Occupancy: The Single Most Watched Real Estate Metric

Occupancy is the share of a building's leasable area that is let, and it is two different numbers. In Nilgiri Real Assets Fund I's office, invented, 3,00,000 square feet are leasable and 2,64,000 are let, so physical occupancy is 88.0 per cent. Two tenants on 24,000 square feet are inside a rent-free period, so economic occupancy is 2,40,000 over 3,00,000, being 80.0 per cent.

Start with something smaller than a building. Suppose a household has built two extra rooms on the roof and lets them out. Both rooms have tenants who have signed and moved in. Ask the household how full the roof is and the answer is obvious: completely. Now ask a second question. How much rent came in this month? The room needed painting and the tenant did the painting, so one of the two tenants agreed to move in on the understanding that the first two months carry no rent. So the roof is fully let and half paid. Both answers are true, neither is a mistake, and the person asking has to say which of the two questions they actually meant. Occupancy is that question and nothing more, and the rest is the same situation at 3,00,000 square feet with the arithmetic written down.

What is occupancy actually measuring, and what sits underneath it?

Occupancy is a ratio, and a ratio is only ever as clear as the two numbers it is built from. The bottom of the ratio, the denominator, is leasable areaThe floor space in a building that can be let to a tenant, measured in square feet.: the floor space in the building that can be let to somebody at all. In Nilgiri Real Assets Fund I's office, invented, that is 3,00,000 square feet, and it is the one figure in this guide that does not move. Lifts, stairwells, plant rooms and the lobby are not in it. Leasable area is fixed by the building, not by anything a tenant does.

The top of the ratio, the numerator, is where all the trouble lives. There are two entirely reasonable things to count. One is the space that somebody has signed a lease for. The other is the space that is actually producing rent on the day the figure is struck. In most buildings on most days those two sets of square feet are the same, and the question never comes up. When they differ, occupancy stops being one number and becomes two, and both of them are correct.

What does physical occupancy count?

Physical occupancyLet area divided by leasable area. counts signatures. The measure asks how much of the leasable area sits under a lease that somebody has signed, and it does not ask whether one rupee has yet arrived under any of them. The numerator is let areaThe floor space that is under a signed lease, whether or not rent is being paid on it., and in Nilgiri Real Assets Fund I's office, invented, that is 2,64,000 square feet of the 3,00,000 leasable. 2,64,000 divided by 3,00,000 is 88.0 per cent, and that is this invented building's own physical occupancy on its own leasable area.

Left over are 36,000 square feet with nobody in them and no signature against them. The 36,000 are not separately quoted anywhere in this fund's schedule and do not need to be, being simply 3,00,000 less 2,64,000. Physical occupancy is a statement about commitment: it says how much of the building has somebody contractually attached to it, and nothing at all about money. It is the figure that moves when a lease is signed and when a lease ends, and it is deaf to everything that happens in between.

There is a good reason this measure exists rather than being an accident of sloppy reporting. A signature is a hard fact with a date on it. Rent arriving is a softer fact that depends on the tenant, on the terms, and on which month is being looked at. Physical occupancy is the honest answer to what has been agreed about a building, and it is the answer that changes least often.

What does economic occupancy count instead?

Economic occupancyPaying area divided by leasable area. counts money, or more precisely it counts the square feet that money is currently attached to. The numerator becomes paying areaThe floor space that is actually producing rent on the day the figure is struck.: the floor space that is producing rent on the day the figure is struck. Everything else falls out, whether it is empty or merely not yet paying.

In this invented office, 2,64,000 square feet are let but two tenants occupying 24,000 square feet between them are not paying yet. Take those 24,000 out of the numerator and 2,40,000 square feet remain. 2,40,000 divided by 3,00,000 is 80.0 per cent, and that is this invented building's own economic occupancy on its own leasable area, on the same day the physical figure reads 88.0 per cent. Economic occupancy is a statement about cash: it says how much of the building is working right now, and nothing about how much of it has been promised.

Notice what economic occupancy quietly does. Economic occupancy pushes two very different kinds of square foot into the same bucket. A floor with nobody in it and a floor with a tenant who is unpacking boxes and paying nothing both count as not occupied. To the arithmetic they are identical. One of them has a signature attached and the other does not, so to anybody who has to decide something about the building they are not remotely identical.

Why would a landlord let space and be paid nothing for it?

