I spent 10 years on the buy side at Stockbridge Capital, the last five as Director of Research. We closed more than $2B in commercial real estate during that run. The cap rate conversation came up on every multifamily deal — usually as the moment where the broker’s number and our underwritten number diverged by 50 to 100 basis points.
That gap is the entire story of this article. The cap rate printed in an OM is a marketing number. The cap rate institutional buyers actually pay is a derived number that starts with comp data and then layers on adjustments for condition, location, tenant profile, CapEx, and the cost of debt that day. The teams that consistently win deals — and consistently get out at exit — are the ones who treat the broker’s cap rate as a starting point and do the adjustment work themselves.
What the cap rate actually tells you
A cap rate is the simplest valuation math in commercial real estate and the most misused number in the industry. Divide net operating income by purchase price. The result is the unlevered going-in yield a buyer is accepting on day one. A 5.5% cap rate on a $50M deal says the buyer expects $2.75M of NOI in year one before debt service, taxes, and CapEx.
That is the math. The meaning is harder.
A cap rate compresses an enormous amount of information into a single number: the market’s view of risk, the cost of debt, rent growth expectations, supply pipeline, insurance and tax trajectories, and the asset’s physical condition. When a broker says a deal is trading at a 5.25% cap, they are telling you what the market paid — not what the buyer underwrote or whether the underlying NOI is defensible.
This guide covers 2026 benchmark ranges by class and geography, the forces moving cap rates this cycle, how institutional buyers actually derive their number, and the traps that cost acquisitions teams tens of millions per year.
2026 benchmark ranges by asset class
The ranges below are illustrative of stabilized multifamily cap rates in primary and secondary U.S. markets as of Q2 2026. They are not a substitute for market-specific comp data, but they frame what a reasonable institutional buyer is paying.
| Asset class | Typical cap rate range | Representative deal profile |
|---|---|---|
| Class A urban core | 4.5% – 5.5% | New-build high-rise, gateway market, institutional sponsor |
| Class A suburban | 5.0% – 6.0% | Post-2015 garden-style, top suburban submarket, strong schools |
| Class B / B+ | 5.5% – 6.5% | 1990s-2010s vintage, value-add or light renovation, strong tenant base |
| Class C | 6.5% – 8.0% | Pre-1990 vintage, workforce housing, heavier CapEx and management intensity |
Three points about these ranges.
First, they are wider than they were in 2021. At the cycle peak, Class A urban traded as tight as 3.25% in Manhattan and 3.75% in Boston. Those compressed cap rates priced in perpetual rent growth, near-zero short rates, and abundant agency debt. None of those conditions hold in 2026.
Second, the spread between classes has widened. In 2021, the gap between Class A and Class C was roughly 150 basis points. In 2026, that spread is closer to 200 to 300 basis points, reflecting repriced CapEx risk and the cost of managing older assets through an insurance-premium and property-tax reset cycle.
Third, these are stabilized cap rates. Lease-up deals, heavy value-add plays, and non-stabilized assets trade on stabilized yield-on-cost or unlevered IRR, not going-in cap rate. Using a stabilized cap rate benchmark on a non-stabilized asset is one of the fastest ways to overpay.
How geography moves the number
National cap rate ranges are a starting point. The spread within those ranges is driven almost entirely by geography.
Coastal vs. Sunbelt
For a decade running into 2022, Sunbelt multifamily traded 50 to 100 basis points tighter than comparable coastal assets. Population in-migration, favorable state tax regimes, lower construction cost inflation, and institutional consensus about “where the growth is going” compressed Sunbelt cap rates even as rents grew faster.
That trade has partly unwound. Phoenix, Nashville, Austin, Tampa, and Charlotte absorbed heavy supply pipelines that peaked in 2023-2024 deliveries. Rent growth turned negative in some submarkets. Insurance premiums doubled in Florida and parts of Texas. Property taxes reset after sales. Cap rates pushed out 75 to 150 basis points from the 2021 low.
Meanwhile, coastal gateway markets with constrained supply — Boston, Seattle, D.C. suburbs, Northern New Jersey — held firmer. Class B garden in these markets often trades inside comparable Sunbelt product in 2026.
Primary, secondary, tertiary
A simpler framing than coastal-vs-Sunbelt is primary / secondary / tertiary, which correlates more reliably with cap rate than geography alone.
- Primary markets (top 10 by institutional transaction volume): NYC, L.A., Boston, D.C., Seattle, Chicago, San Francisco, Miami, Atlanta, Dallas. Cap rates at the tight end of class ranges.
