‘Fix the errors and a lot more people will flow through the system’: The credit ecosystem may need a makeover

While Covid-19’s mark on the US is slowly starting to fade, credit invisibility remains as bold as ever.

Being credit invisible means having little to no credit history for lenders to evaluate whether or not they can trust you with a loan. Around 26 million Americans are considered credit invisible, according to research by the Consumer Financial Protection Bureau. 

Matt Harris is the CEO and co-founder of Bloom Credit, an API platform that acts as a bridge to create access between credit bureaus’ data and fintechs that want to offer credit products. According to him, the credit measuring ecosystem is one messy salad bowl of legacy systems, faulty credit measurements, and banking deserts. And minority communities, which are most likely to be in places with little to no access to FDIC-approved lenders, are most likely to suffer the consequences.

“There’s a system that’s kind of stacked against consumers being able to create credit identity, specifically in minority communities,” said Harris at Tearsheet’s DataDay Conference in May. “And we kind of just discount it. We don’t really have a real conversation about how to establish these things.”

Credit bureaus’ legacy systems may be part of the problem. With infrastructure that needs a serious makeover, the bureaus can only find people that have the specific financial footprints their tech can actually detect. The rest remain ghosts of credit past. 

“Bureaus are ahead of the game in terms of data collections but struggle in modernizing their tech,” said Harris.

And with nowhere else to go in terms of getting a loan, lots of people who are considered credit invisible are turning to alternative lenders. 

But alternative lenders may not have the best ways of measuring data and that tends to lead to inaccurate credit scoring. For now, credit bureaus still have alternative lenders beat in terms of credit measuring. 

“You see some companies attempt to launch alternative credit scores in the United States,”  said Harris. “But the reason why these end up falling flat on their faces is because they’re never going to get the reporting structure necessary to actually be able to get the data they need.”

On the opposite side, simply stepping back and modernizing credit bureaus’ tech may not be on the list of viable solutions, either. According to Harris, bureaus are ‘where the buck stops,’ which means that most of their resources and prioritizations are going into making sure companies they’re working with are being compliant. 

“Bureaus get a lot of flack for not being innovative and I think that’s fair in some way,” said Harris. “But I think what a lot of people tend to discount is the reason why the bureaus tend to do a lot of what they do, which is that credit is a highly legislated section of our economy.” 

But reinventing the bureau isn’t really an option, either. With all the steps that need to be taken digging into old tech, it’s a risky endeavor. The chances of not making a giant mistake along the way are very small.

“It’s really difficult not to get it wrong because you’re working in 20 to 30 year old bureau infrastructure that hasn’t been updated. If let’s say you’re trying to get a system of records that’s written in Python to communicate with something that looks more like COBOL — then you’re going to have to do five simultaneous compliance checks against it,” said Harris. “Long story short, it doesn’t become a major focus for lenders, because it’s a complicated thing to do.”

You also can’t tell lenders to fix the errors in their systems on their own because they may not even know what the errors are. This has to do with the fraudulent disputes blurring their data mishaps.

“If I’m a lender, I’m dealing with thousands of fraudulent disputes, which actually prevent me from being able to fix the errors that are happening,” said Harris.

Instead of trying to suss out and fix the errors, Harris recommends working on preventing the errors that plague credit data in the first place. “The way we look at it is fix the errors and you’re going to start seeing a lot more people properly flow through the system,” he said. “And also when there are errors, they’ll get fixed a lot more easily because ultimately, there will be less errors, which means we can handle the ones that are frivolous a lot more effectively, and actually get to the ones where people are not being properly represented.”

Bloom Credit sits on top of credit bureau infrastructure. That makes it easier for technology firms to build more modern use cases for the underlying data. Harris says the overall credit ecosystem needs more investment in technology. It’s due for an overhaul. 

“From an infrastructure point of view, imagine we’ve built this really extravagant house. It has all these rooms, a lot of people live there, but the sink is clogged,” said Harris. “Rather than saying ‘the house is completely ruined, we need to move to another state,’ why not just unclog the sink?”