Cap Rates Explained: What They Are and What Drives Them

Introduction

A 50 basis point cap rate shift moves property value by roughly 10% on a stabilized asset, all else equal. That sensitivity is why the cap rate is the most quoted number in commercial real estate and why every RE banker tracks the cap rate environment more closely than the interest-rate environment that drives it. The cap rate compresses or expands based on four interacting drivers (interest rates, property fundamentals, the asset's risk premium, and transaction friction), and understanding which driver is moving in a given moment is the difference between a candidate who can talk credibly about deal activity and one who cannot.

The state of play in 2024 made the mechanic visible. Cap rates rose 200+ basis points across most property types from the 2021 lows, with secondary office assets repricing the hardest. Industrial cap rates moved least because the e-commerce-driven NOI growth offset the rate pressure. With the Fed beginning rate cuts late in 2024, CBRE projected cap rate compression of roughly 30 bps in industrial, 24 bps in retail, 17 bps in multifamily, and only 7 bps in office through end-2025. The variance across property types tells you exactly which driver is dominant in each sub-sector.

The cap rate is the unlevered yield on a stabilized real estate asset:

The formula is intentionally simple. A property generating $10 million of stabilized NOI valued at a 6.0% cap rate trades at approximately $167 million (10 divided by 0.06). The cap rate is before debt service, before income taxes, and before capex because those items vary by owner; the cap rate isolates the property's underlying economic yield in a way that is comparable across buyers and capital structures.

Cap Rate (Capitalization Rate)

The unlevered yield on a stabilized real estate property at acquisition, calculated as Net Operating Income divided by property value. Mathematically the inverse of an NOI multiple (5% cap rate = 20x NOI; 7% cap rate = 14.3x NOI). Cap rates are quoted as percentages and serve as the primary market-comparable pricing benchmark across the commercial real estate industry.

The cap rate is the inverse of an NOI multiple. A 5% cap rate corresponds to a 20x NOI multiple; a 7% cap rate corresponds to a 14.3x NOI multiple. The arithmetic relationship matters when bridging real-estate language to corporate-finance valuation: a corporate analyst quoting "10x EBITDA" maps roughly to a real estate analyst quoting a "10% yield" if you treat property NOI as a rough proxy for property EBITDA, though the cleanest cross-walk requires separately adjusting for management overhead and recurring capex.

Why the 50bps-Equals-10% Heuristic Works

The arithmetic behind the "50 bps moves value by 10%" rule of thumb is straightforward. If a property's NOI is fixed and the cap rate moves from 6.00% to 6.50%, the implied value moves from NOI/0.06 to NOI/0.065, a decrease of approximately 7.7%. From 5.00% to 5.50%, the decrease is approximately 9.1%. From 4.00% to 4.50%, the decrease is approximately 11.1%. The rule of thumb averages to "roughly 10%" across the cap rate range RE bankers actually see; the precise math compresses at higher cap rates and expands at lower cap rates.

The sensitivity is what makes cap rate movement so consequential for REIT prices, sponsor underwriting, and transaction activity. A 100 bps cap rate widening cycle (typical of the 2022-2024 rate-up period) takes roughly 20% off property values industry-wide, before any NOI growth offsets. That magnitude is exactly why the listed REIT market re-priced sharply in 2022 and why 2024 take-private activity centered on assets where private-market cap rates had not adjusted as much as listed prices.

Cap rates do not move in response to a single input. Four interacting drivers explain almost every cap rate move:

Driver 1: Cost of Debt

The most cited cap rate driver is the 10-year Treasury yield (or for longer-duration core deals, the 10- to 15-year unlevered cost of capital). The same Treasury yield curve that debt capital markets desks price against is the anchor that resets real estate cap rates as it shifts. The relationship is real but not 1-for-1. The historical spread between average commercial real estate cap rates and the 10-year Treasury since 1990 has run roughly +/- 130 basis points around a long-term mean of about 200-300 bps depending on property type, narrowing to approximately +/- 100 basis points since 2000 as transparency and capital flows increased.

