How cap rates work

A capitalisation rate answers one question: what yield does this year's net operating income represent at this price? Write it as a fraction:

$$\text{cap} = \frac{\text{NOI}}{\text{price}}$$

On $36,000 of NOI and a $480,000 asking price that is $36000 / 480000 = 0.075$, 7.50 percent. If next year's income were identical, and if you paid all cash, that 7.50 percent is the cash yield the building throws off. The moment income moves, or a loan sits on the property, it is no longer the return. It is still the cap.

That is the same object as a dividend yield on a stock: this year's cash over today's price, with no credit for growth and no charge for a change in the price. A 7.50 percent cap and a 7.50 percent dividend yield are cousins. Neither is total return.

The calculator above opens on those figures because they are the first worked example, not because 7.50 percent is a typical market cap. Typical depends on the city, the asset type, and the year, none of which belong inside a formula.

Net operating income is the building's rent minus the costs of keeping it rented: property tax, insurance, maintenance, management, a vacancy allowance. Two things stay out on purpose.

Debt service stays out. The mortgage is a fact about the buyer, not about the building. Two people can bid the same $480,000 on the same $36,000 of NOI, one in cash and one with a loan, and they have the same cap rate. They do not have the same cash-on-cash return. Mixing them is how a listing and an offer argue about a number that was never the same object.

Income tax stays out. The tax a buyer pays depends on the rest of that buyer's return, the jurisdiction, and the entity. Cap rate is a property identity, which is why those sit outside it.

The usual padding is at the NOI line itself. A seller who quotes gross rent as NOI has not subtracted the costs, so the cap is overstated. A buyer who treats a roof replacement as an operating cost has understated NOI, so the cap is understated. The formula divides the number it is given. It will not catch a padded rent roll.

Vacancy is the line that looks small and is not. A building quoted as fully let at $36,000 of rent, with 8 percent vacancy ignored, is not a $36,000 NOI building. Take 8 percent off and the income the cap should see is already lower before any operating cost is subtracted. The cap on the asking price then prints high, which is the seller's number, not the building's.

A one-off capital item is not an operating cost. Spreading a new roof across 20 years of NOI is a judgement about how to think about the building. Putting the whole roof into this year's NOI is a way to make a 7.50 percent cap look like 5. Neither treatment is the formula's job. Naming which one you used is.

A cap on its own is a yield. A cap against another cap is a price. Rearrange the identity:

$$V = \frac{\text{NOI}}{\text{cap}}$$

The same $36,000 of income at a 6 percent cap is worth $36000 / 0.06 = 600000$, $600,000. Offered at $480,000, this building sits $120,000 below that implied value. Equivalently, it is offered at 7.50 percent while the comparison is 6 percent.

Which direction is cheap depends entirely on which cap is the market's. If similar buildings really do trade at 6 percent, a 7.50 percent offer is a lower price for the same income. If they trade at 8 percent, this 7.50 percent offer is a higher price. The calculator will not tell you the market cap. It will tell you what this income is worth at the cap you type.

A second building with $50,000 of NOI at $625,000 is an 8 percent cap. At a 7 percent comparison cap that income is worth $714,285.71, so the $625,000 price sits $89,285.71 below that value. Two buildings, two caps, one comparison rate each, and the gap is now in money rather than in percentage points.

When the comparison cap equals the property's own cap, implied value equals the asking price and the gap is 0. That is a restatement, not new information. It is also the check that the arithmetic is wired the right way round.

Borrowing to buy the building does not change the cap rate, because NOI and price do not include the loan. What borrowing changes is the return on the cash you put in, and how fragile that return is, which is the leverage ratio question applied to a house rather than a company balance sheet.

A higher cap can support more debt service, because more of this year's income is available before the loan is considered. It does not become a higher cap because you took the debt out. A 7.50 percent cap on $36,000 of NOI is 7.50 percent whether the buyer writes a cheque for $480,000 or borrows most of it.

Cash-on-cash return divides cash after debt service by cash in. That ratio moves when the loan moves. Cap rate does not. Report the levered figure as a cap and you can no longer compare it to any listing that used the property identity, which is most of them.

The rent versus buy question is a household version of the same split: the building's yield is one input, the mortgage is another, and folding them into one number hides which of the two is doing the work.

Cap rate is a one-year snapshot. If NOI is going to grow, a buyer will accept a lower cap today, because some of the return is arriving as growth rather than as this year's yield. That is the same split as dividend yield versus total return, and the Gordon identity $k = \text{yield} + g$ is the stock-market version of it. A cap rate is the yield piece. It is silent on $g$.

Two buildings with the same $36,000 of NOI this year are not the same building if one of them has a lease that steps 3 percent a year and the other is vacant next winter. The cap on both can print 7.50 percent. The prices will not match, and they should not.

For a growing income stream priced as a present value, the dividend discount calculator is the closest tool on this site, even though it is written in dividends rather than in rent. The algebra is the same: next year's income over required return minus growth. A two-stage DCF is that idea with an explicit forecast in front of the perpetuity.

What you should not do is take a 7.50 percent cap, add a 3 percent growth rate, and call 10.50 percent the return. That addition is Gordon's $k$ only if the cap is next year's yield on today's price and $g$ is perpetual. This year's NOI over today's price is not next year's. Off by one year of growth is how a tidy-looking 10.50 percent gets baked into an offer.

A higher cap is a higher yield on this year's income, and it usually means a lower price for that income. It can also mean the income is less certain, the building is older, the lease is shorter, or the location is weaker. Cap rate does not sort those causes. It only reports the ratio.

That is why a cap is almost never used alone. It is read against the caps on similar buildings, against the same building's own cap last year, and against the required return the buyer actually faces. A 7.50 percent cap in a 6 percent market is a different sentence from a 7.50 percent cap in an 8 percent market, as the comparison-cap identity already showed in money.

A bond's current yield is the cousin that makes the limit clear. Current yield is this year's coupon over price. Yield to maturity also counts the pull to face value. Cap rate has no maturity and no face, so it is closer to current yield than to YTM. There is no date at which the building is redeemed at a known amount. The terminal value, if you want one, is another cap applied to a later NOI, which is a DCF, not a cap rate.

The real return calculator is the inflation step if the 7.50 percent is nominal and prices are rising. A 7.50 percent cap at 3 percent inflation is not a 7.50 percent gain in what the income buys.