By Zurani
Index and auction data show that how far art’s value moves, and whether it sells, depends on the segment, the price level and the years measured. This is what the main published figures say, and what they leave out.
There is no single figure for how art performs. The published indices and auction data describe several smaller markets, divided by segment and by price, and these move at different speeds and sometimes in opposite directions. Arts Economics estimates that global art sales rose 4% in 2025 to US$59.6bn after two years of decline, which left the market below its 2022 peak of US$68.1bn and 7% below its 2015 level [1]. A total of that size says little about any single work. For someone weighing art as part of their wealth, the risk mostly comes down to which segment a work sits in, its price level, the years over which it is held, and whether it can be sold when needed. What follows is a reading of the main published sources on each of those points, and of what their figures leave out. It makes no recommendation.
What does “risk” mean when the asset is a work of art?
It is the chance that a work is worth less than expected, or cannot be sold on acceptable terms, at the moment it has to be sold.
Market data can show part of that. It shows how prices or sales values have moved within a segment, how unevenly value is spread across price levels, and how often works offered at auction fail to sell. It cannot show the risks attached to a particular object: whether the attribution holds, whether the seller has good title, what condition the work is in, and whether its history can be documented. Those questions belong to due diligence on the work itself, which our articles on art investment fraud warning signs and on provenance cover in detail. A work can sit in a rising segment and still lose most of its value if its attribution fails.
How much did different art segments move?
On Knight Frank’s index, all six art segments fell over the five years to the end of 2025, and only one rose over ten years.
Knight Frank, a real-estate advisory firm, publishes the Knight Frank Luxury Investment Index (KFLII) in its annual Wealth Report. It is a modelled index, a weighted average of asset performance built from third-party data, and for 2025 it was expanded to split art into six segments [2]. The report lists Artnet among its data sources but does not explain how each art sub-index is constructed. The figures below are cumulative changes in the index over each period. They are index movements, not the net return an owner would receive.
Over the 12 months to the fourth quarter of 2025, five of the six art segments rose: Impressionist art by 13.6%, modern art by 7.1%, post-war art by 5.2%, the top 100 artists by 3.6% and European Old Masters by 1.7%. Contemporary art fell 6.0% [2]. Over five years every segment was down, from 2.9% for European Old Masters to 17.8% for contemporary art, with modern art down 13.0% and the top 100 artists down 13.8% [2]. Over ten years, five of the six were flat or lower, and only European Old Masters rose, by 2.2% [2]. The composite index, which also covers wine, whisky, cars, watches and other luxury assets, rose 38.6% over the same decade [2].
These rows come from one firm’s model. The six segments are also not separate slices of the market: the top 100 artists grouping overlaps with the other categories, since an artist in it may also fall within the Impressionist, modern, post-war or contemporary segments, so the rows should not be added together or read as shares of a whole. Nor are they comparable with the Art Basel and UBS figures used later in this article, which measure the total value of sales rather than prices. The chart and table below set the three periods side by side.
| Segment | 12 months | 5 years | 10 years |
|---|---|---|---|
| European Old Masters | +1.7% | −2.9% | +2.2% |
| Post-war | +5.2% | −4.7% | −0.2% |
| Impressionist | +13.6% | −8.5% | 0.0% |
| Modern | +7.1% | −13.0% | −9.3% |
| Top 100 artists | +3.6% | −13.8% | −8.1% |
| Contemporary | −6.0% | −17.8% | −0.3% |
Cumulative change in the Knight Frank Luxury Investment Index to Q4 2025, before any costs. Source: Knight Frank, The Wealth Report 2026. The top 100 artists grouping overlaps with the other segments.
Why does the time period change the answer?
The same segment can look strong or weak depending on the dates chosen, and the published sources say so themselves.
Impressionist art is the clearest case in Knight Frank’s figures: up 13.6% over 12 months, down 8.5% over five years and unchanged over ten [2]. Someone looking only at the last year and someone looking back to 2020 would describe the same segment very differently.
Arts Economics makes the same point about its own auction indices, which track the total value of sales in each price band. In its words, performance “depends on the start and end points” [1]. Measured from 2010, only the band above US$10m ended 2025 higher than it started, and the value of sales between US$50,000 and US$250,000 was down 29%. Measured from the 2009 low, every band rose, the top band by almost 240% and the low end by only 8%, which the report notes would be a decline once inflation is taken into account [1].
Shorter periods can also show large swings, as in these examples, although a few years’ figures do not establish a general rule. By the end of 2024 the value of auction sales above US$1m had fallen to less than half its 2022 high, and the value of sales in the US$10m-plus band was down 67% over the same period [1]. In 2025 sales in that top band rose 30% by value [1]. These are changes in the total value sold, which depends on how many expensive works owners chose to consign. They do not mean that a US$10m painting lost two thirds of its value.