That spread is the cleanest single gauge of how much risk premium the market is pricing into real estate over the risk-free rate, and bankers quote it constantly when arguing whether cap rates are cheap or rich versus history:

A 6.0% cap rate against a 4.5% 10-year Treasury implies a 150 bps spread, near the tight end of the historical band. When the spread compresses toward zero, cap rates are not pricing meaningful risk premium over Treasuries and the market is stretched; when it blows out, the market is demanding a fat cushion. The 2022-2024 widening was driven mostly by the Treasury leg moving, which is exactly why cap rates lagged: a rising risk-free rate compresses the spread before sellers accept the wider cap rate that restores it.

Driver 2: Property-Type Fundamentals

NOI growth expectations directly compress cap rates. Industrial cap rates compressed roughly 200 basis points from 2018-2022 as e-commerce-driven logistics demand reset the sector's long-term NOI growth path. Data center cap rates compressed even harder as AI-driven hyperscaler demand pushed the sub-sector's pricing into the 3-4% range, well below historical norms for any other commercial property type. Office cap rates moved the opposite way during the work-from-home transition because tenant retention assumptions deteriorated and NOI growth expectations turned negative for secondary office buildings.

Driver 3: Asset Risk Premium

Within the same sub-sector, individual assets price at meaningfully different cap rates based on tenant credit, lease structure, lease-roll exposure, submarket positioning, and physical condition. A Class A trophy office tower in midtown Manhattan with investment-grade tenants on 15-year leases trades at a meaningfully tighter cap rate than a Class B office building in a tertiary market with short-dated leases to non-investment-grade tenants. The risk-premium spread between the two assets is the cap rate translation of equity risk premia in corporate finance.

Driver 4: Transaction Friction

Short-term cap rate moves often reflect financing-market friction more than fundamental shifts. When CMBS conduits widen, agency multifamily lenders raise spreads, or banks reduce CRE balance-sheet capacity, the implied cost of debt for a new acquisition rises and cap rates widen mechanically as buyers require larger going-in yields to clear levered IRR hurdles. The 2008-2009 cycle, the 2020 COVID quarter, and the 2022-2023 rate-up period all featured financing-market friction layered on top of the underlying rate move. Modeling cap rates without accounting for financing-market conditions misses a meaningful share of short-term variation.

How to Estimate a Cap Rate When Comps Are Sparse

A common interview scenario: estimate the cap rate on a hypothetical property without recent comparable sales. The standard approach builds the cap rate from first principles using the build-up method, the same logic that constructs a discount rate from a risk-free base plus risk premia: start with the 10-year Treasury yield as the risk-free rate, add the historical commercial real estate risk premium over Treasuries (roughly 200-300 basis points for stabilized core assets), add a property-type premium (industrial below, office above), add an asset-specific risk premium for tenant credit, lease-roll exposure, submarket positioning, and physical condition, then subtract an expected NOI growth offset (for assets in growing sub-sectors like industrial or data centers). The result is an estimated cap rate that should map reasonably closely to where the market would actually clear if a transaction occurred.

The build-up method is mechanical but useful for cross-checking comp-based cap rate quotes that may have idiosyncratic features (a forced sale, a programmatic sale to a related party, a deal where the seller carried back financing). When a quoted cap rate looks dramatically different from the build-up estimate, the analyst should investigate why before accepting the comp at face value.

The same decomposition has a cleaner closed form. Treat a stabilized property as a perpetuity of NOI growing at a constant rate, and the Gordon Growth Model gives its value as the forward NOI divided by the discount rate minus the growth rate:

This is the growth-based view of the same cap rate the build-up method constructs. The discount rate r is the property's required return (the risk-free rate plus the stack of risk premia above), and g is the perpetual NOI growth rate (the same growth offset the build-up subtracts). A property with an 8.5% required return and 2.5% perpetual NOI growth prices to a 6.0% cap rate, which is why the strongest-growth sub-sectors (industrial, data centers) carry the lowest cap rates: a higher g pulls the cap rate down for any given required return. The build-up method and the Gordon Growth identity are two routes to the same number, one adding premia up from the risk-free rate and the other subtracting growth from the discount rate.

Acquisition underwriting uses two cap rates: the going-in cap rate (the cap rate at acquisition, applied to year-1 stabilized NOI) and the exit cap rate (the cap rate at sale, applied to the forward NOI at exit). The relationship between them is one of the most argued-over assumptions in any deal model.