- Secondary markets (Charlotte, Nashville, Denver, Phoenix, Tampa, Austin, Raleigh): roughly 25 to 75 basis points wider than primary for comparable product.
- Tertiary markets (smaller MSAs and suburban submarkets of secondary metros): 75 to 150 basis points wider, sometimes more. Liquidity risk is the driver — fewer institutional bidders means the seller pays a liquidity premium in the form of a higher cap rate.
A Class A asset in Huntsville or Boise is not the same risk as a Class A asset in Atlanta, even if the rent roll looks similar. Institutional buyers apply a tertiary-market discount that shows up in the cap rate.
What is driving 2026 cap rates
Three macro forces explain most of the cap rate movement since 2022.
1. The 10-year Treasury and SOFR. The risk-free rate is the floor under every real estate valuation. When the 10-year moved from 1.5% in late 2021 to the 4.0% to 4.5% range it has held through 2024-2026, cap rates had to expand to maintain a risk premium. The spread between multifamily cap rates and the 10-year (the “risk premium”) has historically averaged around 150 basis points. It compressed below 100 basis points at the 2021 peak, which was the market’s warning that prices had run ahead of fundamentals.
2. Insurance cost inflation. Property insurance premiums roughly doubled between 2020 and 2025 in Florida, coastal Texas, Louisiana, and parts of California. For a 200-unit garden-style asset, that can mean $400K to $800K of NOI erosion per year. Insurance is now the fastest-growing operating expense line in institutional multifamily underwriting. Buyers are pricing in continued premium escalation, which compresses NOI projections and expands cap rates.
3. Rent growth deceleration. The 2021-2022 rent spike is over. National multifamily rent growth normalized to low-single-digits through 2024-2026, with negative growth in supply-heavy Sunbelt submarkets. The “trees grow to the sky” underwriting that justified tight 2021 cap rates has been replaced by conservative rent growth assumptions — which mechanically expands the cap rate required to hit target returns.
A fourth, quieter driver: property tax resets after sale. In Texas, Florida, and several other states, a trade triggers a full reassessment. Year-two property tax can jump 30% to 60%, compressing NOI by 5% to 10% of the going-in number. Sophisticated buyers underwrite to post-reset taxes, not seller’s historical. Less sophisticated buyers use the seller’s T-12 and discover the problem in year two.
Going-in, stabilized, and exit cap rates
The single most common underwriting confusion is mixing three distinct cap rates.
Going-in cap rate uses year-one NOI — the actual NOI the property produces on day one. This is what the property is worth today. Any cap rate quoted in a broker OM without a qualifier is usually going-in.
Stabilized cap rate uses projected NOI after the business plan executes — lease-up completed, renovations done, concessions burned off, taxes and insurance normalized. This is what the property will be worth after you execute. For value-add deals, stabilized cap rates run 50 to 100 basis points tighter than going-in because the risk is lower after stabilization.
Exit cap rate is the cap rate you assume the next buyer will pay when you sell in year 5, 7, or 10. It should be wider than your going-in cap — typically 25 to 75 basis points — to reflect the asset aging, uncertain future rate environment, and the discipline of not underwriting perpetual compression.
The mistake that costs teams deals: applying a stabilized cap rate to a non-stabilized NOI. If the asset has 8% vacancy, $2M of unaddressed CapEx, and concessions still burning off, dividing the current NOI by a stabilized cap rate produces a value that is neither today’s nor tomorrow’s. It produces a number that justifies the price the broker wants.
How institutional buyers actually derive a cap rate
The cap rate printed in an OM is a market observation. The cap rate an institutional buyer uses in their IC model is a derived number that starts with the market observation and then layers on adjustments.
The derivation usually looks like this:
-
Start with the market cap rate. Pull recent comparable trades from Newmark, JLL, CBRE, RCA/MSCI, and CoStar. Filter for vintage, asset class, submarket, and business plan. The output is a range — typically 50 to 75 basis points wide.
-
Adjust for condition. Roof age, plumbing material (Kitec, polybutylene), HVAC vintage, parking lot condition, structural issues. A comparable asset with $5K per unit of deferred maintenance trades at a wider cap rate than one with zero deferred. The adjustment can be 25 to 100 basis points.
-
Adjust for location within the submarket. Flood zone, wildfire exposure, school district, crime trend, adjacent land use, walkability. A deal two blocks off the main corridor trades wider than the corridor comp.
-
Adjust for tenant profile. Rent-to-income ratios, concession burn-off schedule, renewal rates, eviction trends, income mix. A property with 40% of tenants under 2.5x rent-to-income has more default risk than one at 3.5x, even at the same in-place rents.
