The U.S. housing market is weighed down by a vast inventory of low-rate mortgages that could help unfreeze sales, but a fundamental flaw in how Multiple Listing Service (MLS) data is handled means buyers rarely see these opportunities.
Andy Taylor, founder and CEO of RetroRate, estimates that about 22% of all homes with mortgages have assumable loans—financing that buyers could take over, often at rates far below current market levels. Yet these homes remain invisible on major sites like Zillow, Redfin, and local broker platforms.
“It really makes me sad every time a 6.3% loan is originated when it could have been assumed,” Taylor says. “It’s just a total loss. It really is. It’s value destruction.”
The Scale of the Hidden Inventory
Taylor highlights the magnitude of the opportunity: “Right now, 70% of all outstanding mortgages have a rate that’s below 5%,” he explains. “And when I say below 5%, I mean they’ve got a four-point-something rate.”
This concentration of low-rate mortgages is a legacy of the years following the 2008 housing crash and the COVID-19 pandemic, when millions of homeowners locked in historically low rates. These loans, mostly originated during periods of near-zero interest rates, are still active.
Taylor notes that government-backed loans—VA, FHA, and USDA mortgages—are typically assumable, allowing a qualified buyer to take over the seller’s existing loan. “There are probably 22 to 25% of all the homes with a mortgage that have an assumable loan,” Taylor says. Despite this, most of these favorable loans go unused by new buyers.
Why the Data Never Reaches Buyers
The main obstacle, according to Taylor, is a structural flaw in how MLS systems collect and display mortgage information. Assumable loan details are stored in an optional field that listing agents must manually fill out. “The listing agent usually doesn’t ask the seller if they have an assumable loan,” Taylor explains. “And honestly, the seller wouldn’t know if they had an assumable loan if you asked them.”
Because this information is rarely entered in the MLS, it never makes its way to the IDX feeds that power broker portals and consumer websites. “It’s never on Redfin or Zillow, or any of the other sites,” Taylor says. “So because of this data gap, nobody knows these loans are available.”
Taylor describes these properties as “gems that are out in the open that people aren’t really looking for.”
This lack of visibility has a ripple effect. Buyers miss out on the chance to assume mortgages with rates often 2 to 3 percentage points below what lenders currently offer. Sellers lose a key selling point that could attract more buyers and potentially raise their sale price. The broader market remains locked in a rate-driven slowdown that might be partially relieved if these loans were more visible.
The Data Infrastructure Challenge
Addressing the visibility issue is not as simple as pulling mortgage data from public records. “I thought there would be some off-the-shelf data I could just plug in,” Taylor recalls. “But you really have to clean it up pretty hard to make it useful.”
Taylor describes the real estate data landscape as a “patchwork quilt” of APIs and sources, each with varying quality. “Some have really great fidelity. Some are poor. And it depends on which state you’re in,” he explains. By the time data reaches a platform, it has often been handled by hundreds of agents and county clerks.
To improve accuracy, RetroRate uses multiple redundant data sources and a waterfall model that ranks sources and assigns confidence scores. “If you don’t do that, sure, you can get something up and running. But the moment an agent sees an error—like showing a 5.875% rate when the home actually has a 4.25%—they’ll never trust you again,” Taylor warns.
A Solution Focused on Listing Agents
RetroRate’s solution targets the moment when listing agents enter property information into the MLS. “As a listing agent is typing in 123 Main Street, just before listing, we can flag it and highlight: Hey, this home has an assumable loan,” Taylor explains.
The company is partnering with MLS organizations to integrate this data layer as an add-on to existing systems. Taylor emphasizes the competitive advantage for agents: “If you put this into the MLS, it syndicates out to all the real estate sites that use that data. That’s free advertising.”
RetroRate recently announced a partnership with Beaches MLS and aims to expand to 10 of the top 25 MLS organizations by 2026. The company’s API-based system allows the data to be accessed in multiple ways beyond just MLS integration.
Taylor believes the issue will persist even if rates decline. “Let’s say we get to 6% by the end of 2026—there will still be 80% of loans out there with better rates, in the form of assumables,” he says. As more listing agents surface this information, Taylor expects visibility and demand to increase in a “snowballing” effect.
Whether this approach is widely adopted by MLS organizations may determine if the industry can finally make this hidden inventory accessible to buyers—unlocking billions in potential savings and helping to ease the housing market’s current gridlock.
