Vacancy in a Brazilian REIT (FII): what it is, why only brick funds have it, and the CVM data trap

Vacancy measures how much of a property sits empty, producing no rent. It sounds simple — and it is, until you look at the raw CVM data and find it means different things depending on who filled it in. Almost no one explains that trap.

What vacancy is

It's the share of a property's area that is empty — square metres producing no rent. At Dados B3, each fund's vacancy is the area-weighted mean of its properties: a 130,000 m² warehouse weighs far more than a 500 m² store. The figure comes from the quarterly report the fund files with the CVM (inf_trimestral) and is point-in-time — the vacancy as of that report's date, not a snapshot from today.

Why it matters only for brick funds

In a brick (tijolo) fund, rent is the income. Empty area is income that doesn't come in — which is why vacancy is the fund's health signal. A paper (CRI) fund, by contrast, owns no building at all: it holds debt, receivables. There's no property to sit empty, so it has no vacancy — an honest absence, not a missing figure. The difference between the two models is in Brick vs paper.

The CVM field trap. The Percentual_Vacancia field carries inconsistent meaning across administrators. Some fill in the real fraction of empty area (12.4%). Others fill in 0 or 1 as if it were occupancy — and a top-tier shopping-mall fund, with its malls 100% operating, shows up with "1" on every property. Read naively, that "1" becomes 100% vacancy. A raw average ends up publishing "91% vacancy" on a packed mall — an absurd number resting on a filling convention, not on the reality of the property.

How Dados B3 handles it

We flag these cases. When a fund-quarter is dominated by properties marked at ~100%, or when aggregate vacancy comes out ≥70%, the record is marked suspect and excluded from clean rankings and medians. The number isn't deleted — it's kept with the flag, for audit — but it is never shown as if it were clean. The rule is the usual one: we'd rather show the inconsistency than publish a number that looks right and lies. A reader should distrust any source showing a huge vacancy on a fund that is obviously operating.

How to read the vacancy that remains

Once the noise is cleaned out, the number earns its keep:

Low and stable is a good sign — the properties are leased and the income is predictable.
Rising pressures the distribution: less leased area, less rent, less dividend ahead.
Genuinely high in a single-building fund — a corporate office at 50%+ vacancy — is a legitimate signal of distress, not data noise. It's the kind of vacancy that tends to travel alongside a very low P/BV.

Vacancy never decides on its own. Read it together with P/BV and dividend yield: rising vacancy with a falling P/BV tells a far clearer story than either alone. The full checklist is in How to analyze a FII.

Vacancy by fund and segment medians are on the FIIs page; each fund's own figure is at /fiis/TICKER.

See FII vacancy

Honest limitations

It's a quarterly snapshot: between one report and the next, real vacancy may have shifted.
An area-weighted mean doesn't tell a good property sitting empty from a bad one.
Not investment advice. Vacancy is one signal among several, not a verdict.