By The Briefcase Team | April 29, 2026

In September, on an investor call most of America was not listening to, Compass CEO Robert Reffkin said the quietest part out loud.

"Our customers are the agents who work with home buyers and sellers."

Read that twice. The customer is not the buyer. The customer is not the seller. The customer is the agent. Buyers and sellers are, in this telling, the inventory the customer trades. It is the kind of sentence a CEO says when he forgets the microphone is on, except Reffkin said it on purpose, and the Wall Street Journal printed it, and almost nobody in the housing industry seemed to notice.

We noticed.

Because over the last twelve months, while the country was busy arguing about mortgage rates and the 10 million home shortage, the largest residential brokerage in America has been quietly assembling something that has never existed before: a parallel, private, off-MLS housing market that one company controls. They acquired the country's biggest brokerage holding company. They got the Department of Justice antitrust review they should have triggered, and then somehow did not get it. They sued Zillow, lost, and won anyway. They cut a deal with Rocket and Redfin to scrub price-drop history and days-on-market from half a million listings. And the academic data now shows that none of this helps the people whose houses are actually being sold.

This is the story of how that happened. Pour something strong. We are going in.

Quick civics refresher. The Department of Justice has a rule of thumb in antitrust: when one company controls more than 30 percent of a local market, regulators start asking questions. Above 50 percent, they start filing lawsuits. Below 30 percent, you are usually fine.

In January 2026, Compass closed its $1.6 billion acquisition of Anywhere Real Estate. Anywhere is the holding company that owns Coldwell Banker, Century 21, Sotheby's International Realty, Corcoran, ERA, and Better Homes and Gardens Real Estate. With one signature, Compass added six of the most recognizable brand names in American real estate to a roster that already included @properties, Christie's, and a handful of regional powerhouses.

The combined entity now touches roughly one in three home sales in the United States. In specific metros, HousingWire's market share analysis suggests Compass affiliates may already exceed the 30 percent DOJ threshold. In a few, they are flirting with 50.

The DOJ noticed. According to the Wall Street Journal, career antitrust enforcers wanted to open a formal investigation into the Compass-Anywhere merger. They were overruled by political appointees higher up the chain. The deal closed without a second look.

You do not need to be a conspiracy theorist to find this interesting. You just need to be awake.

To understand why the merger matters, you need to understand the Compass business model. It is elegant. It is also, depending on your priors, either brilliant or slightly terrifying.

Step one: recruit agents. In Q4 2025 alone, Compass added 830 principal agents, the largest single quarter of agent recruiting in the company's history. Full year 2025 revenue hit $7 billion. 2026 is projected at $13.68 billion with $780 million in EBITDA.

Step two: when those agents win a listing, route it through the Compass Private Exclusive network first, before it ever hits the MLS. As of September 2025, Compass had roughly 9,000 active Private Exclusive listings. A buyer wants to see one of those homes? They have to work with a Compass agent.

Step three: double-end the deal. Compass agent represents the seller. Compass agent represents the buyer. Compass collects both sides of the commission. In industry slang this is called "double ending," and Reffkin spent the better part of an April 16, 2026 LinkedIn post defending the practice from critics who pointed out that it is, by any reasonable definition, a conflict of interest.

Step four: repeat. Every recruited agent is a node. Every Private Exclusive is a moat. Every double-ended sale is two commissions instead of one. The flywheel spins faster as the network grows, which is exactly what venture-backed flywheels are supposed to do.

There is just one problem. The seller, in this story, is the product. Not the customer. Reffkin said so himself.

Now we get to the fun part.

In June 2025, Compass sued Zillow over Zillow's new Listing Access Standards, which require any home publicly marketed for sale to be posted to the MLS within one business day or be banned from Zillow for the life of the listing. Compass argued the rule was anticompetitive. Zillow argued it was the bare minimum required to keep American home buyers from being herded into walled gardens.

On February 6, 2026, a federal judge denied Compass's request for a preliminary injunction. The judge wrote that Compass was unlikely to succeed on the merits. In legal language, that is a polite way of saying your case is bad and you should feel bad.

Compass lost. Or did they?

Twenty days later, on February 26, 2026, Compass announced a "historic alliance" with Rocket Companies and Redfin. Under the deal, 500,000 listings would flow from Compass to Redfin (which Rocket had recently acquired). Read the fine print and the genius reveals itself: those listings flow to Redfin without days-on-market data and without price-drop history.

If you are a home buyer, those two data points are the most important information you have. Days on market tells you whether the seller is desperate. Price drop history tells you what the seller actually thinks the home is worth versus what they listed it for. Strip those out, and you are negotiating in the dark. The seller's agent, of course, is not in the dark. The seller's agent is Compass.

