VantageScore 4.0 is moving into mortgage. Lenders need to understand what’s actually changing.
For our latest Calyx Sunlight Series, we sat down with Jason C. Jefferies, Director of Client Success – Mortgage Division at VantageScore, and Dr. Andrada Pacheco, EVP & Chief Data Scientist at VantageScore, to go underneath the score and talk about how the model actually works.
There’s a practical reason we wanted to have this conversation.
This isn’t just another number on a credit report.
The model uses different data, can score consumers who may not receive a score under older models and introduces new questions around workflow, policy, training and technology.
This is bigger than a credit score change. It’s an operations conversation.
▶ Watch: VantageScore 4.0 Is Here. What Does It Mean for Mortgage Lenders?
We covered trended data, thin-file borrowers, score differences, rental history and what lenders should think about before bringing another credit model into their operation.
If you missed it live, start with the full conversation. Then use the recap below for the points we’d make sure your team understands.
One of the bigger differences is the use of trended credit data.
Traditional credit scoring has largely been built around a snapshot of a consumer’s credit profile at a particular moment. This model can look at up to two years of account-level history.
That matters because two borrowers can look similar today and still have very different credit behavior behind the number.
One borrower may be steadily paying balances down. Another may be moving in the opposite direction.
During the Sunlight session, Dr. Pacheco compared static credit data to looking in a mirror. You can see where someone is now. Trended data helps show how they got there.
That context is one of the more important things for lending teams to understand.
VantageScore estimates that its methodology can score approximately 33 million additional consumers who may not receive a score under some traditional models.
That includes consumers who are new to credit, have thin credit files or have had limited recent credit activity.
According to figures discussed during the session, roughly three-quarters of those newly scored consumers have a score of 620 or higher.
That doesn’t mean every one of those borrowers suddenly qualifies for a mortgage. Credit scoring is only one part of underwriting.
But it may give lenders additional information for consumers who were previously more difficult to evaluate. In a market where lenders are working hard for every qualified borrower, that’s worth understanding.
This may be the most important thing we heard during the conversation.
A VantageScore and a FICO score are not direct equivalents. They’re different models using different methodologies, weighting, segmentation and data.
So we wouldn’t recommend looking at a 680 under one model and assuming it means exactly the same thing as a 680 under another.
How does this model perform against the actual risk in your portfolio?
During the session, the team encouraged lenders to evaluate score ranges against their own historical performance and credit policies instead of trying to build a simple conversion between models.
If your organization begins evaluating a new scoring model, this should be a portfolio and risk conversation—not a point-for-point score comparison.
Calyx Path, Point and Zenly support VantageScore 4.0.
If you’re already using a Calyx solution, these quick tutorials show how to work with the score inside your product:
Technology support is only one piece of readiness. Lenders should still understand when their investors, credit providers and other lending partners support the model and how it fits into their own policies and processes.
The same underlying model is used across Experian, Equifax and TransUnion. The data at each bureau can still be different, but the scoring model applied to that data is consistent across all three.
Rental and other alternative credit data can also add context when that information is reported to the credit bureaus. A borrower may not have a deep traditional credit file but may have years of consistently paying rent.
The important distinction is that the model isn’t collecting that information directly from a landlord, utility company or phone provider. It has to make it into the credit bureau data first.
Still, it points toward a broader change happening in credit: more of the borrower’s financial behavior may become usable data.
Start with where the score enters your process.
When will you pull it? Who needs to see it? How will it appear in your LOS? And what happens if your team is working with more than one scoring model?
Then look at your policies. Are there automated conditions, pricing rules, overlays, disclosures or internal procedures built around specific score ranges?
Those things may not need to change. But they’re worth identifying before adoption gets further along.
Training matters too. Loan officers, processors and underwriters are going to get questions when borrowers see different scores or when lending partners handle the models differently.
And talk to your investors and lending partners. Not every organization is going to adopt a new model at the same speed.
For brokers and TPO channels especially, understanding which partners accept it and how they use it will matter.
We’d ask instead of assume.
There’s a bigger signal underneath all of this
Mortgage has operated around a familiar credit scoring system for a long time. That’s starting to change.
What we’re seeing now is part of a broader shift toward using more data, more historical behavior and more flexible ways of evaluating borrower risk.
That creates opportunity. It also creates more complexity.
When multiple models, datasets and lending requirements begin moving through the same operation, the technology, workflow and people behind the mortgage process need to be ready for that flexibility.
The score is only part of the change. The operation around the score matters just as much.
The full Sunlight conversation goes beyond the highlights above, including how lenders can evaluate the model against their own portfolios, what brokers should be asking their lending partners and questions submitted by mortgage professionals during the live session.
▶ [WATCH THE FULL SUNLIGHT SERIES RECORDING]
Don’t treat VantageScore 4.0 like somebody changed the formula behind the same old score.
Understand what data is different. Understand how the model evaluates risk. Understand which partners are using it. And understand what has to happen inside your operation if you decide to use it.
Credit scoring is changing.
The lenders that understand the change before they have to react to it will be in a much better position as adoption expands.