Keenan Luxury Market Index: Methodology

What the index is

The Keenan Luxury Market Index (KLMI) tracks what Austin luxury homes actually sell for, measured in price per square foot. It covers closed single-family sales between $2M and $10M across Travis, Williamson, and Hays counties, read two ways: band medians that keep price tiers separate, so a $2.4M sale never drives a $5M-$10M reading, and a composite that deliberately pools the full $2M-$10M universe - one median across all of it, not an average of the band medians. We built it because no public index measures this segment well - citywide medians drown luxury in volume, and the big national indexes lose reliability above $2M, where sales are few and no two homes are alike.

KLMI is computed from MLS-reported closings under a licensed data feed, not from surveys, estimates, or a syndicated third-party product. Every figure we publish traces to one defined computation on real closed sales. This page documents that computation in full - what goes in, what stays out, and where the index can be wrong - so you can judge for yourself whether to rely on it.

How it is computed

The derivation, start to finish:

1. Start with closed sales. We take every closed single-family sale recorded in the MLS for Travis, Williamson, and Hays counties with a sold price between $2M and $10M. Leases, land, condos, and townhomes are out - the published index is single-family only. We track Austin's luxury condo market separately, and those sales are never mixed into these bands.

2. Compute price per square foot for each sale. Sold price divided by total square footage, one figure per sale. Sales that compute below $100 or above $3,000 per square foot are dropped as data errors - almost always a mis-keyed square footage.

3. Take the median within each price band. The bands are $2M-$3M, $3M-$5M, and $5M-$10M, each including its lower edge and excluding its upper, so the index covers $2M up to but not including $10M. We also publish a composite $2M-$10M reading computed from the same pooled sales, not by averaging the band medians. We use the median of per-sale figures, not total dollars divided by total square feet. The median resists distortion from a single unusual sale, which matters in bands where a quarter may hold only a few dozen closings.

4. Read it two ways. Luxury volume is thin, so a single quarter of sales makes a noisy median. Each quarterly reading is a rolling 12-month median ending at quarter close - stepped quarterly. The quarterly series behind our pinned editions uses the 12 months ending at each quarter's close, labeled by the quarter's start date - so the Q1 2026 reading covers the 12 months ending March 31, 2026, and the Q2 2026 reading covers the 12 months ending June 30, 2026. Sales counts are 12-month counts on the same window. The current snapshot on our Market Report reads the last complete quarter's 12-month window. The daily refresh revises the sales inside that window as late closings post; the headline window itself advances only when a quarter closes.

5. Index it. The index applies to the composite alone: the $2M-$10M reading is divided by its value in the base period, 2021-Q1, and multiplied by 100. A reading of 118 means the composite median price per square foot is 18% above its base-period level. The three price bands publish median price per square foot and year-over-year change - they do not carry index values. The series runs back to 2021-Q1, and its first window draws on closings from April 2020 forward.

In the 12 months ending March 31, 2026, the $2M-$10M universe held 477 closed sales.

What it deliberately excludes, and how it can be wrong

A methodology you cannot poke holes in is a methodology nobody wrote down honestly. Here are ours.

Thin bands. The $5M-$10M band can hold very few sales in a given quarter, and at the neighborhood level even fewer. Any band with fewer than 10 sales in its 12-month window is suppressed - we publish no figure rather than a misleading one. On top of that structural floor, our own publication rule requires at least 5 sales in both the current window and the year-earlier comparison window before a figure appears in a published edition; in practice the 10-sale floor makes that automatic. Published figures carry a confidence tier - firm for 30 or more sales, indicative for 10 to 29 - so you can see how much data sits behind each number.

Texas is a non-disclosure state. Sold prices are not public record here. KLMI is built from MLS-reported closings, which is the most complete sold-price dataset that exists for this market - but a sale that closes fully off-MLS is invisible to it. In the $5M+ range, where private sales are more common, assume some undercount.

