The MBA is sticking to its guns that the GSEs should transition from a tri-merge to a no-merge, & even Bill Pulte might be coming around to the idea. But it is not a popular position in some corners of the industry. 

The CHLA, which generally supports the MBA’s policy positions, is not a fan 🐽. Scott Olson recently authored a white paper expressing concern that a move to single-file would increase costs, make it harder for aggregators/investors to price loans & also hurt consumers that shop for multiple quotes. On Thursday, the MBA’s Bob Broeksmit rebutted his points in a buzzed-about blog post (at least on LinkedIn).

In today’s edition of The Scoop, we highlight the key disputes in the credit score civil war, go under the hood of Vesta’s new AI agents 🔒 & reveal some quasi-personal news: The Scoop now has an NMLS ID & is going to become an LO.

(🙏 If you like what you’re reading, tell a fellow mortgage junkie to sign up here.)

There is so much to unpack in Broeksmit’s post, so let’s review a couple highlights. I’ve edited these passages for length.

1) CHLA: Costs could increase for consumers that shop for mortgage rates. If one lender uses one score, one bureau – and another competing lender uses a different one – the borrower could potentially be charged two upfront credit charges to see the pricing differences between the two lenders.

MBA: It isn’t possible today to transfer a credit report pull from one lender to another. Under a single-file model, a borrower who shops at two lenders would pay for two credit reports, rather than paying for six today in a tri-merge framework. It is also extremely rare today for a borrower who is shopping to be charged upfront for a credit pull - though that may very well change if the credit bureaus are allowed to continue to exploit their monopoly & increase prices. 

2) CHLA: A switch to a single bureau model could make it harder for aggregators/investors to price loans – leading to risk premiums in loan pricing that raise rates. The lack of data on pricing of MSRs during the transition period could create risk and uncertainty – which will be priced into loans, as we saw during the COVID crisis. 

MBA: The current government-granted oligopoly system has seen credit costs rise by roughly 400% over 4 years, despite using precisely the same models and almost the same data. The research suggests that the variations, particularly at the top and in the middle of the score curve, are infrequent in the aggregate, and not significant enough to warrant the extra cost of a tri-merge for borrowers in those ranges. There will undoubtedly be a transition period and necessary investor education, but the long-run impacts should not result in risk premiums significant enough to increase costs for borrowers. Moreover, single-file framework would be an option, not a mandate. Any lender that found a competitive advantage from staying with a tri-merge could continue with a tri-merge.

3) CHLA: Not all data sources report to all 3 bureaus. Thus, undisclosed debt risk increases with just one bureau, one score. This could also lead to greater repurchase risk for lenders if they use a score that misses debt or other credit issues. 

MBA: The GSEs and our members have both reviewed and analyzed the same data on the variations in tradeline coverage between the 3 bureaus. Given the size of their books, the GSEs possess the most robust insights on tradeline and score variances. Based on our discussions with the GSEs and our prior work on the rep and warrant framework, we do not believe they would adopt a one-report policy while holding lenders responsible for data that appears on one of the other two bureaus that the lender was not required to obtain. We would insist on that approach as any other policy would defeat the objective of delivering meaningful cost relief to borrowers.

Personally, I found Kevin Stevens’ take thoughtful (h/t Greg Sher).

One of the things I talk about a lot at The Scoop is edge. Not marketing gimmicks or other oft-peddled bullshit, but actual operational edge. That’s why I plan to become a licensed mortgage originator in ‘26. 

To be clear: I’m not trying to chase volume or anything like that. I don’t plan to originate full-time. But I believe that originating on the side will give me a knowledge edge so I can deliver even better content to you guys.

I already have an NMLS ID (2806851). Over the next few months I’ll do the required 20 hours of education, take the SAFE MLO Exam (I’ve done well on practice tests) & then look for a company to hang my shingle.

I’m in New York—who would you recommend I work with? If readers want, I’ll document the process as I go…

After weeks of teasing demos & blog posts, Vesta on Thursday launched its AI agents in production for paying customers. Everyone & their grandma seems to be launching AI agents for mortgage, so I had to ask the Vesta crew: Is what they’re doing really unique? How does it compare to others in the LOS space?

The rest of this story is available exclusively for Mortgage Scoop Insiders.