There has been a lot of concern in the real estate investment community recently about a rule recently enacted by the Federal Housing Finance Agency for Fannie Mae and Freddie Mac loans regarding mortgage fees.
The gist of the complaint is that homebuyers with good credit will now have to subsidize those with bad credit. Technically, it’s true. However, the way it is framed is quite deceptive. The general argument goes something like this: those with a FICO score of 620 will get a 1.75% discount, and those with a FICO score of 740 will pay 1%.
Or another example would be this particularly popular tweet:
Homebuyers with credit scores of 680 or higher will pay around $40 a month more on a $400,000 home loan.
Buyers with 15%-20% down payments will be shocked by the higher fees.
Buyers with riskier credit scores and lower down payments will get lower rates and fees. pic.twitter.com/yVEp3btNJg
— Wall Street Money (@WallStreetSilv) April 19, 2023
Although what is said is technically correct, it sounds much worse than it is.
First and foremost, it would allow only affect Fannie Mae and Freddie Mac loans. This accounts for most loans made to homeowners, but would not affect FHA and VA loans or non-conforming loans that many investors get.
The costs referred to here are called the Price adjustment at loan level or LLPA, which primarily takes into account the borrower’s FICO score and the mortgage’s LTV. To a lesser extent, it also takes into account whether the property is owner-occupied or not, whether it is a condo or single-family residence, whether it is a second or a first mortgage and if there is a cash out during a refinancing.
The LLPA fees are then effectively added to the mortgage. So, for example, if the mortgage is $100,000 and has an LLPA of 1%, the LLPA would be $1,000. This could be paid as a fee, but is more often absorbed by the lender in exchange for a higher interest rate on the loan.
This additional cost on the mortgage is intended to cover Fannie Mae and Freddie Mac from the additional risk of lending to riskier borrowers.
The riskiest borrowers always pay more
The mistake made by many here is that the percentages given are the changes, not the totals. Well, not even quite that. The 1% fee mentioned is what someone with a FICO score of 740 would pay if they took out an 80-85% LTV loan. The 1.75% “reduction” is not the fee someone with a FICO score of 620 would pay, but rather the reduction in that fee from before. And in this case, it’s for someone taking out a 95% or higher LTV loan.
Before this rule was passed, the LLPA fee for someone with a FICO score of 620 taking out a 95% loan was 3.5%. It is now 1.75% (a reduction of 1.75%). Here is a table of Daily Mortgage News showing the effects that changes to this rule would have on loans for borrowers based on LTV and FICO score.
And here are the actual rates people would pay.
As Daily Mortgage News to summarize,
“As you can now clearly see, if you have a score of 640 you will pay significantly more than if you had a 740. Using a loan to value ratio of 80% as an example, your LLPA at 640 is 2.25 . % versus 0.875% for a score of 740. That’s a difference of 1.375%, or just over $4,000 on a $300,000 mortgage. It’s almost half of the previous difference, and it’s certainly a big change.
In fact, this rule change was made on January 1, 2023 and only went into effect now. Here is the announcement from the Federal Housing Finance Agency, and here is the full loan level pricing matrix from Fannie Mae herself.
The long and short story is, however, that those with low credit will always pay more than those with high credit. The real estate world has not been completely turned upside down.
Is it still a grant for people with low credit?
At the beginning of this article, I said that this new rule still involved those with good credit subsidizing those with trouble. Since those with good credit always pay less, how come?
The reason is that those with low credit scores are much more likely to default than those with good credit. And the difference is probably bigger than most people realize.
For example, a FICO white paper concluded that their model showed that “out of a score of 800, we would expect about 180 borrowers to consistently repay their loans on time for each defaulting borrower. This compares quite favorably to consumers with a score of 600, where one in 11 borrowers is expected to have payment problems.
Overall, the relationship they found between FICO scores and mortgage default rates was as follows:
another paper found that between 2000 and 2002, those with a FICO score of 750 or higher had a probability of default of only 1%, while those with a score of 600-649 had a default rate of 15.8%, and those under 500 had a default rate of a whopping 41%. Similar results were found in another study by the SEC of mortgages contracted between 1997 and 2009.
The overall result should come as no surprise, although the size of the gap might be too large (does the 2008 financial crisis make a bit more sense now?).
The LLPA is intended to cover some of this additional risk. But just looking at the table above, it would seem that even the old LLPAs were a bit generous (especially considering the average loss a bank takes on a mortgage that is foreclosed on is something like 40%). Reducing the LLPA for subprime borrowers will likely increase costs even further for Fannie and Freddie. And as basic economics indicates, this loss should be offset by higher rates across the board, including for borrowers with high credit ratings.
So, it is true that this rule will likely mean that borrowers with high credit ratings will subsidize those with low credit ratings.
But no, the clickbait outrage headlines are fake. Borrowers with low credit scores will pay no less than borrowers with high credit scores. And it’s important to be specific about what exactly is happening.
Get the best financing
Quickly find and compare investor-friendly lenders that specialize in your unique investment strategy. It’s fast, free and easier than ever!
Note by BiggerPockets: These are opinions written by the author and do not necessarily represent the opinions of BiggerPockets.