Vacancy Is Not a Percentage You Lose — It Is a Month You Do Not Collect

A rate is an output, not an input

Ask a landlord for a vacancy rate and you get a percentage: five, eight, ten. Ask where it came from and the answer is usually a rule of thumb or a market report. But nobody experiences a vacancy rate. What actually happens is that a tenant gives notice, the flat empties on a Saturday, it is cleaned and painted over three weeks, it is advertised, it is viewed, and someone moves in six weeks later. The rate is what you get when you divide that event by the period between two of them. It is an output of a physical process, and treating it as an input is what makes it useless for decisions.

The formula is as simple as it looks. If V is the void in months and T is the length of the tenancy that preceded it, the effective vacancy rate over the cycle is V ÷ (T + V). The denominator is the whole cycle, not just the occupied part, which is the mistake most spreadsheets make: dividing by T alone overstates the rate, because those void months are inside the period, not added to it. Six weeks after a one-year tenancy is 1.5 ÷ 13.5 = 11.11 percent, not 1.5 ÷ 12 = 12.5 percent.

Written that way the rate stops being a market statistic and becomes two things you can influence: how long people stay, and how fast you re-let. Everything a landlord can actually do about vacancy acts on one of those two terms. A market report cannot tell you either of them for your own flat, which is why importing its percentage into your own model imports someone else's tenancy behaviour along with it.

The same void, three tenancy lengths, three different rates

Hold the void fixed at six weeks — an unremarkable figure for a flat that needs painting and a letting agent — and vary only how long tenants stay. After a twelve-month tenancy the cycle is 13.5 months and the rate is 11.11 percent, which is 40.6 days empty in an average year. After twenty-four months the cycle is 25.5 and the rate is 5.88 percent, 21.5 days. After thirty-six months the cycle is 37.5 and the rate is 4.00 percent, 14.6 days. After five years it is 2.44 percent, under nine days a year.

The interesting comparison is the pair that lands on the same number by two different routes. A twelve-month tenancy with a two-week void gives exactly 4.00 percent, and so does a thirty-six-month tenancy with a six-week void. Identical rates, opposite operations: one landlord runs a slick re-letting machine on a high-churn property; the other lets the flat sit empty for six weeks and does not care, because it happens once every three years. Both would report the same figure to a lender, and they are not running the same business.

There is a floor under all of this that no amount of skill removes. Even a perfect handover leaves days: notice periods rarely align with move-in dates, deposits have to be settled, an inventory has to be taken, and in most jurisdictions the outgoing tenant has the right to occupy until a date that has nothing to do with when the next one wants to arrive. A landlord who reports zero vacancy over many years is either very lucky, running back-to-back tenancies with overlapping rent, or not counting the days properly.

The void is only part of what a turnover costs

A vacancy rate counts rent not collected and stops there, which understates the event badly. A turnover also costs money out of pocket. On a $1,500 rent, six weeks empty is $2,250 of rent that never arrives. Make the flat ready — clean, repaint the hallway, replace whatever the deposit does not cover — and call it $900. Pay a letting agent one month's rent to find and reference the next tenant and that is another $1,500. Total: $4,650, of which less than half is the void everyone measures.

Spread that over the cycle and the picture changes shape. On the thirteen-and-a-half-month cycle, potential rent is 13.5 × $1,500 = $20,250 and the turnover consumes $4,650 of it: an effective loss of 22.96 percent, more than double the 11.11 percent vacancy rate. On the thirty-seven-and-a-half-month cycle, potential rent is $56,250 and the same $4,650 is 8.27 percent — still double the 4.00 percent rate, but on a much smaller base. Annualised, the high-churn property loses $4,133.33 a year to turnover and the low-churn one $1,488.

Who pays the letting fee is not universal, and in several markets it is not the landlord at all. In Germany the Bestellerprinzip introduced in 2015 for rentals means the party that instructed the agent pays — in practice the landlord, and a tenant may not be charged a commission for a letting the landlord commissioned. Other markets allow a shared or tenant-borne fee within caps. Where the fee genuinely lands on the tenant, the landlord's turnover cost falls to the void plus the make-ready, but the tenant's cost of moving rises, which pushes the same way: longer tenancies.

The break-even rent increase — the number this article exists for

Here is the decision that actually arrives, once a year, in an envelope. Your tenant is paying $1,500. The market says $1,650. Do you raise the rent? Raising it earns you the increase for as long as the tenant stays. It also raises the chance the tenant leaves, and if they do you pay the void and the turnover cost. The break-even is where those two exactly cancel.

Write it out. Over a horizon of H months, an increase of g on a rent R earns H·R·g. If the increase triggers a departure, the cost is the void at the new rent, V·R·(1+g), plus the cash cost C of the turnover. Setting them equal and solving gives g = (V·R + C) ÷ (H·R − V·R). With R = $1,500, V = 1.5 months and C = $2,400, the numerator is $2,250 + $2,400 = $4,650 and the denominator over a twelve-month horizon is $18,000 − $2,250 = $15,750. So g = 29.52 percent, which on $1,500 is $442.86 a month. That is the size of increase a one-year view demands before a six-week void is worth risking, and no ordinary market rises by that much.

Stretch the horizon and the threshold collapses, because the gain repeats and the cost does not. Over twenty-four months the break-even is 13.78 percent, over thirty-six 8.99 percent ($134.78 a month), over sixty 5.30 percent ($79.49). This is why the same increase can be obviously right and obviously wrong depending on nothing but how long you expect the new tenant to stay, and why the honest version of the question is never should I raise the rent but how confident am I that this tenant stays either way. If you are certain the tenant will renew at $1,650, raise it: the void never happens and the break-even is irrelevant. The calculation only bites when the increase is what causes the departure.