Going-In vs Exit Cap Rate

Going-in cap rate is the cap rate applied to year-1 stabilized NOI at acquisition. Exit cap rate is the cap rate assumed when the property is sold, applied to the forward NOI at exit. The standard convention is that exit cap rates should be 25 to 50 basis points wider than going-in cap rates to account for asset depreciation over the hold period (older building, shorter lease durations, more recurring capex). Aggressive underwriting sometimes assumes flat or compressing exit caps, which makes the levered IRR look better than it should.

The mechanical rationale for the spread is that an older building generally trades at a higher cap rate than a newer building of the same quality, because the cap rate prices in expected capex, lease-roll risk, and submarket repositioning needs. A property bought at year 0 and sold at year 7 is meaningfully older, and the exit buyer will demand additional yield to compensate. The 25-50 bps spread is the rule of thumb across most stabilized core and core-plus deals.

Aggressive underwriting (typically by sponsors trying to win a competitive process) sometimes assumes flat or compressing exit caps: the exit cap rate equals or is lower than the going-in cap rate. The justification is usually that the property will be in better condition after capex investment, in a stronger submarket trajectory, or in a tighter rate environment at exit. The lender response is to apply higher exit caps anyway as a stress case, because optimistic exit cap assumptions are the single biggest source of overstated levered IRRs in deal underwriting.

The cap rates above describe the private market for individual property transactions. The mirror-image concept for listed REITs is the implied cap rate, calculated by working backward from the REIT's enterprise value to ask what cap rate the public market is applying to the portfolio's NOI. The math: Implied Cap Rate equals trailing or forward NOI divided by Enterprise Value (market equity capitalization plus preferred stock and debt, minus cash and joint venture interests).

The implied cap rate is the single most-watched number in every RE PE sponsor's daily morning screen. When a listed REIT trades at an implied cap rate 100 basis points or more above the private-market cap rate for comparable assets, the gap signals a take-private opportunity at attractive economics. The sponsor can acquire the listed equity at the implied valuation (cheap relative to underlying assets), pay a 20-30% takeout premium to the shareholders, and still end up owning the portfolio at a meaningful discount to NAV. The 2024 Blackstone take-private of AIR Communities at $39.12 per share rested on exactly this mechanic: AIR's implied cap rate sat well above the private-market multifamily cap rate, and the take-private captured the spread.

When Implied Cap Rates Sit Below Private Cap Rates

The reverse situation drives the opposite recommendation. When listed REITs trade at implied cap rates below private-market cap rates (a premium to NAV), the listed entity should issue equity and acquire properties at the cheaper private-market pricing, recycling the public-market arbitrage into asset growth. The 2020-2021 industrial REIT premium-to-NAV environment drove a sustained wave of REIT-led acquisitions financed by accretive equity issuance, exactly because the implied cap rates were below private cap rates for that subsector.

Implied Cap Rate

The cap rate the public market applies to a listed REIT's portfolio, calculated by dividing trailing or forward NOI by enterprise value (market equity cap + preferred + debt - cash). Higher implied cap rates than the private market signal a REIT discount to NAV; lower implied cap rates signal a premium. Daily monitoring of implied cap rates across the listed universe drives almost every take-private and follow-on equity decision.

A cap rate quoted off in-place NOI can be misleading when the property's in-place rents differ materially from market rents. A Class A office building leased at $50 per square foot when the submarket is renting at $70 per square foot will see NOI step up materially as leases roll to market, and the going-in cap rate quoted off in-place NOI will understate the underwritten yield. The reverse situation (in-place rents above market) is equally common in distressed office or post-COVID retail.

Sophisticated underwriting uses a lease-roll-adjusted cap rate: project the NOI trajectory as leases roll to market, and quote the cap rate off a stabilized future-NOI figure rather than the in-place number. The convention varies by sponsor and broker, but the underlying point is the same: the cap rate is only meaningful relative to the NOI baseline it is applied to, and a property's value depends on what its NOI will actually be three to five years out, not the contractual in-place figure today.

Mark-to-Market Gaps Especially Matter in Office

The mark-to-market gap matters especially in office. A trophy Manhattan office building leased at $80 per square foot in a market renting at $95 per square foot has roughly a 15-20% mark-to-market upside as leases roll. That upside is real and worth paying for, but the going-in cap rate quoted off in-place NOI will look tighter (lower yield) than the lease-roll-adjusted cap rate that captures the upside. Buyers and sellers spend considerable negotiation time on which NOI figure the cap rate quote anchors to.

