Method · Multifamily

How to read pricing power
in a multifamily market

Comps tell you what your neighbours are asking. They do not tell you what the market can absorb. This is the model Sentinel runs to answer the second question — built entirely from public data, and published here in full.

The two rules

Everything below depends on two constraints. Break either one and the model produces confident nonsense.

  • Macro is a guardrail, comps set the dollar. Public data establishes the band a market can support — the ceiling and the floor. It never produces the asking price for a specific unit. A model that names a rent from federal data alone is overreaching.
  • Fair Market Rent and ACS median rent are affordability floors, never comparables. They describe what a market must support to remain liveable. Sliding them into a comp set is the most common way a well-intentioned pricing model starts recommending rents nobody can pay.

The seven signals

Each resolves to the county level, then combines into a single banded read — not a number, a band. The band is the output that survives contact with reality.

Signal 01

Rent level vs. the affordability floor

Where asking rent sits against HUD Fair Market Rent and the local income distribution. This is a floor and a ceiling check, never a comp — see the second rule below.

Signal 02

Income and cost burden

Census ACS household income and the share of renters paying more than 30% of it on housing. A market where renters are already stretched has less headroom regardless of what comps say.

Signal 03

Employment base

BLS wage and employment data by county. Rent follows payroll with a lag; a market adding jobs behaves differently from one holding steady at the same rent level.

Signal 04

Supply in the pipeline

Census building permits. Permits issued today are competition eighteen months from now, and are the single most reliable early warning of a softening market.

Signal 05

Observed rent trend

Published rent indices over a trailing window, used for direction and momentum rather than absolute level.

Signal 06

Occupancy

Your own occupancy, normalised. The one signal that is internal — and the one that decides whether a strong market read is actually actionable this month.

Signal 07

Concession pressure

Whether the market is buying occupancy with free rent. Concessions are the market telling you the truth before the asking rents do.

Sources, all public

  • HUD — Fair Market Rents and Income Limits. Note the entity id is a ten-digit county code; bare ZIP or CBSA codes will simply fail.
  • U.S. Census Bureau — American Community Survey for income, rent and cost burden; Building Permits Survey for supply in the pipeline.
  • Bureau of Labor Statistics — QCEW for employment and wages by county.
  • Published rent indices — for trend and momentum only, never for level.

No proprietary feed is required to run this. That is deliberate: a method you cannot check is a method you should not trust.

The dataset

The banded read across the markets Sentinel tracks publishes once the current cycle closes. We publish the method before the numbers — a dataset you cannot reproduce is marketing, not research.

Questions about the method

What is pricing power in multifamily?
Pricing power is the headroom a property has to raise rent without losing occupancy. It is a property of the market and the asset together — not a number you can read off a comp set, because comps tell you what people are asking, not what the market can absorb.
Can you calculate pricing power from public data?
Partially, and that is the point. Public sources — HUD Fair Market Rent, Census ACS income and cost burden, BLS employment, Census building permits, and published rent indices — establish the band a market can support. They do not set the dollar figure. Combining them gives you a strategic guardrail, which is a different and more durable thing than a price.
Is HUD Fair Market Rent a comparable?
No, and treating it as one is a common and expensive error. Fair Market Rent is an affordability reference — a floor that describes what a market must support, not evidence of what a particular unit should ask. The same is true of ACS median rent. Both belong in the guardrail, never in the comp set.
How often should a market read be refreshed?
The underlying federal sources update on quarterly and annual cycles, so refreshing more often than quarterly mostly re-reads the same data. Occupancy and concession signals move weekly and should be read on that cadence.

Want this read on your portfolio?

Sentinel runs this model daily against the markets you operate in. Tell us where you own and we will show you the band.