The Unexpected Winners of UWM’s Ultimatum

On March 4, ’21, Mat Ishbia gave mortgage brokers 11 days to choose: UWM or Rocket. They couldn’t work with both.

He called it the “All-In” initiative. I was sitting in my wife’s car outside a fertility clinic 🐣 on Long Island when the news broke. A very weird way to experience a major mortgage-industry event.

Anyway, “All-In” immediately divided the industry into two camps. One saw UWM as a paternalistic champion protecting the broker community. The other saw a nakedly anticompetitive ultimatum designed to kneecap 🤕 its biggest rival.

A new academic study suggests the reality was more complicated, & that some of the biggest winners may have been borrowers who never received a loan from UWM or Rocket.

Also in today’s edition: the antitrust case against Optimal Blue is losing steam, we track the biggest buyers & sellers of MSRs in Q1 ’26, and — exclusively for Mortgage Scoop Insiders — reveal which lenders are piling up non-QM/DSCR volume & who’s ceding ground.

The working paper, “Incomplete Exclusivity: Pricing Spillovers in Brokered Mortgages,” comes from Spencer Stone, a finance researcher at the University of Kentucky. Stone studied what happened to wholesale lenders that weren’t directly targeted by “All-In” after UWM-affiliated brokers were barred from sending loans to Rocket & Fairway. Violators faced a penalty of $5K per loan or $50K, whichever was greater.

Stone found that other wholesale lenders cut their rates by an average of 5 bps almost immediately after “All-In” took effect. The discount persisted throughout the full four-month period he studied.

On a $300K mortgage at roughly 3%, that translates to more than $500 in present-value savings for the borrower. Notably, the savings came through lower interest rates rather than reduced fees, suggesting lenders — not brokers — were giving up margin.

Why would UWM’s contract with its own brokers make competing lenders cheaper?Stone’s theory is that the ultimatum displaced rate-sensitive borrowers who otherwise would have matched with Rocket through the broker channel. That highly elastic demand instead flowed to smaller wholesalers, which sharpened their pricing to win the business. To isolate the effect, Stone compared wholesale and retail pricing within the same lenders for observably identical borrowers, using public HMDA and GSE loan data.

IMO, the most surprising finding is who benefited. Exposure to “The Ultimatum” varied dramatically by geography. Some metros had heavy overlap between UWM & Rocket brokers; others had almost none. Yet the price reductions were remarkably consistent across markets.

Stone argues lenders possessed enough data to target discounts toward the markets most affected by “All-In,” but likely couldn’t do so safely. His exposure measure correlates w/ borrower race & ethnicity, meaning geographic price differences could have invited disparate-impact scrutiny. The result was a positive spillover: borrowers received lower rates even in markets where the ultimatum had little direct effect.

The paper doesn’t relitigate the legality of “All-In.” UWM has prevailed in pretty much every case so far, though Ishbia is still expected to testify in one prominent lawsuit.

But Stone argues the episode offers a broader lesson for regulators. Because an exclusivity agreement can rarely eliminate every competitor, it may reshape pricing among the rivals left outside the contract. The same dynamic could apply to noncompetes, technology licensing agreements & manufacturer-distributor relationships.

Look, I think “The Ultimatum” was a major reason why UWM captured a massive amount of market share. But there are legit questions about how effective the strategy is today, & how sustainable it will be over the long term.

Back in the fall, a borrower in Tennessee filed a whopper of a class-action lawsuit against PPE heavy Optimal Blue. Mendez v. Optimal Blue accuses the country’s top mortgage lenders (Rocket, UWM, Wells Fargo, etc.) & the industry’s dominant PPE of colluding to keep rates & fees artificially high nationwide, in violation of the Sherman Act. Plaintiffs allege that the lenders & OB are acting like a “cartel,” in a suit that rhymes closely w/ similar actions brought against multifamily software giant RealPage & some of the country’s largest apartment landlords.  

The case is sprawling & I haven’t heard much in a while, so I figured I’d share some updates from court filings —> 

The defendants' first formal response came in January, when all 29 originally named companies told the court that OB's “Competitive Analytics and Competitive Data License” tools provide "backward-looking, aggregated, anonymized, and de-identified data" — info even the Federal Reserve uses(!). They also noted that several named lenders never used the products at all, & pressed plaintiffs to drop them.

Looks like the tactic worked. The amended complaint filed Feb. 23 cut the defendant list from 29 to 12, dismissing Rocket Mortgage, UWM, Wells Fargo, Chase, BofA, Citibank, U.S. Bank, loanDepot, Fairway, Freedom, NewRez, Flagstar, Mr. Cooper, CMG, AmeriSave, Better, & FirstBank. 

Per the suit, that left Rate, CCM, PennyMac, Guild, NAF, Churchill, First Community, Movement & Beeline. The remaining defendants answered on April 30 w/ a coordinated dismissal package. A joint motion argues plaintiffs still fail to state a plausible conspiracy, asserting the complaint never alleges OB told lenders how to use the data, encouraged price increases, or monitored compliance. Black Knight/ICE filed separately, arguing its former majority ownership of Optimal Blue cannot make it a conspirator. Constellation Software, Optimal Blue's Canadian parent, moved to dismiss for lack of personal jurisdiction.

Under the current schedule, trial would not arrive before ‘29. So yeah, bring a book to read.

Roughly $137.8 billion in unpaid principal balance changed hands through servicing transfers in Q1'26, representing 1.6% of total servicing outstanding. This is a decrease from $209.5 billion in Q4'25.

Carrington Mortgage Services & Truist Bank were among the most active buyers, expanding their portfolios by $36.8 billion and $16.9 billion, respectively. PennyMac ($24.2 billion) and Rocket ($20.3 billion) were the largest sellers. Overall, nonbank servicers accounted for 64.2% of MSR purchase activity, consistent with their growing role in both agency and government servicing markets.

Notable buyers in the quarter who had not been top 10 MSR buyers in the previous 12 months include Truist Bank ($16.9 billion), PNC Bank ($8.0 billion), and Huntington National ($5.5 billion). Notable sellers in the quarter who had not been top 10 MSR sellers in the previous 12 months include PennyMac ($24.2 billion), The Money Source ($12.5 billion), and Essex Mortgage ($12.0 billion).

Low weighted average coupon (WAC) loans were the most actively traded loans in the quarter. The largest interest rate bucket of MSRs transferred was 2.5-2.99%, accounting for 23% of transfers compared to 22% of total serviced volume, followed by MSRs with an interest rate between 3.0-3.49%, accounting for 19% of transfers compared to 16% of total serviced volume. The largest origination vintages of MSRs transferred were the 2021 and 2020 vintages.

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, Freddie Mac, and Ginnie Mae MBS Data Disclosure; servicing transfers exclude M&A activity