Only 28% of Commercial Real Estate Deals Tell You What They Cost

You are trying to price a building. You do what everyone does: you read the trade press, you collect the comparable sales, and you build a range. The range feels defensible because it came from real deals that real parties actually closed.

It is defensible. It is also a sample, and not a random one.

Between 27 May and 5 September 2026 our sources published 16,930 items. Of those, 6,235 described an actual transaction: a sale, a financing, a lease, or a ground-up development commitment. Only 1,765 of them, 28.3%, named a dollar figure.

The other 4,470 deals happened. Buildings changed hands, loans closed, tenants signed. The market simply never learned what any of it cost.

Silence is not evenly distributed

If the missing 72% were scattered at random, you could ignore them. They are not.

Deal type	Headlines	Named a price	Rate
Financing	1,249	664	53.2%
Sale	2,853	982	34.4%
Development	1,135	74	6.5%
Lease	998	45	4.5%

A lender is nearly twelve times more likely to tell you the size of a deal than a landlord is.

That is not a quirk of etiquette. It is who owns the announcement. A loan closing has a press release because the lender wants the next borrower to call, and the loan amount is the entire proof of capability. A lease has no such champion. The landlord gains nothing from publishing a rent that his next four negotiations will be measured against, and the tenant gains less than nothing.

A lender is nearly twelve times more likely to tell you the size of a deal than a landlord is.

So the visible market tilts toward debt. If you read the trade press as a proxy for where capital moved, you are reading a record written disproportionately by the people who lend it, not the people who own the asset.

And it is not evenly distributed by sector either

Sector	Transaction headlines	Named a price	Rate
Multifamily	1,158	559	48.3%
Retail	822	338	41.1%
Mixed use	90	33	36.7%
Land	38	12	31.6%
Hospitality	214	59	27.6%
Office	1,008	259	25.7%
Industrial	699	164	23.5%
Capital markets	840	170	20.2%

Multifamily discloses at nearly twice the rate of office.

Part of that is structural. Multifamily is financed substantially by agency debt, and agency originations are brokered by firms whose business model includes announcing them. We have not measured what share of apartment trades passes through those intermediaries, and we are not claiming a number for it. What the corpus shows is the output: apartment deals arrive already carrying a figure far more often than office deals do.

But part of it is selection, and this is the part that should worry anyone building an office comp set. A seller discloses a number when the number flatters him. In a sector where some trades are clearing below the seller's last basis, and this corpus cannot tell you how many, the quiet 74% is unlikely to be a random 74%. The deals that stay off the record are the ones with a reason to.

Which points at an uncomfortable possibility: the office comps you can see may be, on average, the better half of the office deals that happened. We cannot prove that from headlines. Nobody can disprove it from headlines either, which is exactly the problem with valuing anything off them.

What a deal actually looks like when it does speak

Among the 982 sales that named a price, the median was $34.4 million. Half of all disclosed sales fell between $12.7 million and $89.0 million.

That is the market. Not the portfolio recapitalisation that four outlets cover inside two days, and not the trophy tower. A thirty-four million dollar building, brokered locally, financed regionally, and read about by almost nobody.

Coverage volume does not work this way. The largest deals generate the most headlines per deal, because every outlet covers them and several cover them twice. The result is a reading experience weighted toward the top of the market and a reality weighted toward the middle of it. Anyone calibrating expectations from the volume of coverage rather than the count of deals will systematically overestimate what trades.

We do not publish a mean deal size, and the reason is instructive. Any automated read of headline dollar figures will occasionally capture a market total rather than a transaction, and a single $36.7 billion quarterly volume figure caught by mistake moves an average more than four hundred genuine deals do. Medians survive that. Averages do not. Where you see a mean in this industry, ask what caught it.

Disclosure is rising, slightly

Month	Sale headlines	Named a price
June	850	32.5%
July	923	32.2%
August	843	37.7%

August broke five points above the summer's baseline on a comparable count of headlines. One month is not a trend, and we will not call it one. But the direction is the one you would expect if sellers were becoming more comfortable putting numbers on paper, and it is worth watching through the autumn.

What to do with this

Three things follow, and none of them require better data than you already have.

Treat every comp set as a survivorship sample, and say so in writing. If you are valuing office, note in the memo that roughly three-quarters of comparable trades in the period disclosed nothing, and that silence correlates with disappointment. That sentence has more analytical content than another decimal place on your cap rate.

Stop inferring activity from coverage. A sector that generates a lot of headlines is a sector with active press operations. Multifamily discloses at 48.3% and capital markets at 20.2%. That gap tells you about press offices, not about capital.

Read the debt to find the equity. Where owners will not talk, lenders will. At 53.2%, financing announcements are the densest source of real numbers in the market, and a loan amount plus a loan-to-value assumption gets you closer to a price than the silence around the sale ever will.

The market you can see is not the market. It is the part of the market that had a reason to speak.