The U.S. rebound: recovery for whom?
Auction concentration, guarantees and the geography of buyers
The 2025 rebound combines important collections, stronger historical categories and greater concentration. Bank transaction data adds a different perspective on where buyers are located.
Original research · Bank of America × ArtTactic · 2026
Reading & commentary · Art Value Lab
2026 U.S. Art Market Report- Sales value rose while volume and artist representation contracted.
- Guarantees and single-owner collections increasingly shape outcomes.
- Auction locations and buyers’ regions describe different geographies.
Figures refer to the source samples and survey period; expectations are not realised outcomes.
Does a stronger total mean broader participation, or the return of a few expensive transactions?
Timing and breadth of the rebound
First-half sales fell 5.6% year on year; second-half sales rose 54.1%, producing 23.1% annual growth. A year-end headline can obscure the weak first half.
Lots sold fell 19.9%, and represented artists declined from 4,112 to 3,315. Recovery in value coexisted with narrower participation, while totals remained below 2021–2023 levels.
Source report pages · 6–7、10–21New York’s strength depends on what comes to market
New York accounted for 69% of value in the report’s global auction sample, above its ten-year average of 62.1%. This is not the U.S. share of all global art transactions.
Single-owner sales grew from 5.1% of global sample value in 2015 to 27.3% in 2025. Major collections combine scarcity and provenance, but their irregular availability creates volatility.
Take with youAlongside demand, ask what owners were willing to sell.
Source report pages · 14–21Guarantees redistribute risk
Guaranteed value reached 78.3% of New York evening sales, the highest share since 2015 in the report. Consignor certainty increasingly depends on arrangements made before the auction.
Strong sell-through or performance against estimates does not remove risk. Selection, estimates and withdrawals also matter, while guarantors, houses and bidders have different incentives.
Source report pages · 6、14–21Historical categories lead an uneven recovery
Impressionist and Modern sales grew, while Post-War, Contemporary and Young Contemporary categories contracted. Active supply does not necessarily translate into stronger prices.
Women artists show notable long-term growth, but gender is neither a style nor a uniform investment category. Period, scholarship, supply and institutional support remain important.
Take with youCategory trends begin an inquiry; they do not replace work-specific context.
Source report pages · 22–31Where buyers live is not where auctions occur
Bank transaction data offers a demand-side geography. Buyers in California, Florida, New York and Texas accounted for 46% of overall transaction volume in that dataset in 2025, with greater concentration at higher price levels.
A California collector can buy at a New York auction. Because the bank sample also includes galleries and online markets, it complements rather than contradicts auction-location data.
Source report pages · 32–35、42Resale statistics and owners’ actual outcomes
Evening-sale resales averaged 4.4% annualised performance in 2025, down from 5.3% in 2024, while holdings below five years were negative. This selected sample highlights timing without proving that waiting guarantees profit.
Transaction, insurance, transport and care costs affect realised outcomes. The evidence supports attention to long-term stewardship more readily than predictions about the next purchase.
Take with youWorks, timing and costs separate market recovery from individual outcomes.
Source report pages · 6、14–21、42The following editorial analysis explores mechanisms, practical approaches and open questions. It extends the source research without presenting new findings on behalf of its authors.
What does a selective recovery mean for market health?
When value rises while lots and artist counts fall, the recovery is unevenly distributed. Recognisable works may attract attention without expanding opportunities for everyone.
Major collections organise supply around provenance and narrative. Their return differs from a broad increase in demand, and their absence next year could change totals without an equivalent change in underlying interest.
Ask whether gains extend across bands, whether artist breadth recovers, and whether sales without major collections improve too. These distinguish event-driven rebounds from more sustained activity without assuming one proven cause.
Analysis references · source pages · 6–7、14–31How guarantees, estimates and selection shape outcomes
Auctions select works, set estimates and negotiate terms before bidding begins. Sell-through reflects these choices as well as demand.
Estimates provide a variable reference point. The same realised price can look stronger against a lower estimate, so comparisons require knowing whether the benchmark is low, middle or high and whether premiums are included.
Guarantees allocate some uncertainty in advance. They can enable important consignments while making the prior arrangements relevant to interpreting public results. Context strengthens rather than negates the information in auction prices.
Analysis references · source pages · 6、14–21Three meanings of the U.S. market
| Unit | What it can address | What it cannot directly establish |
|---|---|---|
| New York auction sales | Activity at a selected sale location | That all buyers live in New York or the U.S. |
| Bank client transactions | Regional patterns within observable clients | A census of all U.S. collectors |
| Evening-sale resales | Price changes for identifiable repeat sales | Net returns and liquidity of all collections |
See pp. 6–7, 32–35 and 42. The datasets complement one another without sharing the same scope.
How can you use this research?
Collectors
Locate the relevant category and price band before applying a recovery headline to a work.
Galleries and researchers
Distinguish sale and audience locations and record discovery-to-purchase paths.
What changes when we read these together?
Methodology & source
Auction analysis uses hammer prices at Christie’s, Sotheby’s and Phillips in New York, 2015–2025, with specified cities and categories for global comparisons. The 2021 Las Vegas MGM Picasso sale is classified as New York. Regional spending uses proprietary Bank of America transactions across auctions, galleries and online markets, not a census of U.S. activity. Resales and guarantees use narrower samples; economic forecasts remain forecasts.
2026 U.S. Art Market Report
Bank of America × ArtTactic · 2026