For more than 30 years, the FICO credit score has been the standard for evaluating credit risk in consumer credit markets — from credit cards to auto loans. Because of its wide adoption by lenders, Fannie Mae and Freddie Mac, the government sponsored enterprises (GSEs) subsequently adopted the FICO score for use in originating conforming mortgages.
In the decades since, mortgage originators, investors and the housing-finance industry at large have come to use the FICO score as the independent, reliable indicator of credit risk in the mortgage market. In recent years, at the direction of the Federal Housing Finance Agency (FHFA), the GSEs tested newer credit-scoring models to determine their predictive potential in the housing market.
The FHFA in 2017 then initiated a “request for input” to engage stakeholders and carefully determine the operational and competitive considerations necessary for changing or updating the GSEs credit-score requirements. Before FHFA could complete the decision making process, Congress intervened, requiring a new rulemaking process as part of the Economic Growth, Regulatory Relief, and Consumer Protection Act. This delayed a final decision on the matter, and FHFA is now developing a new proposed rule.
FHFA is the appropriate entity to regularly test credit-scoring models because taxpayers ultimately hold the associated risk, unlike other credit markets where lenders bear the risk, such as credit cards or auto loans. FHFA can use the rulemaking process established by Congress to make meaningful improvements to the current system, including the use of a more predictive score.
As part of this rulemaking process, any change in the standard measure of credit risk in the mortgage market also must include a careful consideration of the impact to the entire housing-finance ecosystem, including increased costs to mortgage originators and private mortgage insurers. FHFA will be soliciting public comment on this rulemaking, and there is an opportunity for mortgage originators to lend their voice to this debate.
The FICO credit score has long been the industry standard as a reliable predictor of credit risk. And it’s not just for mortgages. According to research by the Mercator Advisory Group, in 2016 FICO credit scores were used in 90 percent of the lending decisions related to automobile, credit card and mortgage financing. There are no requirements to use the FICO score in these markets. Lenders and investors trust the FICO score to be independent, predictive and reliable.
The only alternative to FICO that has been considered by FHFA is VantageScore, a joint venture of the three main credit bureaus that claims to be able to expand access to mortgages. VantageScore lowers the minimum scoring criteria, which have long been a source of reliable performance in credit scoring.
VantageScore’s approach includes scoring a borrower seeking a mortgage with as little as one month of credit behavior or a credit report that hasn’t been updated in more than 24 months. Should lower standards be accepted, this introduces potentially dangerous unreliability and risk into the mortgage market, which has no historical performance data to truly reflect the risk associated with these mortgage originations.
This past December, the FHFA published a proposed rule — which is still subject to public comment and final approval — that calls for prohibiting the GSEs from adopting VantageScore or other credit-scoring models developed by credit-data providers such as the major credit bureaus, given the potential conflicts of interest posed by such an ownership structure. Regardless of the fate of that proposed FHFA rule, there is a better way to responsibly expand access to credit for borrowers and provide a bridge to homeownership. By innovatively using data which is not housed in traditional credit-bureau files, the credit box can be expanded while maintaining sound underwriting.
To do this requires new data sources that meet industry standards while properly assessing a borrower’s credit risk and assigning a meaningful and reliable score to people with thin or stale credit-bureau files. FICO Score 9, for example, evaluates medical debt differently from other debts, increasing the score’s predictiveness and further increasing access to credit. In addition, the recently announced UltraFICO Score pilot allows borrowers, for the first time ever, to contribute their checking, savings or money market account data to potentially enhance their FICO scores.
Expanding access by innovatively using trusted data will benefit both consumers and industry stakeholders, building on the integrity of the system. This is in contrast to simply lowering scoring standards, which could revive the predatory-lending practices that led to the financial crisis.
It’s also important to evaluate the cost associated with introducing alternative- scoring models for industry participants across the mortgage life cycle. The transitional costs alone would be significant, with stakeholders facing the prospect of large-scale system adjustments and model management oversight. As the Independent Community Bankers of America noted, “the shift to an entirely different score brings to bear higher transitional costs that disproportionately affect community banks without providing any real benefits for making the change.”
These costs could be exacerbated by the fact that while models such as VantageScore replicate the well- understood FICO score range, the scores do not share the same odds-to-score relationship, meaning the risk at a given score is different, and that difference can fluctuate over time. Such large differences will not only impact segments around the minimum score cutoff for approval, but throughout the entire population for pricing.
In order for the industry to accept a different score, it will require pricing and actuarial model changes, forward/backward compatibility testing, and reporting and disclosure changes, ultimately leading to a higher cost of mortgage credit for both the industry and consumers. It’s important to note that because it shares the same odds-to-score relationship as the classic FICO scoring system — which is the score currently in use for conforming mortgages — adopting FICO Score 9 would allow the industry to adopt a more predictive model with much lower transition costs.
The benefits of introducing a secondary score like VantageScore are negligible, according to Quantalytic and FHFA. A study by Tom Parrent and George Haman of Quantalytic found that 99.9 percent of new scores generated by VantageScore would likely not result in new mortgages annually. Furthermore, an evaluation by FHFA “revealed only marginal benefits to requiring a different credit score than classic FICO.”
For decades, the FICO credit score has provided an independent indicator of risk that investors trust. A shift to a model such as VantageScore, which is owned by the three major credit bureaus — Equifax, Experian and TransUnion — would primarily serve to consolidate the credit bureaus’ power in the mortgage market.
• • •
Unlike other consumer-credit markets, the GSEs’ tri-bureau requirement that mortgage lenders pull prospective borrowers’ credit reports from all three main bureaus — Equifax, Experian and TransUnion — during the mortgage origination process means that the credit bureaus do not compete with each other in the conforming-mortgage market. This lack of competition would give the credit bureaus broad power in a system where they have control over the pricing and distribution of both credit reports and scores, which would only increase if the GSEs accepted VantageScore.
Given these high costs and the marginal benefits, any changes to credit scoring undertaken by FHFA and the GSEs should be subject to a robust, independent cost-benefit analysis examining the impact on all the participants in the mortgage market, from originators to insurers to consumers.