Cap Rate Benchmarks by US Metro

Cap rate benchmarks are useful for exactly one thing: telling you whether the property in front of you is priced in line with its market. They are close to useless for deciding which market to buy in, and they are actively misleading when compared across asset classes or vintages.

This piece covers what moves the spread between metros, where to get current numbers, and how to read them without reaching the wrong conclusion.

Why metro cap rates differ

Four forces do most of the work.

Expected rent growth

The dominant one. A market where rents are expected to grow 4% annually will trade at a materially lower cap rate than one expected to be flat, because buyers are pricing tomorrow’s income, not just today’s.

This is why high cap rates cluster in slower-growth metros. That spread is not inefficiency waiting to be arbitraged — it is the market’s assessment of the income stream’s trajectory. Buying an 8% cap in a shrinking market and a 5% cap in a growing one can produce identical total returns, or leave the 8% behind.

Liquidity and buyer depth

Markets with deep institutional buyer pools trade tighter. More bidders means more competition for the same income, which compresses cap rates and — importantly — makes the exit more predictable.

Thin markets carry an illiquidity premium. You get paid more current yield, and you accept that when you sell there may be three buyers rather than thirty. That risk is real and it shows up precisely when you least want it, which is when conditions have turned.

Tax and insurance burden

Effective property tax rates vary by a factor of four across US jurisdictions, and insurance costs have diverged sharply in coastal and wildfire-exposed markets.

Both hit NOI directly, which means they hit the cap rate calculation on both sides. High-cost jurisdictions tend to show higher headline cap rates partly because the NOI is already suppressed — and partly because buyers demand more yield to accept expense lines that have been growing faster than rents.

Regulatory environment

Rent stabilization, eviction procedure, and the practical cost of enforcing a lease all affect what an income stream is worth. Markets where a non-paying tenant takes eight months to remove carry a risk that shows up in pricing.

Where to get current numbers

Cap rates move with interest rates, and any benchmark more than about six months old is a historical fact rather than a decision input. These are the sources worth checking:

The last row is the one investors underuse. A broker active in a specific submarket knows what actually traded last quarter, at what cap, and on whose numbers. Published surveys report averages across a metro that may contain submarkets 200 basis points apart.

Benchmark table

Multifamily, all classes blended — this data set does not break out by A/B/C, which is itself worth noting: a metro average across classes tells you less than a class-specific figure would, and you should weight it accordingly. Based on 929 transactions totaling $41B across the 30 most active US multifamily markets, transactions through September 2025. Source: Yardi Matrix, via Bisnow, published October 14, 2025.

This snapshot is roughly a year old as of this writing. By the six-month rule stated above, treat the levels below as directional, not current — the spread between markets moves more slowly than the absolute levels do, so the relative ordering (gateway metros tight, Sun Belt wider) is more durable than any single number.

*Spread computed against the 10-year Treasury yield at 4.65% (August 27, 2026). Because the cap rate snapshot predates the Treasury figure by roughly eleven months, treat these spreads as illustrative of the method, not a precise current reading — recompute against a live Treasury yield and a fresher cap rate print before using this for an actual decision.

Always state the survey date and the asset class. A cap rate without both is not a benchmark — it is a number. Class A and Class C in the same metro routinely sit 150–250 basis points apart, and a benchmark that does not specify which one it describes cannot be used for comparison. This table’s biggest limitation is exactly that: it is all-class blended, because that is what was available without a paywall — see the note above the table.

How to actually use a benchmark

Compare within a market, not across markets. The right question is “is this property priced in line with comparable properties in this submarket?” A 6.2% cap on a building where similar assets trade at 5.6% is worth investigating. Whether 6.2% is better than a 7.4% cap in a different state is a question benchmarks cannot answer, because the two numbers describe different risks.

Match the asset. Vintage, class, unit count, and condition all move cap rates within a single submarket. Benchmarking a 1962 twelve-unit against a survey of institutional Class A product tells you nothing useful.

Check the spread, not the level. Cap rates move with the risk-free rate. A 5.5% cap rate means something very different at a 1.5% 10-year Treasury than at a 4.5% one. Track the spread over the 10-year and you have a measure that survives rate cycles; track the absolute level and you will conclude that every market got worse in 2022 and better in 2020, which is a statement about interest rates rather than about real estate.

Treat divergence as a question, not an answer. A property priced well above the local benchmark is not automatically a bargain. Find out why. Deferred capital, a below-market lease with years left, a pending assessment appeal, or a submarket the survey averages over can all explain it. Occasionally the answer is that the seller needs to sell — but you should confirm that rather than assume it.

The limit of benchmarks

A benchmark tells you what other people paid for other buildings. It does not tell you whether the income you are buying is real.

That comes from building the NOI yourself — reassessed taxes, quoted insurance, a management fee whether or not you intend to pay one, and a reserve for capital you know is coming. A property that benchmarks well against its market on the seller’s NOI and poorly on yours is a property priced on numbers that will not survive your first year of ownership.

The full sequence for getting to a defensible number is here, and the tools that make it fast enough to run on every property rather than your favorites are compared here.

FAQ

What is the average cap rate in the US right now? A national average blends asset classes, vintages, and markets that have little to do with each other, which makes it nearly useless for any individual decision. Use metro and class-specific data from a dated survey instead.

Why are cap rates lower in expensive coastal markets? Primarily expected rent growth and buyer depth. Investors accept less current yield where they expect income to grow and where the exit is liquid.

Should I buy in a high cap rate market? High cap rates compensate for something — slower growth, thinner liquidity, higher expense burden, or greater regulatory friction. Sometimes that compensation is generous. The point is to identify what you are being paid to accept before deciding it is worth it.

How often do cap rates change? They track the interest rate environment, so they move continuously and can shift 50–100 basis points within a year when rates move sharply. Any benchmark older than about six months should be treated as historical.