Does the price level of a work matter?
Yes. Auction totals are concentrated in a small number of very expensive lots.
At fine art auctions in 2025, 95% of lots sold went for less than US$50,000, yet together they made up only 15% of sales value. Works sold for more than US$1m were under 1% of transactions and 54% of value, down from 61% in 2022, and the US$10m-plus band alone accounted for 22% [1]. High-value sales can disproportionately influence auction-market totals.
The survey points the same way for the collectors it covers. Of the 3,100 high-net-worth collectors surveyed for Art Basel and UBS in 2026, across 10 markets that did not include the Gulf, 77% spent less than US$50,000 on fine art in the first half of 2026 and 84% less than US$100,000; 3% spent more than US$1m [4]. For most of these collectors, spending was well below the price levels that dominate auction totals.
How easily does art sell?
Not always. Of the lots offered at fine art auctions in 2025, excluding lots confirmed as withdrawn and lots whose prices were not disclosed, 30% were bought in, meaning they did not sell [1].
That share had fallen for the second year running to its lowest level in six years, and buy-ins at Sotheby’s, Christie’s and Phillips fell to a reported 13% or lower as demand improved during the year [1]. Failure reaches the top of the market too: Alberto Giacometti’s Grande Tête Mince (1955), estimated at around US$70m, was bought in at Sotheby’s in May 2025 [1].
A buy-in rate measures failed auction sales. It does not show how long a private sale takes or what discount a seller might have to accept. In its 2026 collector survey, Arts Economics notes that art remains very illiquid and seldom produces income, and suggests this may be one reason the average share of wealth that the surveyed collectors hold in art has fallen since 2022 [4]. The quality of the work, the estimate, the venue and the timing all affect whether a sale succeeds.
What do these figures leave out?
They leave out most of the costs an owner actually pays, and usually inflation as well.
Index movements and sales totals describe the market. An owner’s net return depends on the full purchase cost, including any buyer’s premium, the sale proceeds after the seller’s commission, and the costs of holding the work in between: insurance, storage, conservation, shipping and tax. None of the figures in this article is net of those costs. The sources also differ in what their prices include: the Bank of America and ArtTactic figures are hammer prices, excluding the buyer’s premium [3], so they are lower than the totals buyers paid. Holding costs such as art insurance are paid every year whatever the market does. Unless a source says otherwise, its figures are also nominal, and Arts Economics points out that an 8% gain since 2009 would be a loss after inflation [1].
Sales-value indices reflect supply as well as price. A year in which several large collections come to auction lifts the total without any change in what comparable works are worth. A valuation is an opinion for a stated purpose at a stated date, and a formal appraisal is how to test what a specific work might be worth now. The sources used here also measure different things: Knight Frank models segment indices, Arts Economics measures global auction and dealer sales, and Bank of America and ArtTactic cover hammer prices at three New York auction houses. Their figures cannot be added together or ranked against each other.
How should a collector read art market data?
Ask five questions of any performance figure before relying on it.
- Does it measure prices, or the total value of sales?
- Which segment and price band does it cover, and does the work in question belong there?
- Which years does it use, and what happens if the start or end point moves?
- Is it gross or net of transaction and holding costs, and is it adjusted for inflation?
- Who publishes it, on what underlying data, and can that data be checked?
Before buying, it is worth weighing the segment, the likely holding period and the route to an eventual sale, and treating any headline number as one reading among several. Zurani’s approach to valuation, liquidity and risk is set out on the art investment page, and governance and reporting for collections held alongside other assets are covered on the family offices page.
If you are weighing a purchase or a sale and want to test it against the evidence, the team can go through it with you. Contact Zurani to request an introduction.
This article is general information for collectors and their advisers. It is not investment, legal or tax advice. Index and survey figures are third-party data, gross of costs, and past movements are not a reliable indicator of future results. It does not describe any transaction or engagement involving Zurani.
Sources
- Arts Economics (Dr Clare McAndrew) for Art Basel and UBS, The Art Basel and UBS Art Market Report 2026 (PDF), March 2026.
- Knight Frank, The Wealth Report 2026 (20th edition), Knight Frank Luxury Investment Index, data to Q4 2025, April 2026: report download (registration required) and Knight Frank’s summary of the 2026 index.
- Bank of America Private Bank, powered by ArtTactic, 2026 U.S. Art Market Report (PDF), February 2026.
- Arts Economics (Dr Clare McAndrew) for Art Basel and UBS, The Art Basel and UBS Survey of Global Collecting 2026 (PDF), October 2026.
Related reading: Art investment fraud: five warning signs · Provenance in art · How to value artwork · Art advisory
