A Very Sneaky Way to Avoid an EPO

On Friday, I shared a screenshot of a post from a popular mortgage FB group showing what appeared to be a broker trying to dodge an EPO from the investor. The setup is a little convoluted, so bear with me.

The borrower refi’d two months ago into a 15-year w/ a 2-1 buydown & a higher note rate. The plan, apparently, was to make six payments, refi again & roll the remaining buydown funds into the new loan. Then, ah shit: the borrower’s spouse loses their job 🙃. Suddenly they need a 30-year just to make the monthly payment work.

And here’s where things get weird. The borrower apparently signed a side agreement requiring them to repay the loan costs, the lender credit used for the buydown & the origination charges if they refinance early. There’s just one small problem: the Closing Disclosure explicitly says there is NO PREPAYMENT PENALTY.

Still w/ me? Good, because this whole mishegas raises a bunch of interesting questions, which we’ll tackle today.

Elsewhere in this edition: PMI is clawing back share after a soft Q1, RETR has fresh numbers on what acquirers actually lose when LOs walk post-close, a big Sierra Pacific/UHM team defected to AnnieMac & I hear that a family-owned Midwest IMB is quietly up for sale.

For starters: Why was this borrower put into a 15-year in the first place?

“I wonder if there was a gigantic rate/cost benefit closing the original mortgage on a 15-year term,” said Jordan Hamling. “It's been years since I've seen a rate/cost scenario on a 15-year that made more sense than just closing on a 30-year (flexibility, invest money elsewhere, etc.).”

From the broker’s perspective, the EPO problem is real. You spend weeks getting a loan closed, the borrower refinances almost immediately & suddenly you’re writing a check back to the lender. “There’s some mortgage companies that go under because they can’t afford to pay back EPOs or EPDs,” Nathan Gort noted.

That sucks, no doubt. But, uhhh, you probably can’t solve it by having the borrower sign something that looks an awful lot like a prepayment penalty while simultaneously giving them a CD that says NO PREPAYMENT PENALTY.

“It’s deceptive, unenforceable & sadly so many try to delay client prepayments to keep commissions & corp revenue,” said Phil Tocci. “Our worst enemies to our business are within our industry.”

There are still some unanswered questions, like was it a cash-out refinance? Did the borrowers qualify w/ the lower 30-year fixed rate? Mike Lee recommended that the borrowers directly contact the lender & servicer.

“The upside for the lender is to retain the customer, avoid repurchasing the loan if it defaults, & to do the right thing, including following up w/ this broker,” he said. “Put out the fire first - lender should do everything to assist under the circumstances, especially eisnce the broker is an approved originator for the lender.”

There is a broader potential solution, though I suspect borrowers wouldn’t exactly love it. “There's the argument to just add 6-month prepays to all loans,” said Colin Robertson.

Private mortgage insurance ("PMI") activity grew in Q2'26. During the quarter, PMI covered $78.5 billion original unpaid principal balance ("Original UPB") of newly insured GSE mortgage-backed securities compared to $70.3 billion in Q1'26 and $66.2 billion in Q2'25; accounting for 36.5% of total Fannie Mae and Freddie Mac issuance, increasing from 34.6% in Q1'26 and decreasing from 38.5% in Q2'25.

The share of purchase money mortgages with PMI was 48.0% in Q2'26, compared to 47.3% in Q1'26 and 46.1% in Q2'25. The share of refinance loans with MI was 16.8%, compared to 17.1% in Q1'26 and 11.6% in Q2'25.

Average FICO score of insured mortgages was 757 in Q2'26, compared to 756 in the prior period and 755 a year prior.

The share of mortgages under $200K with MI was 21.9%, while loans with a balance between $200k-$300k was 34.5%, and loans with a balance greater than $300k was 38.8%.

This report is brought to you by Milliman. Milliman provides strategic and quantitative consulting services across the mortgage market with expertise in origination, servicing and capital investment.

Source: Fannie Mae and Freddie Mac MBS Data Disclosure. Note: This data is separate and independent from Milliman's quarterly Private Mortgage Insurer (PMI) industry report. Quarterly totals do not incorporate or relate to insurer-specific financial statements or published results.

Have you checked out RETR’s analysis detailing how well acquiring companies do at retaining LOs post-acquisition? If not, you should 100% check it out. It’s excellent. 

The Good People of Mortgage wanted to know whether buyers were shedding low producers, or if real originators were walking too?

Quick answer: both! RETR looked at the current rolling 14-month production of all 1,165 LOs who transferred across the eight acquisitions in its study. Among LOs still active today, retention climbs steadily w/ production. So basically, the bigger the book, the likelier they're still w/ the buyer. Same pattern on volume.

But here's the other half. Those active transferred LOs have collectively closed north of 30,000 loans & $11.3B over the trailing 14 months. The acquirers hold 21,946 loans & $8.13B of it. Competitors hold 8,233 loans & $3.23B, roughly 27% of the loan count & 28% of the volume.

Per the study, it's not just marginal talent leaving. Thirty LOs currently doing 61+ loans went to another lender. Sixteen departed LOs are running 100+ loans at another shop. Pair that w/ Part 3, where exits kept clustering at the same competitors, & you're looking at entire teams leaving, not stragglers.

One caveat I previously noted is retention bonuses. Top producers typically get them, often w/ two- or three-year repayment provisions. RETR can't see who got what, so that’s obviously going to muddy the waters a bit. But it’s definitely a factor.

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