 C.F.R. 1003.2
4 Regions referred to here are defined by the U.S. Census Bureau