How Nashville Homes housing research is built

What data does the research use?

The September 26, 2026 edition analyzes ILHM’s supplied Nashville September report, public Realtor.com ZIP summaries, Zillow and Redfin city metrics, Redfin metro search flows, Greater Nashville REALTORS historical category tables, and Census county estimates. These populations and reporting windows are kept separate. No listing-level MLS export was used.

Source: ILHM publisher information for the supplied Nashville report · Realtor.com Nashville market summary, retrieved September 26, 2026 · Greater Nashville REALTORS historical category tables

How are the luxury ratios calculated?

Divide reported monthly sales by reported active inventory. For a combined price group, add the sales counts and inventory counts separately before dividing; do not average the published percentages. The analysis applies ILHM’s thresholds to unrounded ratios: buyer below 12%, balanced from 12% to below 21%, seller at least 21%. The three combined single-family groups follow published band boundaries, not statistically estimated turning points. The PDF leaves its Nashville boundary unspecified.

How were the 15 ZIP codes selected?

The sample contains the 15 ZIPs shown in the Nashville summary’s listing-price table. It is a convenience sample rather than every Nashville ZIP or a transaction-weighted city measure. Each ZIP contributes one vote to directional counts. The summaries are labeled September 2026 while their charts run through August; we describe a September 26 retrieval, not completed September results. The individual 37204 page’s 81-day DOM is used rather than the parent comparison’s 75.

Which fields were excluded?

The raw 37218 sold median of $7,702,546 was quarantined because its cause could not be established. The public ZIP download contains inventory and DOM fields only. Other extreme sold-median changes are not labeled appreciation. Redfin’s ambiguous city sales-count text is excluded from volume comparisons. An unsupported August GNR year-over-year field in the supplied analysis code is not used. GNR category comparisons are limited to the recovered January–March 2026 rows.

Which comparisons are intentionally separate?

An ILHM attached home is not automatically a GNR condominium. A Redfin metro search is not a completed move, and its city sales geography is not its search-flow geography. Zillow’s modeled value is not Realtor.com’s listing median. We never average those prices into a synthetic Nashville price. The ten displayed Redfin inbound counts define only the denominator for the reported top-ten share.

Can you reproduce the calculations?

The downloads retain the input counts and reporting windows. Sales-to-inventory is sales ÷ inventory. Percentage change is (current ÷ previous − 1) × 100; percentage-point change subtracts two percentages. The historical GNR comparison divides category inventory by that month’s closings and is not its published months supply or a prediction of selling time. Thirty-one independent arithmetic checks reconciled the principal article calculations to the supplied CSV inputs.

What would listing-level MLS data add?

Deduplicated listing histories could test concessions, original-to-final price changes, relisting behavior, property characteristics and unsold listing age. The current public aggregates cannot answer those questions. A future MLS release would identify permitted fields, exact dates, geographic boundaries, exclusions and calculation code before making those findings.