Comparing Cap Rates Across Sales and Surveys

Cap rates compare cleanly only when each one divides the same kind of annual NOI by the same kind of price. Closed-sale rates, appraisal averages and survey responses record different things, so a gap between them is part measurement and part market. Rebuilding every comparable on the subject’s income basis removes the measurement part, and what remains is a question about the properties themselves.

Which twelve months of NOI sit behind each rate

Every cap rate is one year of NOI divided by a price or value, so two published rates match only when the year and the deductions match. Multifamily underwriting checklists make the same point: the cap rate applied should match the income basis used for the subject, whether that is actual or trailing twelve-month income or pro forma income.

Three income bases circulate under one label. Historical income describes a period that has already been operated. Projected first-year income describes the year after closing as the buyer expects it to run. Stabilized income describes a condition the property has not yet reached. Each produces a correct ratio, and each answers a different question.

The entry cap rate is typically defined as projected first-year NOI divided by the acquisition price. An entry rate based on projected NOI is forward-looking, and an investor who sets it beside a rate built on trailing income is comparing a forecast with a result.

At a fixed price, more NOI means a higher cap rate. A projection that assumes rent growth, or a setup that carries unusually light expenses, lifts the ratio with no change in what the buyer paid. Reading that higher rate as cheaper pricing mixes an income assumption into a pricing conclusion.

Two questions are worth asking of any comparable: whether its NOI came from a setup with very light expenses, and whether projections were used. A comparable that fails either test can still be used, but only after its NOI is rebuilt on the same deductions and the same period as the subject.

The decision test is simple to state. If the comparable’s income definition could be applied to the subject without changing the operating period or the deductions, the two rates can sit side by side. If not, the adjustment comes first and the comparison second.

“Ask questions such as: Was it from a broker’s setup that used very light expenses? Were projections used?”

How acquisition and disposition formulas differ by property type

Public acquisition and disposition cap rates are built from issuer-specific formulas, and the formulas do not line up across property types.

One apartment owner defines its acquisition cap rate as anticipated next-twelve-month NOI, adjusted for estimated management costs and in-unit replacement capital, divided by gross purchase price. For lease-up acquisitions the same definition reaches out to stabilized NOI in year two or three. Its disposition yield uses the same construction on the sale side: anticipated next-twelve-month NOI less management costs and replacement capital, divided by gross sale price.

Those deductions matter because a rate computed before management and replacement capital will exceed one computed after them at the same price. A comparable that omits them has a larger numerator than the subject’s underwriting will carry.

A net-lease owner starts elsewhere. Its acquisition and disposition rate is straight-line contractual annual rent plus anticipated annual net tenant recoveries, divided by purchase or sale price. Straight-line rent spreads scheduled increases across the term, so the numerator can differ from the cash rent collected in the first year. Setting that percentage beside an apartment rate built on operating NOI compares two different income constructions.

Hotels show the size of the basis effect. In one hotel study, going-in cap rates built on projected first-year NOI and purchase price ran 20 to 90 basis points above cap rates built on historical NOI, which reflects investor expectations of NOI growth. Price is common to both calculations, so the spread comes entirely from the basis chosen.

Even after the formulas are rebuilt, the original definition keeps its value. A stabilized rate shows how much of the reported percentage depends on the property reaching a condition it has not reached, and a veteran wants that dependence visible in the schedule.

Why closed sales and survey rates do not measure the same thing

A closed sale records an income figure against a price somebody paid. A survey response records the rate a respondent wants. A gap between the two is expected, and it cannot be read as a pricing error or a market move until the nature of each observation is accounted for.

Hotel valuation work states the point directly for cap-rate surveys and transaction data from the same period.

Appraisal averages add a separate layer. In one institutional office database, the average appraisal cap rate was 5.6% in one quarter, while 23 properties sold out of that database in the same quarter at an average transaction cap rate of 8% based on in-place income. Appraisals value the whole database as it stands. Sales come from the subset that traded, and owners choose when to sell. The two averages describe different populations, and membership in one database does not make them the same.

The averaging method matters as well. The transaction figure was an equal-weighted average, which gives every property the same influence regardless of price. A rate computed as combined NOI over combined price gives larger deals more influence. The same trades can yield two different published averages, so a veteran checks the weighting before reading a difference as a trend.

Basis needs checking too. In one annual dataset of full-service hotels appraised at the time of sale, the average cap rate on historical NOI was 7.5%. A sale rate built on projected NOI would sit above a historical-NOI rate for the same trades if projected NOI were higher, and the dataset states which basis it uses.

Aggregates give context. Aggregate data should not set the cap rate for an individual asset, whose rate or pricing parameters depend on its own attributes.

“As the survey data reflect desired rates of return in contrast to rates derived from closed deals, a variance would be expected.”

Worked example: a six percent quoted rate rebuilt on full expenses

This hypothetical compares two readings of one apartment property over the same projected twelve-month period. Both use a purchase price of $20,000,000, revenue of $1,800,000 and no acquisition costs. The only change between them is which expenses are deducted.