-
Adjust for CapEx. How much unreserved CapEx will this asset require in the hold period? Unreserved CapEx compresses IRR and should push the cap rate wider. A buyer who does not underwrite a realistic CapEx reserve is overpaying.
-
Adjust for financing. At 6.5% agency debt, the LTV the deal can support drops. The equity return compresses. Cap rates expand until the levered return target is met.
Two institutional buyers can look at the same OM and derive cap rates 50 to 75 basis points apart because their adjustments differ. The buyer who wins the deal is usually the one with the lowest derived cap rate — which is either the sharpest underwriting or the most optimistic.
Common cap rate mistakes
The cap rate errors that show up in underwriting reviews, in roughly descending order of frequency and cost:
1. Using the seller’s T-12 without normalizing. The seller’s NOI often includes one-time non-recurring income, below-market insurance that will reset at renewal, payroll that does not reflect the required management structure, and tax expense that will jump on reassessment. Underwriting to that NOI produces a going-in cap rate that looks better than reality.
2. Applying a stabilized cap rate to unstabilized NOI. Covered above. If the asset is not stabilized, the stabilized cap rate is the wrong lens.
3. Ignoring insurance and tax resets. For Texas, Florida, and other reassessment states, post-close property tax can exceed seller’s by 30% to 60%. Insurance renewals in coastal markets frequently come in 30% to 100% above seller’s in-place premium. Year-two NOI looks nothing like year-one if you ignore these.
4. Using seller’s forward NOI. Brokers often present a “forward 12” that assumes rent growth, concession burn-off, and operational improvements between listing and close. That NOI has not been achieved. A sophisticated buyer underwrites to trailing NOI and projects forward under their own assumptions.
5. Using the broker’s cap rate. The cap rate in an OM is calibrated to the price the seller wants. It is a marketing number, not an underwriting number. Derive your own.
6. Not adjusting exit cap for vintage drift. A 2015-vintage asset held seven years is a 2022-vintage asset at exit. The exit cap should reflect the asset’s age at sale, not at purchase.
Data sources professionals actually use
Institutional teams triangulate across four primary sources and validate against recent trades.
- Newmark Multifamily Capital Markets Quarterly. Transaction volume, cap rate ranges by market, pricing trends. Strongest on mid-market and secondary deals.
- JLL Multi-Housing Outlook. Quarterly cap rate and market overview. Strong on primary markets and Sunbelt coverage.
- CBRE North America Cap Rate Survey. Published semi-annually. Cap rate survey data by asset class and market from CBRE’s capital markets professionals.
- Real Capital Analytics / MSCI. Transaction-level data. The gold standard for closed-deal pricing, but expensive.
For deal-specific comps, teams layer in CoStar, RealPage, RealtyRates, and broker OMs for adjacent trades. No single source is authoritative. The discipline is comparing all four and noting where they diverge.
How DDee.ai fits into cap rate analysis
A cap rate is only as good as the NOI underneath it. The biggest source of error in institutional underwriting is not the cap rate selection — it is the NOI the cap rate is applied to.
That NOI is built from primary documents: rent roll, T-12, tax bills, insurance quotes, leases, operating statements. In most shops, an analyst keys the data into a spreadsheet over two or three days, then normalizes it for non-recurring items, pending tax resets, concessions, and insurance renewal risk.
DDee.ai eliminates the keying. It reads the PDFs, extracts the line items, flags inconsistencies, and produces a normalized NOI with each figure linked back to its source document. The analyst receives clean inputs for the cap rate analysis in under an hour rather than three days. Clean NOI does not replace judgment on the cap rate itself — it makes sure the number you are capitalizing is the right number.
See how DDee.ai builds the NOI that sits underneath your cap rate analysis →
The bottom line
The cap rate is one number that compresses a dozen forces into a single decimal. In 2026, those forces have repriced substantially: the 10-year Treasury, insurance inflation, rent deceleration, property tax resets, and the end of the zero-rate era. Cap rates have widened across every class and geography, spread between classes has grown, and the Sunbelt-coastal premium has partly unwound.
Reading a cap rate correctly means three things. First, understanding whether it is going-in, stabilized, or exit. Second, deriving your own number through explicit adjustments for condition, location, tenant profile, and CapEx — not accepting the broker’s. Third, making sure the NOI under the cap rate reflects reality, including tax resets, insurance renewals, and normalized operating expenses.
The cap rate is not hard math. It is hard judgment. The teams that get it right are the ones that invest in clean data and disciplined adjustment, not the ones that trust the broker OM.