CRE market evidence

EQUIRE holds four independent commercial real estate evidence bases. They answer overlapping questions on incompatible bases, so the product reports them side by side with each basis labelled and does not combine them into a single figure.

The four bases

Securitized transactions

What it measures. Loan and property observations taken at the moment a mortgage was securitized: cap rate as NOI over valuation at securitization, physical occupancy at securitization, DSCR, NOI per square foot, appraised value per square foot, original loan size, interest-only share, and tenant concentration.

Where it comes from. Asset-level disclosure filed with the SEC on Form ABS-EE, exhibit EX-102, under Regulation AB II. The connected read returns available filings for the requested market and property type.

Grain and coverage. In the 14 September 2026 snapshot, the producer reported 348 securitizations from 348 filers, with 39,836 property records and 26,874 loans across 778 metros and 14 property types. These are historical dataset counts; current availability varies by metro and property type. Underlying loans span vintages, so each cap rate is dated by its securitization, not by today; use the returned sample count and source period for a specific read.

The caveat that matters. Every figure on this base is measured at securitization, not currently. A property securitized in 2019 contributes its 2019 occupancy. This base is also skewed toward institutional and larger assets, because those are what get securitized.

Broker asking listings

What it measures. The asking side: quoted lease rates, asking cap rates, asking price per square foot, available square feet, and the count of lease versus sale offerings.

Where it comes from. EQUIRE's indexed listing inventory, collected from brokerage sites.

Grain and coverage. The 14 September 2026 snapshot recorded 119,618 active listings across 832 metros and 9 property types. Current asking-side availability varies by geography, property type, and inventory refresh. Use each result's geography, sample counts, and source/collection dates.

The caveats that matter. An asking figure is an offer, not a signed trade. The headline asking rent is pooled across lease bases, so use the NNN-specific figure when the basis matters. Use each result's geography, sample, and collection date; coverage varies by market and property type. The residential rent indicator in economic indicators is a separate measure.

CBRE Cap Rate Survey

What it measures. Surveyed cap rate ranges by market and property type, reported as a low, high, and midpoint. EQUIRE surfaces the Class A stabilized rows from the most recent edition, and the full half-yearly history for trend.

Where it comes from. CBRE's published U.S. Cap Rate Survey.

Grain and coverage. The 14 September 2026 snapshot contained 10,183 rows spanning H1 2011 through H2 2025, across 59 markets and 5 property types. The connected survey read exposes metro/property-type rows and a half-yearly history. Check the period, market, property type, and stabilization basis attached to each returned figure.

The caveat that matters. A survey reports practitioner sentiment about where cap rates sit. It is not a record of closed trades, and a survey range and a transaction median are not interchangeable.

REIT quarterly disclosures

What it measures. Two different things. Operating fundamentals: same-store revenue growth, same-store NOI growth, economic occupancy, and effective rent change year over year. Transaction yields: the weighted-average initial cash yield REITs disclose on acquisitions and dispositions.

Where it comes from. REIT quarterly earnings supplementals filed with the SEC on Form 8-K, exhibit EX-99.

Grain and coverage. The 14 September 2026 snapshot contained 791 operating observations across 32 metros and 11 sectors, and 293 transaction cap observations across 19 sectors. Coverage depends on what each REIT discloses. Figures may be metro-level or national by sector; the product labels the returned geography, sector, and quarter.

The caveat that matters. The connected REIT read reports same-store rent change, not an effective-rent level. Other provider-specific level data may exist on a different basis; it is not part of this read.

Why the bases are not blended

A blended median of securitization and asking observations is arithmetically trivial and analytically worthless, because the mix behind any one cell is unpublished. Two metros can differ by 80 basis points purely because one cell is mostly securitized institutional product and the other is mostly small broker offerings. The difference reads as a market signal and is a composition artifact.

The upstream producer marks the blended cap rate and price-per-square-foot columns as deprecated for exactly this reason, and replaces each one with a stated-basis pair: an appraised-at-securitization figure and an asking figure, each with its own sample count. EQUIRE strips those deprecated columns out of the metro-index rows it hands the assistant, so a pooled distribution is not available to quote from that read. The benchmark rows returned alongside them, which the producer labels national, are stripped on the same principle: they are percentiles taken across metros of a single metro-index column, not national totals, so each one inherits that column's basis, and the one built on the deprecated vacancy column is withheld rather than labelled. Each benchmark figure that does survive arrives with its basis stated and a note that the row's market count is not the sample behind any figure in it.

The practical consequence: when you ask about cap rates in a metro, you should expect more than one number and a label on each. A single unqualified figure would be hiding the question of which market you were asking about.

The metro index

The four bases are joined into per-metro, per-property-type rows so a surface can read them together without merging their meanings. The 14 September 2026 snapshot covered 741 metros; current availability varies by metro, property type, and source refresh. Market research shows observation dates separately from refresh time, because a refreshed index can still carry older observations.

Where this evidence appears

The metro index is deliberately not available in deal chat. A deal has its own market ledger in Market research, which confirms the metro and property type, records the run, and shows each figure beside its basis and sample. An assistant answering the same question from the raw index alongside that page would produce a second, unrecorded set of market figures in the same workspace. The stated-basis siblings stay available on every surface.