Because it was a term of the deal rather than an accident. A rent-free periodA stretch at the start of a lease during which the tenant occupies and pays no rent. is a stretch at the beginning of a lease during which the tenant has the space and pays no rent for it. The rent-free period is written into the lease, agreed before anybody moved in, and it is one of the things a landlord can offer in exchange for a tenant committing to everything that follows.

The street-level version is everywhere. A tea stall owner lets a friend put a sandwich counter in the corner of the shop and takes nothing for the first two months. The friend will be there for two years and the corner was empty anyway. A wedding hall lets a caterer store equipment free until the season starts. In each case somebody is giving up income now in exchange for something longer. Nothing about a rent-free period is a fault in the building or a failure by anybody, and treating it as one is the most common misreading of the gap it creates.

How a rent-free stretch gets negotiated, how long one usually runs, how it is priced against the rent that follows and what a landlord gives up to get a signature are all questions about how to let a building, and they belong to a separate subject. The rule that matters is narrower and completely mechanical: for as long as a rent-free period is running, those square feet sit inside let area and outside paying area, and that is precisely why the two occupancy measures come apart.

How can one building be 88.0 per cent and 80.0 per cent full on the same day?

Here is the whole split, worked to the square foot on Nilgiri Real Assets Fund I's asset 1, invented. The asset is a grade-A office building the fund bought for Rs 1,20,00,00,000. Leasable area 3,00,000 square feet. Let area 2,64,000 square feet. Physical occupancy 88.0 per cent. Inside that let area, two tenants on 24,000 square feet are in a rent-free period. So the building holds 36,000 square feet that are empty, 24,000 square feet that are let and paying nothing, and 2,40,000 square feet that are let and paying. Economic occupancy is 2,40,000 over 3,00,000, being 80.0 per cent.

Eight points of difference, on one building, on one day, and neither number is wrong. Nothing was estimated, nothing was assumed and no judgement was applied. The two figures answer two different questions and they were both answered accurately.

The split is really three-way and only ever gets reported two ways. Now look at the same building from above rather than as a pair of bars.

What is the eight point gap worth in money?

Eight points sounds like a rounding argument until it is converted into rupees, and this building allows that exactly rather than approximately. 24,000 times 11 is 2,64,000 with nothing left over, so the 24,000 rent-free square feet are exactly one eleventh of the 2,64,000 square feet of let area. So the rent that area is not paying is one eleventh of the building's gross rental incomeThe rent a building produces in a year before any running cost is taken off..

Nilgiri Real Assets Fund I's office, invented, carries a contracted rent of Rs 62.00 a square foot a month on its let area, and its gross rental income for the year is Rs 19,63,44,000. One eleventh of that is Rs 1,78,49,455 to the nearest rupee. The building's net operating income for the year is Rs 15,00,00,000, this invented fund's own figure on its own cost of Rs 1,20,00,00,000. Divide the one by the other and the yield on costA year's income divided by what was paid for the asset, and nothing more. is 12.5 per cent. Strip out the rent the rent-free area is not paying and Rs 13,21,50,545 is left. On the same Rs 1,20,00,00,000 of this fund's own cost, Rs 13,21,50,545 is 11.01 per cent.

Eight points of occupancy is 1.49 points of yield on this invented fund's own cost, and the only thing that changed between the two figures is which square feet were counted. The 11.01 per cent is carried to two decimals on purpose. The 12.5 per cent is the figure this fund's own record states for the building. The 11.01 per cent is worked out from it and keeps its own precision rather than borrowing the single decimal of a figure it was not taken from. Rounded to one decimal it would read 11.0, and 12.5 less 11.0 is 1.5. The subtraction does not actually give 1.5. Nothing was bought, nothing was sold, no tenant left and no rent was renegotiated.

One warning about that rent figure. A careful reader will try to check it. The Rs 62.00 a square foot a month is the headline contracted rent on this invented building's let area, and the Rs 19,63,44,000 is the gross rental income its schedule states for the year. Neither figure is derived from the other. Where a rent a square foot is needed for arithmetic, the average across the let area is Rs 61.98 a month. Take the annual gross, divide by twelve months and then by 2,64,000 square feet. A headline rate and an annual total are two separate facts on a schedule, and multiplying one into the other asserts something the schedule did not say.

What is each of the two numbers actually good for?

Neither measure is the better one. The two measures answer different questions, and the reason to keep both is that a reader almost always has one of those two questions in mind and rarely says which.

Put the two questions in front of somebody and the choice is usually obvious. The difficulty is that people rarely ask a question out loud before reading a number off a schedule. People read the number first and attach a question to it afterwards. The order is exactly backwards.

