I’m writing this after six weeks in Europe, where I spent time with more than a thousand agents from 25 countries across three in-person events, masterminding the issues they’re facing in their own markets. Most of it will sound familiar to anyone in this business: high inflation, affordability strain in a post-COVID world, thin inventory, social media, AI.
But the one problem that came up in every single room, from the UK to Portugal, Italy, Spain and France, was data. Every one of them wanted better access to closed transactions and to what’s actually available for sale. And every one of them was envious, genuinely envious, of the MLS system we have in the U.S. and Canada.
For the past 18 months, the fight over how homes get marketed has been fought on the terrain of fairness: whose listings get seen, whose don’t and who profits from the gap.
I’ve made my position clear over and over: A marketplace that lets sellers hide inventory in private networks isn’t consumer choice; it’s consumer cost, and it falls hardest on the buyers who already have the least access, first-time buyers and buyers without an insider’s Rolodex.
But there’s a question in this debate that hasn’t gotten nearly enough attention, and it cuts to the heart of consumer fairness: whether the numbers underneath every mortgage in America are still going to add up.
What this could do to every mortgage in America
Home lending in this country runs on comps. When an appraiser values a house for a loan, the rule is simple: at least three settled, comparable sales, pulled overwhelmingly from the MLS. That agreement is the plumbing of the entire U.S. mortgage system.
It’s why an appraiser in Reston, Virginia, can trust a comp as much as one in Sacramento, California. It’s the same plumbing underneath more than $13 trillion in outstanding U.S. mortgage debt, every dollar of it underwritten on the assumption that the comps behind it are real, complete and current.
That system only works when virtually every sale flows into the same shared pool. Increasingly, it doesn’t. A growing share of transactions now move through private, closed marketing arrangements before, or instead of, ever reaching the open market.
Some of those sales never make it into the shared system at all. Others show up late or missing the very fields, price history, days on market, that tell an appraiser whether a price reflects the real market or a fluke.
I recently spoke with a veteran appraiser about what this actually looks like on the ground. His answer has stuck with me: When a comp three doors down never makes it into the shared pool, he can’t use it. Not because the price wasn’t real, but because he can’t verify it.
Time on market and price history aren’t noise; they’re how an appraiser checks whether a sale price reflects the real market or a fluke. Hand him a comp with only bare-bones information, and a careful appraiser won’t touch it and may fall back to tax records instead, which lag the market by months.
A less careful one might use it anyway, on half the picture, and the homeowner is the one who eats that liability when the number comes in wrong, not the brokerage that kept the sale private.
He also raised something worth sitting with: Appraisers already spend hours tracking down comps by calling agents one by one. That’s manageable when it means six calls to six agents inside one shared system. It stops being manageable when it means six separate offices, each running its own private network with its own rules.
The friction doesn’t just slow appraisers down; it quietly sorts them at the consumer’s expense. The professionals who do the extra digging can still get to a reliable number, at a cost. The ones who can’t, or won’t, default to whatever comps happen to be easy to find, whether or not they’re representative.
Multiply that across every market where a dominant brokerage runs its own parallel comp pool, and you get a national appraisal dataset with holes in it. Different holes in different metros, depending on local market share.
And those holes are opening at the exact moment federal regulators are pushing appraisers toward more automated, data-driven valuations. Fragment the comps, and you’re not just making appraisals harder; you’re degrading the inputs.
That should worry Freddie and Fannie. It should worry FHFA. The Consumer Federation of America has already asked federal regulators to examine the private-listing partnerships reshaping how homes get marketed.
Beyond the fair housing and market power concerns they raised, I’d add this: What do fragmented comps do to the reliability of the appraisal itself, and what does that mean for the confidence lenders and investors place in the mortgage-backed securities built on those appraisals?
The National Association of Realtors and the Council of MLSs have told the Department of Justice and the Federal Trade Commission that MLSs are “critical infrastructure for the housing market.” So what does a patchwork of regional private listing networks do to comp reliability and appraisal confidence at scale?
We built the most liquid, accurate and complete housing dataset on the planet on a simple rule: If you want to pull from the cooperative pool, you contribute to it. That bargain is what let lenders trust appraisals, let investors trust mortgage bonds and let this country underwrite homeownership at a scale nowhere else in the world can match.
We shouldn’t give it up one regional data deal at a time.
Leo Pareja is CEO of eXp Realty.