Quoted case. The marketing setup deducts $600,000 of operating expenses. NOI is $1,800,000 minus $600,000, or $1,200,000. The cap rate is $1,200,000 divided by $20,000,000, multiplied by 100, or 6%.

Rebuilt case. The buyer’s underwriting adds an owner management fee of $90,000 and replacement reserves of $110,000 that the marketing setup left out. Those two items total $200,000, so expenses become $600,000 plus $200,000, or $800,000. NOI is $1,800,000 minus $800,000, or $1,000,000. The cap rate is $1,000,000 divided by $20,000,000, multiplied by 100, or 5%.

The headline gap is 6% minus 5%, or 1 percentage point, which is 100 basis points. Price did not move. The entire gap comes from $200,000 of deductions.

The error grows once the quoted rate is carried into a valuation. Capitalizing the quoted $1,200,000 at 5% gives $24,000,000. Capitalizing the rebuilt $1,000,000 at 5% gives $20,000,000. The difference is $24,000,000 minus $20,000,000, or $4,000,000, which is $4,000,000 divided by $20,000,000, multiplied by 100, or 20% of the price.

The 6% rate reflects a numerator with fewer deductions and signals no cheaper entry. On identical expense treatment the two readings agree at 5%.

Reading a rate change: did price move or did income move

Holding one variable fixed isolates the other. With NOI constant, a higher price produces a lower cap rate, and rising values compress rates when income does not change. With price constant, higher NOI produces a higher cap rate. A rate that moved reveals nothing about the cause until one of the two is pinned down.

Two separate reconciliations keep the causes apart. The income reconciliation records the period, the revenue assumptions and each deduction. The price reconciliation records the denominator, such as gross purchase price. Whatever difference survives both belongs to the properties, such as location, tenancy, lease terms or condition, and it can then be analysed as a property difference.

Entry and exit rates sit in different places in a model. Because the entry rate divides by price, a change in it between two deals could come from either a higher or lower price on the same income or a different NOI projection on the same price, so the denominator and numerator must be checked separately before concluding which one moved. The exit rate, sometimes called the terminal or reversion rate, is the expected capitalization rate at sale. Neither is a holding-period return, because a cap rate relates one year of NOI to price or value and says nothing about yield over a hold or IRR.

For an investment committee or family office reviewing a schedule of comparables, the test is whether the comparison supports the conclusion attached to it. A rate built on forecast income supports a statement about forecast income. A statement about relative pricing needs matched income definitions and a property comparison that accounts for the subject’s attributes.

Cap rate comparison mistakes that survive a polished schedule

Mixing income periods is an arithmetic slip that a tidy schedule hides. A trailing rate and a stabilized rate in adjacent columns look consistent because both are percentages. Their numerators cover different years, so the column difference mixes timing with pricing. A correct schedule records the period behind each numerator and shows any bridge between them separately.

Reading a lighter expense load as a better price is a second error. At an unchanged price, fewer deductions raise NOI and the reported rate together, as the 6% versus 5% case above shows. A schedule that lists NOI without its deductions cannot tell an operating difference from a calculation difference, and management costs and replacement capital belong inside the definition so a reader can reconstruct what was capitalized.

Treating a survey rate as a transaction is a third. A desired return and a rate derived from closed deals measure different things, and matching the income basis does not close the gap. Each percentage needs its observation type beside it.

Applying an average to a single building is a fourth. A population mean carries no adjustment for the subject’s tenancy, location or condition, so a precise average can still be an inapplicable comparable. The average describes the group it was drawn from, and the asset is priced on its own attributes.

Equating an entry-exit spread with a return is a fifth. The difference between an entry rate and an exit rate does not compute what an investor earns over a hold. That result needs the ownership-period cash flows and the capital invested, and a cap rate by definition covers one year of NOI against price or value.

Checklist questions for a comparable cap rate schedule

What annual income period does each rate use? Historical income and projected first-year income describe different operating periods, and the schedule needs a column for it.

Does any rate depend on a later stabilized condition? A stabilized acquisition rate carries a timing assumption that stays visible beside its percentage.

Which expenses are deducted before NOI is stated? Light expenses raise the cap rate at an unchanged price.

How are management costs and replacement capital treated? Their inclusion or omission changes the annual income being capitalized.

Is the numerator straight-line contractual rent plus recoveries, or an operating NOI? Those constructions differ and need a bridge before the percentages share a column.

What price sits in each denominator? Gross purchase or sale price is a specific definition, and the comparison needs it applied consistently.

Does the rate come from a closed deal or from a survey response? Only a closed deal records a price actually paid.

How was each aggregate rate averaged? Equal weighting and combined NOI over combined price can give different results from the same trades.

Which property differences remain after the arithmetic is reconciled? An individual asset’s rate still depends on its own attributes.