Which number does a schedule usually show?

Usually the physical one, and usually with no label saying so. The missing label is not a conspiracy. Physical occupancy is the figure that falls out of a lease register with the least work: count the square feet with a signature against them, divide by the leasable area, done. Economic occupancy needs somebody to look inside each lease and pull out which ones are currently paying. The extra work produces a figure that changes more often.

So the line a reader is handed tends to look like a fact and is actually half of one. An occupancy figure carries a number and almost never carries its basis. The reader has to go and get the missing half.

What is the one question to ask of any occupancy figure?

Which one is it. Let area or paying area. The basis is the whole question, it takes four seconds to ask, and every other question about the figure depends on the answer to it.

The question buys something narrow, and the narrowness is worth stating. Knowing the basis does not settle whether the building is doing well, and no level of either measure is good, poor, healthy or worrying in itself. The answer buys the ability to know what has been said. The number on its own is not the information; the number plus its basis is the information, and a figure with no basis attached is the half that is easiest to misuse.

If the answer comes back as physical, a second thing follows for free: the economic figure is either the same or lower, and it can never be higher. A square foot cannot be paying rent without being under a lease, so paying area is a subset of let area by construction. So physical occupancy sets a ceiling on economic occupancy on every building on every day, and the distance between them is the rent-free area and nothing else.

Why is a high occupancy figure not by itself the whole story?

Two reasons, and they are different from each other. The arithmetic one comes first, and the numbers already in hand prove it.

Economic occupancy at 80.0 per cent reads exactly the same whether the missing eight points are empty floors or floors that are let and inside a rent-free period. The measure is linear in paying area and it does not care why a square foot is not paying. On Nilgiri Real Assets Fund I's office, invented, those 24,000 square feet have signed tenants sitting in them. A building with 60,000 square feet standing genuinely empty and nothing rent-free would report the same 80.0 per cent economic occupancy on the same 3,00,000 square foot denominator, and the two buildings are in visibly different positions. The blindness is not a flaw in the measure, it is the definition of the measure, and a reader who wants to tell those two buildings apart has to ask for the split rather than for a better ratio.

The second reason is about time. Both occupancy measures are struck on one day and say nothing about how long the leases behind them have to run. Nilgiri Real Assets Fund I's office has 1,90,000 square feet sitting on leases with more than three years left and 74,000 square feet on leases expiring within eighteen months, and those two figures add back to the 2,64,000 square feet of let area exactly. A building at 88.0 per cent with 74,000 square feet expiring soon is a different proposition from a building at 88.0 per cent with none, and the headline figure cannot tell them apart because it was never trying to. The denominator matters wherever that 74,000 appears: it is 24.7 per cent of the 3,00,000 leasable and 28.0 per cent of the 2,64,000 let, and a sentence quoting it without saying which has said something imprecise. The schedule of when leases end, and what that schedule shows, are covered under lease expiry.

What does somebody reading an asset schedule actually do with these two numbers?

Three different people read the same line for three different reasons, and each of them wants a different one of the two measures. The three readings are what turn a definition into a habit.

An analyst inside the manager, preparing the quarterly asset schedule that goes to investors, has to decide which figure to print and label it. If the schedule prints physical occupancy without a label, every reader downstream inherits the ambiguity, and the analyst is the last person in the chain who could cheaply have removed it. The habit here is a column, not a footnote: basis, stated on the row.

A lender sizing a loan against a building is doing something narrower. The lender is asking what cash is available to service interest over the next several quarters, so the physical figure is close to useless on its own and economic occupancy is the starting point. A rent-free period that runs for another two quarters is two quarters of cash that will not be there, and a lender who read 88.0 per cent and stopped has mis-sized the cover. The measure a reader needs is decided by the horizon of the decision, and a lender's horizon is the next few payment dates rather than the life of the leases.

An investor deciding whether the two figures on a schedule are consistent with each other has the easiest job of the three, and almost nobody does it. Take the reported occupancy, take the reported income, and see whether they agree. On this invented office, 88.0 per cent of 3,00,000 square feet at the contracted Rs 62.00 a square foot a month does not reproduce the schedule's own Rs 19,63,44,000 of gross rental income for the year, and the reason it does not is that 24,000 of those square feet are not paying. A reader who does that one check finds the rent-free period without anybody having disclosed it. The consistency check is the practical payoff of the whole distinction.

What goes wrong when two buildings are compared on occupancy?