Cap Rates: What They Measure and What They Hide
The capitalisation rate is one year of income divided by price. That simplicity is its usefulness and its danger.
Investing · 592 words · updated 2026-08-26
Cap rate = Net operating income ÷ Value
It is a snapshot yield on unlevered, in-place income for a single year. Everything else an investor cares about — growth, capital expenditure, lease rollover, financing, exit — sits outside the formula.
Getting the NOI right
Cap rate comparisons fail far more often on the numerator than the denominator. A defensible NOI:
- Uses effective gross income — contract rent plus reimbursements and other income, less vacancy and credit loss — not gross potential rent.
- Deducts a market management fee even if the current owner self-manages.
- Deducts a replacement reserve for recurring capital, typically stated per unit or per square foot.
- Excludes debt service, depreciation, amortisation, income tax, and non-recurring items.
- Is stated as trailing twelve, in-place annualised, or forward year one — and the three are not interchangeable. A cap rate quoted on a forward NOI that assumes a lease-up is a projection, not a yield.
Ask which NOI a quoted cap rate uses before comparing it to anything.
Going-in, exit and the spread that matters
- Going-in cap rate — year one NOI ÷ purchase price.
- Exit / terminal cap rate — assumed NOI in the year after sale ÷ sale price. Underwriting convention is to assume an exit cap at or above the going-in cap; assuming compression is assuming the market does the work.
- Spread to the risk-free rate — the cap rate less the ten-year government bond yield. This spread, not the absolute cap rate, is how capital markets price the asset class over time. A 6.0% cap against a 1.5% ten-year is a very different proposition from a 6.0% cap against a 4.5% ten-year.
What a low cap rate actually says
A low cap rate is a high price per dollar of current income. It is rational where the buyer expects income growth, where the income is unusually secure, or where the asset is in a supply-constrained location. It is irrational where the buyer is simply extrapolating the last cycle. The relationship is:
Cap rate ≈ Discount rate − Growth rate
So a 4.5% cap implies either a low required return, a high expected growth rate, or both. Testing whether the implied growth is plausible against the market's actual rent and supply data is the whole exercise.
Where cap rates mislead
- Heavy near-term rollover. A stabilised-looking 6.5% cap on a building where 60% of the space expires in 24 months is not a 6.5% asset. Look at the weighted average lease term and the mark-to-market on expiring rents.
- Deferred capital. NOI does not deduct the roof, the chillers or the façade. Two identical cap rates with a $6 million capital backlog on one are not the same trade.
- Below- or above-market in-place rent. An asset let 20% below market has embedded upside that the cap rate understates; one let above market has downside it hides.
- Ground leases and non-standard structures. A leasehold interest and a fee interest do not price on the same yield basis.
- Thin comparable sets. In smaller markets and specialised asset types there may be too few genuine comparables to establish a market cap rate at all. Say so, rather than manufacturing precision.
Using cap rates well
Use them to normalise price across otherwise comparable assets, to convert a stabilised income assumption into a value, and to sanity-check a discounted cash flow's terminal value. Do not use them alone to decide whether an asset is cheap. Unlevered IRR and equity multiple on a cash flow that models rollover, capital and financing will tell you far more.