Nobody buys a building. They buy a stream of income, and then everyone in the room argues about the price. The cap rate is where that argument settles.
The math is simple, and it has to be, because everything in this business eventually hangs on it. Cap rate equals net operating income divided by price. A building nets $200,000 a year and sells for $4 million, that's a 5 cap. Same building, same income, same tenants — the cap rate is nothing more than the price translated into a percentage. It tells you what return the buyer accepted for owning that income stream.
Here is why an operator should care more about cap rates than an appraiser does: the formula runs backward. Value equals NOI divided by the cap rate, which means at a 5 cap, every recurring dollar of NOI is worth twenty dollars of building. Raise rents $50 a month across 40 units and collect it — that's $24,000 a year, which is roughly $480,000 of value at a 5 cap. This is the whole reason small operational wins matter so much. Nobody gets rich saving $200 on a landscaping contract; they get rich because that $200 repeats every month and then gets multiplied by twenty.
Cap rates move, and right now they're telling a clear story. CBRE's H1 2026 cap rate survey found stabilized multifamily in New York City trading at 5% to 5.5%, up from 4.5% to 5% in the back half of 2025. Nationally, cap rates held flat through the first half of the year even as Treasury yields whipsawed, with about 60% of surveyed investors expecting them to hold steady. When cap rates rise on the same income, values fall. That widening in New York is the market repricing risk, and it happened without a single tenant leaving.
What moves a cap rate is perceived risk and expected growth. A well-located, stable building trades at a lower cap because buyers accept a thinner yield for safety. A building with hair — vacancy, deferred maintenance, a tough submarket — trades at a higher cap because the buyer needs the extra yield to stomach the risk. Strong expected rent growth compresses cap rates for the same reason: buyers will pay more today for income they believe will grow tomorrow.
The traps are all in the inputs, never the division. The most common one is taking the seller's pro forma NOI at face value — projected rents, fantasy vacancy, expenses that somehow exclude management — and dividing it into the asking price to get a cap rate that looks great and means nothing. Run the cap rate on collected, verified income. The second trap is comparing caps across markets like they're interchangeable. A 5 cap in Manhattan and a 7 cap in a secondary market are not two prices for the same thing; they're two different risk profiles wearing the same formula.
And remember the mirror image: the exit cap in your own model should almost never equal your going-in cap. Markets move, buildings age, and the buyer who takes your building off your hands in year five will underwrite it with fresh eyes. Assuming you exit at the same cap you bought at is just assuming the market's opinion of your building never changes. It always does.
You don't negotiate the cap rate. You discover it, deal by deal, in whatever the market is paying that quarter. What you control is the NOI underneath it — the rents you collect, the expenses you cut, the vacancies you fill. The market sets the multiple. Your job is to move the one number you can actually touch.
← All insights