Fannie Mae and Freddie Mac have published official pricing grids for single-family mortgages using the traditional Classic FICO and new VantageScore 4.0 models. Early analyses suggest the new credit score model could result in higher borrower costs in many scenarios.
The updates, released Wednesday as the Federal Housing Finance Agency (FHFA) opened delivery of VantageScore 4.0 loans to all lenders, formalize a temporary workaround used in a limited rollout: Price adjustments designed for FICO are applied to VantageScore 4.0 at 20 points higher. The top purchase tier for FICO remains 780 and above, which now aligns with an 800-plus VantageScore bucket.
According to sources at the government-sponsored enterprises (GSEs), they expect competition between VantageScore and FICO to reduce credit report and borrowing costs over time while making mortgages available for “credit invisible” borrowers by considering rent payments and trended data in new models.
The sources added that they’ve noticed VantageScore scores are higher than FICO scores, leading to lower loan-level price adjustments (LLPAs) and costs for borrowers. So far, however, only a small share of loans delivered to the GSEs used VantageScore 4.0. Keefe, Bruyette & Woods reported that VS 4.0 represented 5.6% of originations in August.
Spokespersons for the FHFA, the GSEs and VantageScore did not immediately reply to HousingWire‘s requests for comments.
Borrowing costs
A Pivot Financial analysis of the new pricing grids concludes that while VantageScore 4.0 may save lenders money on credit score costs, it can make loans “materially more” expensive for borrowers in most cases over the life of the loan.
Pivot’s analysis provides complete LLPA grid outcomes for the GSEs. It models $400,000, 30-year fixed-rate mortgages, first with a base rate of 6.71% and then adjusted upward with the additional VantageScore pricing adjustments (ranging from +0.125% to +1.125%), which is defined as the “extra hit” in pricing (calculated as VantageScore 4.0’s LLPA minus Classic FICO’s LLPA).
The firm found that the impact on borrowers is most severe in lower- to middle-score ranges and in higher-risk cash-out refinance scenarios.
For low loan-to-value (LTV) loans below 30%, Pivot found there is essentially no difference between the FICO and VantageScore grids. But for purchase and refinance loans at 75% to 80% LTVs and credit scores of 700 to 719, the difference reaches about 0.375 percentage points in additional LLPAs for loans purchased by either GSE.
On the $400,000 example loan, that translates into an increase of $100.32 in the monthly mortgage payment and about $36,114 in additional interest over 30 years.
The largest modeled gap appears in cash-out refinances with 70.01% to 75% LTVs and 680 to 699 credit scores. In that segment, Pivot calculated an LLPA difference of 1.125 percentage points, which equates to a payment increase of $305.41 per month and nearly $110,000 in additional interest over the life of the loan.
“The GSEs do not know how VantageScore is going to perform,” said Jennifer McGuinness-Lubbert, CEO of Pivot Financial. “These are unknown risk adjustments and although the data from the credit bureaus feed these two models, the models are materially different: a 750 FICO score is not the same as a 750 VantageScore.”
Mitigating risk
Milliman, an actuarial and financial consulting firm, examined the pricing issue using publicly available mortgage-backed securities data from Fannie and Freddie for loans originated between 2013 and 2023, when only Classic FICO was delivered. The firm leveraged the GSEs’ historical disclosures across multiple credit score models
The firm compared how often a loan’s LLPA would be lower, the same or higher assuming the 20-point adjustment to VantageScore 4.0.
With the 20-point adjustment, Milliman found that about 25% of the time, VantageScore 4.0 produced a lower LLPA than Classic FICO. The pricing was the same roughly 36% of the time, while in about 40% of cases, the borrower would have received a cheaper LLPA using Classic FICO.
Jonathan Glowacki, a principal with Milliman’s mortgage group, said the calibration for VS 4.0 could reflect both uncertainty and the “lender choice” policy, which allows originators to price and deliver a loan using either score.
“It creates adverse selection, meaning that the originator has more information than Freddie and Fannie when the loan is priced and delivered; what happens is the lender is incentivized to use the highest score of the two,” Glowacki said. “The 20-point adjustment could mitigate some of that risk.”