Two refinements matter in practice. First, the formula above assumes the departure is certain if you raise; if you think the probability is p, multiply the cost side by p and the threshold falls proportionately — at p = 0.3 over three years the break-even drops to about 2.62 percent, which is why moderate annual increases usually survive the test and large catch-up increases usually do not. Second, in markets with statutory rent-review formulas or caps, the size of the increase is not yours to choose, and the calculation becomes a test of whether to apply the permitted increase at all rather than how much to ask.

A portfolio rate hides the distribution

Ten flats, one of them empty for a whole year, nine let continuously. The portfolio vacancy rate is 12 months out of 120, or 10 percent. Not one of those ten flats experienced 10 percent vacancy: nine experienced zero and one experienced 100. Anyone who buys the ninth flat on the strength of the portfolio number has bought a flat with a 90 percent chance of nothing happening and a 10 percent chance of a catastrophe, priced as though it had a small, steady leak.

The same flattening happens to city-level statistics. A published rental vacancy rate is a stock measure taken at a moment: units available for rent divided by the rental stock, at a survey date. It answers a question about market slack, not about how long your particular flat will sit empty. Two neighbourhoods with the same rate can differ completely in how the vacancy is distributed — a handful of unlettable units in one, a general two-week friction across everything in the other — and only the second tells you anything about your own re-letting time.

Seasonality is the last thing an annual average erases. Lettings cluster: students move in late summer, families move around the school year, and in most European cities the weeks around the winter holidays are the worst possible time to advertise a flat. A tenancy that ends in November therefore carries a longer expected void than the same tenancy ending in June, purely because of when it lands. This is directly actionable at the point where it costs nothing: when you sign or renew, choose an end date in a letting season, even at the price of a slightly odd term length. A fifteen-month first tenancy that ends in June may cost less than a twelve-month one that ends in March.

What to do with the number once you have it

Put the effective rate, not a market rate, into the rental account. In the companion article on what a rental actually earns, vacancy at 6 percent removes $1,980 from $33,000 of gross rent and reduces the management fee with it. Substituting 11.11 percent for 6 percent on that same property removes $3,666.30 and cuts cash flow by roughly $1,550 — more than the whole first-year cash flow of $1,392.56. Vacancy is not a rounding item in that account; it is one of the two or three inputs capable of turning the answer negative on its own.

Then attack the two terms separately, because they respond to different things. The void shrinks with logistics: start advertising during the notice period rather than after the keys come back, do the make-ready in the notice window where the tenancy allows viewings, keep a decorator you can book at short notice, and price the re-let to fill rather than to win. The tenancy lengthens with retention: fix things quickly, do not push increases to the last euro, and make renewal the easy option. The second lever is much larger than the first, and it is the one most landlords ignore because it has no invoice attached.

Frequently asked questions

Why divide by the tenancy plus the void rather than by twelve months?

Because the void happens inside the cycle, not on top of it. A twelve-month tenancy followed by six weeks empty covers 13.5 months of calendar, and during that whole stretch you own the flat and pay its fixed costs. Dividing 1.5 by 12 answers a different question — what fraction of the occupied period the void represents — which is not a vacancy rate and is always higher. The 12-month denominator also breaks completely once the tenancy is not a year: with a three-year tenancy it would give 12.5 percent again, which is obviously wrong.

Is a low vacancy rate always good?

No, and the counter-example is common — but it does not survive a full count, which is the real lesson. A flat that never sits empty is very often a flat let below market, and the discount is paid every month while the void it avoids happens once every few years. Run both: a $1,500 rent empty 4 percent of the time collects $17,280 a year, while a $1,650 rent with an 11.11 percent effective rate collects $17,600 — on collected rent alone the higher rent wins despite the worse-looking rate. Now subtract the annualised turnover cash cost, $768 against $2,266.67, and the order reverses: $16,512 against $15,333.33. Vacancy is a cost to be optimised, not a metric to be minimised, and the quantity to maximise is collected rent net of turnover cost — which is precisely the quantity a vacancy rate on its own cannot show you.

Does the break-even increase mean I should never raise the rent?

It means the opposite of what people usually take from it. The threshold is high precisely because it assumes the increase is what causes the departure — that is, that the tenant would have stayed at the old rent and leaves at the new one. Most tenants move for reasons that have nothing to do with you: a job, a partner, a baby, a house purchase. Where the departure was going to happen anyway, the increase costs nothing and you should take it. The calculation is a test of one specific hypothesis, not a general argument against ever asking for more. What it does argue against is the large catch-up increase after years of holding the rent flat, which stacks the entire adjustment onto a single renewal decision and maximises the chance of triggering the void.

How should I set the vacancy rate for a property I have not bought yet?

Build it, do not look it up. Estimate two things about the property you are considering: how long a tenant is likely to stay in that particular kind of home, and how many weeks it takes to re-let in that particular street. A family house with a garden near good schools produces long tenancies; a studio in a student district produces twelve-month ones almost by construction. Then put the two into V ÷ (T + V). If you want a sanity check, run the purchase again at double the void and see whether the deal still works — a property that only clears at two weeks of void is a property with no margin for a single bad let.

Does the deposit cover the make-ready cost?

Only the part attributable to damage, and almost never the part attributable to wear. Every jurisdiction here distinguishes between damage a tenant caused and the ordinary deterioration of living somewhere, and repainting after three years of normal occupation falls on the landlord in essentially all of them. That is why the make-ready line survives in the model even when the deposit is returned in full: it is a cost of the turnover, not a claim against the tenant. Budgeting it as a recoverable is the fastest route to both an unexpected bill and a deposit dispute.