Beyond stabilized properties, value-add and development underwriting uses the trended cap rate: the cap rate applied to a stabilized future-NOI figure that assumes the business plan executes. A value-add multifamily acquisition might quote a going-in cap rate of 4.5% on in-place NOI alongside a trended cap rate of 6.5% on year-3 stabilized NOI after renovation premium, occupancy ramp, and operating expense normalization. The trended number is the more meaningful underwriting input because it reflects the actual investment return the sponsor is targeting.

Development underwriting takes the concept further. Ground-up construction projects quote yields on cost (not cap rates on value) at completion, computed as the stabilized NOI at year 3 or 4 divided by total project cost. A typical industrial development might target a 6.5-7.5% yield on cost when stabilized core industrial trades at a 5.0-5.5% cap rate, the spread between the two figures representing the development profit margin the developer earns for taking the construction and lease-up risk. When the spread compresses below 75-100 basis points, new development becomes economically unattractive and pipelines slow; when the spread widens, developers accelerate.

Cap rate ranges vary widely across property types. The current snapshot for stabilized institutional-quality assets:

The ranges shift with the cycle, but the relative ordering is structural. Hyperscale data center cap rates almost always sit below industrial, which almost always sits below multifamily Class A, which almost always sits below trophy office, and so on through the risk spectrum. Holding the ordering in mind matters because every cap rate quote should be sanity-checked against the sub-sector band: a 5.0% cap rate on a Class B suburban office is a tell that something else is going on (a forced sale, a strategic buyer, a hidden lease-roll story), because the number sits far outside the structural range for that asset profile.

Why Forward Cap Rate Forecasting Is Harder Than It Looks

Forecasting where cap rates will be in 12 to 24 months is the implicit job of every RE PE underwriter and every REIT corporate-development team. The standard forecasting frameworks combine a rate-cycle view (Fed policy path, term structure, inflation expectations), a fundamental-cycle view (sub-sector supply, demand, NOI growth trajectory), a capital-flows view (dry powder deployment pressure, REIT premium-to-NAV reactivation), and a liquidity view (CMBS issuance pace, agency lender appetite, bank balance-sheet capacity). The output is a base-case cap rate forecast by sub-sector with explicit upside and downside scenarios.

Why Sub-Sector Variance Matters in Forecasts

CBRE's late-2024 cap rate compression forecasts (roughly 30 bps for industrial, 24 bps for retail, 17 bps for multifamily, and 7 bps for office through end-2025) embedded specific assumptions on each driver. The forecast was deliberately uneven across property types because the sub-sector fundamentals (industrial demand strength, office sublease overhang) were diverging meaningfully. Generic "cap rates will compress by X bps" forecasts that do not distinguish across sub-sectors mask the underlying mechanics that bankers actually use to underwrite real deals. A senior banker reading a research note that quotes a single industry-wide cap rate forecast typically dismisses it as not actionable; the sub-sector variance is where the deal-relevant signal lives.

The forecast horizon also matters for the methodology:

Each timeframe uses different inputs, which is why "where are cap rates going" is not a single question. The 6-month and the 5-year answers will draw on different inputs and can point in opposite directions without being inconsistent. Forecast research notes that quote a single number without naming the horizon are usually doing a 12-month rate-driven view dressed up as a structural forecast, which is why senior bankers tend to discount them in pitch conversations.

Reading a cap rate quote well means sidestepping two easy misreads. The first is treating cap rates as moving 1-for-1 with the 10-year Treasury. They do not. The 2022-2023 cycle saw 300 bps of rate movement translate into 50-100 bps of cap rate widening because real estate transactions are voluntary and buyer-seller views took time to converge. A candidate who anchors to "rates move, cap rates move the same amount" is using a model that does not match the data. The second misread is quoting a going-in cap rate off in-place NOI without checking the mark-to-market gap. A property leased at $50 per square foot in a $70 market has a meaningfully different underwritten yield than the in-place cap rate quote suggests, and the lease-roll-adjusted figure is the number that actually drives buyer behavior. Both misreads come from treating the cap rate as a static market price rather than a yield that prices in rate expectations, NOI trajectory, and lease-structure mechanics together.