Two speeds. Home prices in 5 of 15 big metros are at a record; in 7 they are still 3 % to 5 % under a peak.

Five metros are at a record and seven are 3 % to 5 % under a peak.

A home price index says where prices stand against where they have been. We took the S&P Case-Shiller index of 15 big metros, seasonally adjusted and monthly, and set the August 2026 reading against the highest reading the series has ever had, from January 1987 or whenever each series begins.

Miami, New York, Chicago, Boston and Washington are at a record: their index in August 2026 is the highest it has been. They have not moved together. Miami is 81.4 % above January 2020 and New York 73.2 %, while Washington is up 42.6 %. Three more, San Diego, Atlanta and Los Angeles, are 0.6 %, 0.6 % and 0.7 % under peaks set in 2025.

Seven metros are not close. Seattle (4.8 %), Phoenix (4.8 %), San Francisco (4.9 %), Tampa (4.4 %), Denver (4.1 %), Dallas (3.7 %) and Las Vegas (3.3 %) are all more than 3 % under a peak. For five of them that peak is from spring 2022, 50 to 52 months ago.

Tap a band below to see its metros, how far each is from its peak and how much each has gained since January 2020.

What an index does not say. It tracks repeat sales of single-family homes in a metro, so it does not show what an apartment building, an office or a particular street is worth, and it is a three-month moving average, so the August reading includes June and July. A metro 4 % under its peak is not a metro in distress: all seven are still 33.8 % to 63.0 % above January 2020. How we read the index

The 3 % line is a convention we use to count, not a signal. At 1 % the count of metros under a peak is 7 of 15, at 3 % it is 7, at 4 % it is 5. Why 3 %

The metros that rose fastest into 2022 are the ones that stalled.

If prices are not flat, what separates the two groups? Not the size of the rise since January 2020: the median metro is up 57.3 %, and the seven that sit under a peak are up 33.8 % to 63.0 %, the five at a record 42.6 % to 81.4 %. The difference is when the gain came.

Measure the rise from January 2020 to the highest reading up to the end of 2022. Phoenix rose 67.6 %, Tampa 67.1 %, Miami 63.4 %, Dallas 54.4 % and Seattle 52.9 %. Chicago rose 28.4 %, Washington 28.5 %, New York 35.2 % and Boston 36.5 %. Of the four slowest risers into 2022, all four are at a record now. Of the two fastest, Phoenix and Tampa, neither is: they are 4.8 % and 4.4 % under a peak. Miami is the exception, third fastest into 2022 and still at a record.

From August 2022 the gap is plain. New York is up 28.1 %, Chicago 26.8 %, Boston 16.0 %. Phoenix is 2.8 % lower, Tampa 2.3 %, Dallas 2.1 % and Denver 1.6 %. Four years later, four of the 15 metros are below where they stood in August 2022. The median metro is 7.8 % higher.

The last 12 months keep the split. Ten metros are up, from 0.5 % in Atlanta and Phoenix to 6.7 % in Chicago and 6.2 % in New York. Five are down: Seattle 2.0 %, Las Vegas 1.3 %, Dallas and Denver 0.8 % each, Tampa 0.4 %. The median is 1.4 % and the nation 2.1 %. Switch the views to see all five measures.

Price is not the whole story and we do not claim a cause. The pattern is that the metros that ran furthest in the first two years of the cycle had the least left, and the ones that ran least have kept climbing. Why is outside the data we read: income, supply, insurance cost and the number of homes for sale all differ by metro, and a price index on its own cannot rank them. What we did not test

What it means for a lender. A value assumption built on one national trend is wrong in both directions: too low in New York, Chicago or Boston, where the market is at a record and an exit at today’s level is no stretch, and too high in Phoenix, Dallas or Seattle, where the same assumption would put a buyer at a price the market has not paid since 2022. A loan written in 2022 in the second group is on collateral that has not recovered; a loan written then in the first has been helped by the market. The check is by metro and by date of origination, not by average.

A shoulder in 2022 separates the metros under a peak from the metros at a record.

The chart sets each metro’s index at 100 in January 2020 and follows it to August 2026, 80 months. All 15 climbed from 2020 into 2022. Then they split. Phoenix, Dallas, Seattle, Denver and San Francisco turned down in 2022 and none has regained its 2022 high; the others kept rising or paused, and five are at a record now.

Phoenix reached an index of 167.6 in June 2022 and is at 159.5 now. Chicago is at 162.5; it was at 128.3 at its 2022 high. Miami has the highest index of the 15, 181.4, and San Francisco the lowest, 133.8. The nation, the dashed line, ends at 155.5.

Pick a metro to bring its line forward against the grey of the other 14. The shape to look for is the shoulder in 2022: the metros with a sharp shoulder have a gap under their peak; those without one do not. That is our reading of the lines.

For an underwriter the date is the point. A property bought or refinanced in the middle of 2022 in a metro with a shoulder carries an entry price the market has not matched in four years; a sale or refinance now is priced against that. An exit assumption that prices grow from here must be tested against the metro’s own path, and the sensitivity to a flat or lower price is the case to run for the metros in the third band.

What to do with this

If you lend on or buy homes and apartment buildings: read value by metro and by the date a loan or a purchase was made, not against one national line. Two metros 5 % apart under a peak can be 40 points apart in what they gained into 2022.

Titleman reads the credit file of a property loan, the offering memorandum, rent roll, operating statement and collateral, and returns the issues and the questions for the borrower. The review can set a deal’s value and exit assumptions against the price path of its metro, so the check is the same for the 200th loan as for the first.