Recent readings can revise. Closings are sometimes reported to the MLS days or weeks late. The index recomputes from the full MLS record every day, so a late-reported closing automatically revises recent readings. A quarter's reading is provisional until late reports settle - a quarter that closed three or more months ago carries minimal remaining revision risk - and we treat the newest, still-open window as in-progress rather than headline it. If a published figure revises, the edition pages stay pinned to their original readings and the live index carries the correction.

Price per square foot is a blunt instrument. It says nothing about condition, lot, views, or renovation quality - the things that actually set luxury prices. KLMI measures the market's level and direction. It does not price your home, and we say so to clients directly. More on the metric's blind spots: Austin Luxury Price Per Square Foot: Why This Metric Misleads.

Out of scope by design. Sales under $2M (not Austin's luxury tier in practice) and over $10M (too few and too idiosyncratic for a quarterly median to mean anything - we track them, we do not index them).

Where the data comes from, and compliance

The underlying records come from ACTRIS (Unlock MLS), the MLS covering Austin and Central Texas, replicated through the MLS Grid API under license. The database syncs continuously through the day, the published index recomputes daily, and every computation runs against the licensed record directly - there is no manual entry step between a closing and the index.

We publish aggregates only: medians, counts, and index values. No individual sold price appears in KLMI outputs or on our public pages, in line with MLS rules and Texas non-disclosure practice. Verified account holders may view sold prices under our MLS VOW license terms. That public-page constraint is permanent, and it is why you will see a band median here but not what any specific home closed for.

How to cite the index

You are welcome to reference KLMI figures with attribution. Use this line:

Source: Keenan Luxury Market Index, Keenan Group (Compass), Austin TX - thekeenangroup.com

For a specific edition:

Keenan Luxury Market Index, Q2 2026 edition. Keenan Group, Austin TX. Methodology: thekeenangroup.com/the-source/keenan-luxury-market-index-methodology

License: Keenan Luxury Market Index figures, tables, and the compiled index are released under the Creative Commons Attribution 4.0 International license (CC BY 4.0). You may copy, republish, chart, and build on the index, including for commercial use, provided you credit Keenan Group and link to this page or to the edition you are quoting.

Questions about the data, or a figure you think is wrong? Email us. If you find an error, we would rather correct it than defend it.

Update cadence

The live index recomputes daily, revising readings inside their windows as late closings post; the headline window advances when a quarter closes. Quarterly editions publish within a few weeks of quarter close - the Q2 2026 edition pinned its readings on July 12, 2026 - and are pinned: their figures stay locked to what the data showed at publication. If this methodology changes, we will date and describe the change on this page. Method changes are rare by intent: a benchmark you keep redefining is not a benchmark.

Frequently asked questions

How is the Keenan Luxury Market Index calculated?

We take every closed single-family sale between $2M and $10M in Travis, Williamson, and Hays counties from licensed MLS data, compute price per square foot for each sale, drop data-error outliers, and publish the median within each price band. Each quarterly reading is a rolling 12-month median ending at quarter close, stepped quarterly; the current headline reading is the last complete quarter's window, recomputed daily as late closings post and advanced when the next quarter closes. The KLMI index value applies to the composite $2M-$10M series only - its reading divided by its 2021-Q1 level, set to 100 - while the bands publish median price per square foot and year-over-year change. The full derivation is on this page.

Why price per square foot instead of median price?

Median sale price moves whenever the mix of homes sold moves - a quarter heavy with estate sales reads as a price jump even if nothing appreciated. Price per square foot normalizes for size, which makes quarters comparable. It has real blind spots (condition, lot, views), which is why we band by price and publish medians rather than pretend one number prices a home.

Why $2M to $10M?

Below $2M is not Austin's luxury tier in practice - it behaves like the broader market. Above $10M, sales are so few and so idiosyncratic that a quarterly median is noise; we track those sales but do not index them. $2M-$10M is the range where there is both a genuinely distinct luxury market and enough closings to measure it honestly.

Can I republish KLMI figures?

Yes, with attribution. Use the citation line on this page and link to it or to the edition you are quoting. Figures are released under CC BY 4.0, which permits reuse, including commercial reuse, as long as Keenan Group is credited. If you are publishing analysis built on our figures and want the underlying aggregates, contact us.