Civexa AI/ Platform/Market intelligence
Market intelligence

Know what the market will support

Civexa reads the market evidence that exists in public record for the parcel's own area, and labels each source precisely — because a rent benchmark and a market asking rent are different things, and underwriting them as the same thing is expensive.

The question

What have comparable properties actually sold for, and what rents would underwrite here?

Recorded sales, not estimates

Transactions that happened, with addresses and dates

Comparable sales come from the county's recorded-sales layer: the address, the recorded price, the date and the distance from the subject parcel. These are transactions that completed and were recorded — not listings, not automated valuations, and not an index.

Trend from the same record

Median by year, with the sample size behind each one

The price trend is the median of recorded residential sales in the area by year across a three-year window, reported with the number of sales in each year so you can see how much weight the median can carry. A year with nine sales is shown as a year with nine sales.

Rent benchmarks, labelled as benchmarks

HUD Small Area Fair Market Rents, and what they are not

Rent figures are HUD Small Area Fair Market Rents for the parcel's ZIP: 40th-percentile gross rent including utilities, published as a programme benchmark. They are useful as a conservative underwriting floor. They are not market asking rents, and Civexa says so on every surface they appear rather than presenting them as comparables.

This is the honest limit of what Civexa reads today. Where you have your own rent survey, entering it replaces the benchmark and the model recalculates.

In the report

What the market section reports

Every figure names its source, and the sources are public record.

1806 E Mercer St$1,335,000 · 0.33 mirecorded
1808 E Mercer St$1,375,000 · 0.33 mirecorded
Median trend, 2023–2026+5.2% totalderived
2-bed rent benchmark$2,630 / monthHUD SAFMR
Rent basis40th-percentile gross, incl. utilities

Illustrative figures from a real Civexa run. Run your own address for live, code-traced numbers.

What it will not do

The answers Civexa declines to give

A model that always produces a number is a model whose confident answers cannot be told apart from its guesses. These are the cases where Civexa reports a gap instead.

  • A benchmark is never relabelled a comparable. HUD Fair Market Rents are identified as programme benchmarks everywhere they are used, including inside the financial model.
  • Absorption is not forecast. Civexa does not project lease-up pace or sell-out velocity from the data it reads, because that data does not support the projection.
  • Competitive supply is not modelled. Pipeline and competing-project analysis is not something Civexa reads today, and the site does not imply otherwise.
  • No automated valuation is produced. Civexa reports what sold and what the trend was. It does not issue a value opinion for the subject property.
Questions

Market intelligence, in detail

Are these asking prices or recorded prices?
Recorded. They come from the county's recorded-sales layer, which reflects transactions that completed, with the date each was recorded.
Why use HUD rents rather than market rents?
Because HUD Small Area Fair Market Rents are published, consistent, ZIP-level and free of survey bias, which makes them a defensible conservative floor for a screening model. They read low against market asking rents, which is why Civexa labels them as benchmarks and lets you replace them with your own figure.
Does Civexa forecast the market?
No. It reports what is in the public record and what the trend has been. Forecasting absorption or competitive supply would require data Civexa does not read, and inventing it would defeat the point of the rest of the product.
Next

What this connects to

Development capacity

What the programme is that the market has to absorb.

See capacity →