On March 17, 2026, Zillow softened its ban with a feature called Zillow Preview, allowing limited pre-marketing exposure under specific conditions. The next day, March 18, Compass dropped the lawsuit entirely.

So let us recount the score. Compass lost the lawsuit. Compass got the Rocket-Redfin deal. Zillow blinked and softened its rule. Compass dropped the case as if nothing had happened. The "loss" was the cover story. The deal was the heist.

You almost have to admire it.

Compass's pitch to sellers is simple: list privately first, test the market, find the perfect buyer, do not let your home become "stale" on the MLS. It is a great pitch. It is also, according to every neutral study we could find, mostly nonsense.

In April 2025, Bright MLS chief economist Lisa Sturtevant published a study of more than 100,000 home sales. The findings, in plain English:

Sturtevant's conclusion was the kind of sentence economists write when they are trying very hard not to swear in print: there are "no clear benefits to sellers."

A more recent University of Georgia preprint by economist Dr. Darren Hayunga, released in March 2026, looked at more than 700,000 Dallas-Fort Worth home sales and found a 1.7 percent premium for pocket listings before NAR's Clear Cooperation Policy took effect. After CCP, the premium evaporated. As HousingWire summarized it, "NAR's CCP didn't stop the practice. It just made it pointless financially."

Zillow's own research estimated that sellers using private listing networks left roughly $1 billion on the table in 2023 and 2024 combined. The Northwest MLS has formally accused Compass of "deceptive" pocket listing practices.

So the academic literature says private listings do not help sellers. The MLS data says private listings do not help sellers. Even the studies Compass might cite say private listings used to help sellers, slightly, until NAR closed the loophole in 2020.

If the product does not benefit the customer, who is it for?

Reffkin already answered that question.

Here is where the story stops being a real estate story and becomes a regulatory story.

In 2024, Invitation Homes, the country's largest single-family rental landlord, paid the Federal Trade Commission $48 million to settle charges that it had hidden mandatory fees from renters until late in the leasing process. In late 2025, Greystar, the country's largest apartment manager, settled similar charges for $24 million. On March 12, 2026, the FTC issued an Advance Notice of Proposed Rulemaking on rental housing fee disclosure, and on December 19, 2025, the DOJ filed a Statement of Interest in Davis v. Hanna Holdings signaling federal enforcement priorities.

The throughline across every one of these cases is the same: information asymmetry between the housing industry and the consumer is now a federal enforcement priority. Hide a $200 amenity fee from a renter and the FTC will fine you eight figures. Strip days-on-market and price-drop history from 500,000 home listings, and... so far, nobody has done a thing.

We are not lawyers, and this newsletter does not give legal advice. But if you are a regulator at the FTC, the DOJ, or any of the 50 state attorneys general offices, and you are reading this, the question writes itself: why is hiding a fee from a renter a federal violation, but hiding the seller's negotiating position from a buyer is "innovation"?

Vermont, Massachusetts, and Washington have already begun examining state-level disclosure rules for off-MLS transactions. None has acted yet. None has acted yet is the key phrase.

In the spirit of fairness, it is worth steelmanning Compass's position.

There are real reasons a seller might want a private listing. Celebrities and high-profile families do not want strangers in their living rooms. Estate sales after a death often benefit from discretion. Some sellers genuinely want to test pricing before committing to a public launch. These are legitimate use cases, and they have existed for as long as real estate has existed.

The fight is not about whether pocket listings should be illegal. The fight is about whether one company should be allowed to systematize them, monetize them at scale, route a third of American home sales through them, scrub the price discovery data that buyers depend on, and then describe the agent, not the homeowner, as the customer.

The free market case against Compass is not that they are too capitalist. It is that they are not capitalist enough. Markets work when buyers and sellers have the same information. Markets break when one side knows everything and the other side knows nothing. That is not innovation. That is the oldest trick in the book, dressed up in a Series F deck.

Robert Reffkin is not a villain. He is a competent CEO running the playbook venture capital handed him in 2012 and refining it ever since. He is winning. He is winning because the system is letting him win.

What we are watching, in real time, is the largest residential brokerage in America building a private market inside the public housing market, and inviting the rest of the industry to either join or get crushed. The DOJ saw it and looked away. Zillow saw it, fought, and partially folded. The MLS researchers saw it and published papers nobody in Washington has read yet. Buyers and sellers are mostly still seeing the polished version on the listing page.

There are three things to watch over the next 12 months.

First, whether any state attorney general decides that information asymmetry in the country's largest financial transaction is worth a press conference and a subpoena. Second, whether the next administration's FTC extends the Greystar logic from rentals to sales. Third, whether the agents themselves, who Reffkin called his customers, decide they would rather work somewhere where the homeowner is the customer instead.

The Compass flywheel works as long as the music keeps playing. The music keeps playing as long as nobody looks too closely at who is dancing with whom.

We just looked.