Main Content

December 2025 San Francisco Market Report

In 2025, Bay Area real estate markets started the year with strong buyer demand, but had the wind taken out of their sails by political and economic uncertainty that surged in early spring. Then, in mid to late summer, stock markets rebounded to hit new highs, and interest rates began a sustained decline, initiating a shift in buyer and seller psychology that is still developing.

In San Francisco, however, the market has been supercharged by its AI startup boom, a dynamic that dramatically accelerated this past fall. After years of social, economic, and demographic challenges, the city has shifted from being the weakest market in the Bay Area to the strongest. Stupendous amounts of new wealth are being created, and market conditions may well exceed those last seen during the IPO boom of 2019.

Demand is far outpacing the supply of homes for sale, and we expect the city’s median house sales price to reach a new high in 2026. The luxury home market is particularly competitive, and in our most expensive district—which includes Pacific Heights, Presidio Heights, Cow Hollow, and the Marina—median house and condo prices already reached new peaks in 2025. The condo market, while not as heated as the house market, has begun a substantial recovery fueled by high-tech workers flocking back to the city.

As 2026 begins, interest rates are near multi-year lows and stock markets are at or close to all-time highs. Early in the new year, it is common for buyers—re-energized to move forward with major life decisions—to reenter the market faster than sellers list their homes for sale. This imbalance of increasing demand versus a limited supply of new listings typically persists through the spring, making these months the most heated selling season of the year. Last year, this pattern was disrupted by the tariff shock. Barring new, unexpected economic alarms, we expect this spring to be very heated indeed.

 

ID must not be a non-empty string.
Slide Image #2
Slide Image #3
Slide Image #4
Slide Image #5
Slide Image #6
Slide Image #7
Slide Image #8
Slide Image #9
Slide Image #10
Slide Image #11
Slide Image #12
Slide Image #13
Slide Image #14
Slide Image #15
Slide Image #16
Slide Image #17
Slide Image #18
Slide Image #19
Slide Image #20
Slide Image #21
Slide Image #22
Slide Image #23
Slide Image #24
Slide Image #25
Slide Image #26
Slide Image #27
Slide Image #28
Slide Image #29

 

DISCLAIMER: Statistics are generalities, essentially summaries of widely disparate data generated by dozens, hundreds or thousands of unique, individual sales occurring within different time periods. They are best seen not as precise measurements, but as broad, comparative indicators, with reasonable margins of error. Anomalous fluctuations in statistics are not uncommon, especially in smaller, expensive market segments. Last period data should be considered estimates that may change with late-reported data. Different analytics programs sometimes define statistics – such as “active listings,” “days on market,” and “months supply of inventory” – differently: what is most meaningful are not specific calculations but the trends they illustrate. Most listing and sales data derives from the local or regional multi-listing service (MLS) of the area specified in the analysis, but not all listings or sales are reported to MLS and these won’t be reflected in the data. “Homes” signifies real-property, single-household housing units: houses, condos, co-ops, townhouses, duets and TICs (but not mobile homes), as applicable to each market.

City/town names refer specifically to the named cities and towns, unless otherwise delineated. Multi-county metro areas will be specified as such. Data from sources deemed reliable, but may contain errors and subject to revision. All numbers to be considered approximate. Many aspects of value cannot be adequately reflected in median and average statistics: curb appeal, age, condition, amenities, views, lot size, quality of outdoor space, “bonus” rooms, additional parking, quality of location within the neighborhood, and so on. How any of these statistics apply to any particular home is unknown without a specific comparative market analysis.

Median Sales Price is that price at which half the properties sold for more and half for less. It may be affected by seasonality, “unusual” events, or changes in inventory and buying trends, as well as by changes in fair market value. The median sales price for an area will often conceal an enormous variety of sales prices in the underlying individual sales.

Dollar per Square Foot is based upon the home’s interior living space and does not include garages, unfinished attics and basements, rooms built without permit, patios, decks or yards (though all those can add value to a home). These figures are usually derived from appraisals or tax records, but are sometimes unreliable (especially for older homes) or unreported altogether. The calculation can only be made on those home sales that reported square footage.

Photo use under the Creative Commons License: https://creativecommons.org/licenses/by-sa/2.0/

Compass is a real estate broker licensed by the State of California, DRE 01527235. Equal Housing Opportunity. This report has been prepared solely for information purposes. The information herein is based on or derived from information generally available to the public and/or from sources believed to be reliable. No representation or warranty can be given with respect to the accuracy or completeness of the information. Compass disclaims any and all liability relating to this report, including without limitation any express or implied representations or warranties for statements contained in, and omissions from, the report. Nothing contained herein is intended to be or should be read as any regulatory, legal, tax, accounting or other advice and Compass does not provide such advice. All opinions are subject to change without notice. Compass makes no representation regarding the accuracy of any statements regarding any references to the laws, statutes or regulations of any state are those of the author(s). Past performance is no guarantee of future results.