Patrick Donnelly, CFA, CFP ^®, Manager, Client Consulting

Fannie and Freddie Making Headlines

Effective May 1, after several years of no change, the LLPA matrix was updated. For loan-to-value (LTV) ratios 70% and below, pricing either decreased or remained unchanged for almost all borrowers. The controversy enters the picture for LTVs between 75% and 95%. Borrowers with credit scores above 680 saw an increase in almost all instances, while scores below 680 saw an across-the-board decrease. All else being equal, borrowers with better credit still pay less than riskier borrowers, but the spread has narrowed; an action many headlines are calling a subsidy.

The FHFA responded in late April with several clarifying statements to what they called a “fundamental misunderstanding about the fees charged by the Enterprises.” In their release, they mention that it had been several years since the pricing matrix was deeply examined and that the intention was to better align fees with the expected long-term financial performance of the loans. The Urban Institute also points out that while lower credit borrowers are seeing a decrease in agency costs for LTVs greater than 80%, those loans are protected by Private Mortgage Insurance (PMI) which charges a premium for poor credit.

Giving the FHFA the benefit of the doubt and assuming that the new LLPA matrix is a true attempt to reprice risk, the questions shouldn’t center around whether the new matrix subsidizes the poor, but whether it subsidizes poor credit. The two don’t necessarily go hand-in-hand. Since income isn’t factored into the calculation of a credit score, the correlation between the two is less than perfect. In fact, a 2018 Fed study proclaimed that the correlation between income and credit scores is only modest, that scores across all income bands are dispersed, and that income is not a strong predictor of credit scoring. A more recent study from Credit Sesame illustrates some of the dispersion: 41% of low-earning individuals making between $15,000 and $30,000 per year have strong credit scores above 720. Conversely, 28% of the individuals earning $75,000 to $100,000 per year have credit scores below 720. All of this to say, home affordability for lower-income households with strong credit may very well be negatively impacted.

From a demographic standpoint, regions that are otherwise similar from a cost of living and income standpoint may benefit more than others. Take Texas and Michigan for example: both states land at almost the exact same point on the Composite Cost of Living Index. The median household income is actually about $4,500 higher in Texas, yet the average Texan will likely catch a break to close a Freddie or Fannie loan over those in Michigan. Why? The average credit score in Texas is a 674 while Michigan has a 705, despite those in Michigan earning slightly less with similar expenses.

In 2021 the FHFA did provide a more direct subsidy that ironically flew under the radar. Closing costs for cash-out refinances and second home purchases, loans typically made to higher income households, saw their fees increase. This increase subsidized a decrease in loan fees for certain types of loans for households earning less than the “regional average." While this seems to align more directly with the affordability aspect of the mission, regional averages can be an exclusive hurdle. For example, households in Ohio making more than $62,000 per year would likely not qualify; if they have strong credit, they may very likely take an increase.

There is one flier that should at least be considered given the nature of Freddie and Fannie themselves. Despite being government-sponsored, and in the conservatorship of the government since the Great Recession, both agencies are privately operated, publicly traded, and responsible for generating enough revenue to back the MBS that they issue. Said another way, they have shareholder obligations. The guarantee they provide the MBS market is not backed by the full faith and credit of the US Government, rather the government guarantee is implicit. That implicit guarantee last came in the form of a controversial 2008 bailout. With increased rates, decreasing home values, and an impending recession, both agencies have recently written billions of dollars of default reserves into their financials. They rely on fee generation to be able to guarantee the mortgages they package into MBS.

In 2009 after the mortgage industry collapsed, we saw the average FICO score on new Fannie Mae loans increase from 722 to 761 while the number of loans issued to FICOs below 660 decreased by 4%. A less extreme example, but following the rate hikes in 2018, we saw the average FICO score for a Fannie loan increase from 743 to 749 with those below 660 decreasing by 3%. This makes sense intuitively; lending tightened, and banks wrote higher quality loans. It is reasonable to expect that we are entering similar conditions where banks will tighten, spreads will widen, and the average new loan will be of higher quality. Is it possible that the agencies are factoring this in and increasing fees where they anticipate receiving more volume?

Sources: LLPA Matrix update 3.22.23 | LLPA Matrix update 5.1.23

Investment advisory services provided through CBIZ Investment Advisory Services, LLC, a registered investment adviser and a wholly owned subsidiary of CBIZ, Inc.