Art is a $2.56 trillion store of wealth for the ultra-rich but a poor financial asset on the evidence. Deloitte’s Art & Finance Report 2025 shows the Artnet Fine Art Top 100 index returning 3.2% a year over 20 years against 10.4% for the S&P 500, and −2.9% a year over ten years. Buying costs 15–28% in auction premiums, selling costs a further 5–20%, holding costs 0.1–2% a year in insurance plus storage and conservation, and the median auction lot takes months to sell at a 32% unsold rate. Correlation with equities is low mainly because art is appraised rather than priced, and forced sales typically realise 20–30% below fair value. Art can diversify a very large portfolio, fund lending and deliver non-financial returns, but anyone expecting equity-like performance from it should read the numbers first.
Art as an asset class is one of the most heavily marketed and least well-measured ideas in wealth management. In this guide the Kurums Startup team sets out what the 2025–2026 data say about returns, volatility, correlation, liquidity and the full cost of owning art, using the Deloitte and ArtTactic Art & Finance Report 2025, the Art Basel & UBS Global Art Market Report 2026 and the public disclosures of platforms and lenders. It belongs to the Arts & Creative Economy hub and follows our overview of the global art market’s size and structure; the vehicles that package art for investors are covered in art funds and fractional ownership.
What has art actually returned?
Deloitte’s 2025 report cites the Artnet Fine Art Top 100 index at 3.2% annualised over 20 years and −2.9% over 10 years, against 10.4% and 13.3% for the S&P 500. Index figures exclude transaction costs of 20–40% per round trip, so realised investor returns are lower still for most works.
What does it cost to hold art?
Insurance runs 0.1–2% of value a year (below 0.5% for large, well-secured collections), plus storage, conservation and periodic appraisals every 3–5 years. Buying at auction adds a 15–28% buyer’s premium; selling costs 5–20% in commission and fees. A $1 million work can cost $20,000–30,000 a year before any price change.
Is art a good diversifier?
Partly. Measured correlation with equities is low, but much of that reflects appraisal smoothing and infrequent trading rather than independent price behaviour. In stress periods art becomes unsellable rather than uncorrelated: works that fail at auction lose an average of 32% on resale, and forced-sale values run 20–30% below fair value.
What returns has art delivered compared with stocks and bonds?
Over long periods the broad art indices have underperformed public equities by a wide margin. The Deloitte and ArtTactic Art & Finance Report 2025 cites the Artnet Fine Art Top 100 index at 3.2% a year over 20 years versus 10.4% for the S&P 500, and −2.9% a year over 10 years versus 13.3%.
Those figures need two qualifications, one in each direction. Against art: index returns are gross of transaction costs, which in art are enormous (a buyer’s premium of up to 28% on the way in and a seller’s commission plus fees on the way out), and they are built from repeat sales of works that did sell, which excludes the roughly one in three lots that fail to find a buyer at all. Artprice reported an unsold rate of 32% for fine art at auction in 2025. For art: top-100 indices track the most traded blue-chip names, and narrower segments have done better in particular windows. Artprice’s own Artprice100 index, which is constructed from the 100 best-performing artists, is marketed as “regularly outperforming” the S&P 500 since 2000, a claim that depends heavily on survivorship, since the constituents are chosen in hindsight.
The 2025 data make the dispersion vivid. Masterworks, the largest fractional platform, reports in its SEC filings 33 artwork sales since 2020 with net annualised returns from 4.1% to 788.9% and a median of around 17%, with no sale at a loss. But the same SEC filings show that of 265 unsold works marked to net asset value in May 2026, 212 (80%) were carried below their $20 offering price, with a median of $16.94. The platform’s realised sales are a selected sample; the unrealised book tells a different story. We take the structure apart in art funds and fractional ownership.
Guarantee data give a third angle. ArtTactic found that in the first half of 2025, lots sold with a guarantee at the three major houses returned 4.6% (compound annual growth from prior purchase price) while non-guaranteed lots returned 36.4%. Guarantors earned 7.2% on their capital, down from 9.5% in 2024. In other words, the works that need financial support to sell are the ones with the weakest price histories, and the market’s own risk capital earns less than a corporate bond for taking that risk.
Why is art’s correlation with other assets lower than it looks?
Art prices show low measured correlation with equities and bonds because they are observed rarely and smoothed by appraisal, not because art is immune to the economy. When liquidity disappears, as in 2009 and 2024, art does not fall in price so much as stop trading, which an index records as stability.
The mechanics are simple. Indices such as Artnet’s or the old Mei Moses series are built from repeat sales, which for an individual work occur every ten to thirty years. Between sales the work is “valued” by appraisers who anchor to the last comparable auction result. Fund and platform net asset values are set the same way. The result is a series that moves in steps and lags the economy by a year or more, which mechanically lowers measured correlation and volatility. Academic work on private equity and real estate found the same “appraisal smoothing” effect; art is the extreme case.
Realised data from downturns show the true behaviour. Global art sales fell 12% in 2024 (Art Basel & UBS) while the S&P 500 rose 23%; they rose 4% in 2025 when US equities again rose. That is not negative correlation with the stock market; it is a lagged response to interest rates, which reduced the discretionary spending of leveraged collectors in 2023–2024 and then eased. The Deloitte report notes that secondary-market discounts of 20–40% from prior peaks were common for mid-career contemporary artists in 2024, discounts that never appeared in any index because the works simply were not offered.
Art does carry some real diversification value. Its drivers (the number and wealth of ultra-high-net-worth individuals, estate timing, taste, museum validation) differ from corporate earnings, and currency exposure is unusual because the top market is priced in dollars in New York regardless of where the owner lives. For a family office with hundreds of millions in liquid assets, a 5–10% allocation to art that is already owned for other reasons is defensible. For a portfolio measured in hundreds of thousands, the costs below swamp the benefit.
What does it really cost to buy, hold and sell art?
A full round trip through the auction market costs roughly 25–45% of the hammer price once buyer’s premium, seller’s commission, insurance, storage, shipping, appraisal and conservation are counted. Deloitte’s 2025 report is explicit that index comparisons ignore these costs, and they are the main reason realised returns trail index returns.
Buying. Since February 2026 Sotheby’s charges a buyer’s premium of 28% on the first $2 million of hammer price, 22% from $2 million to $8 million and 15% above $8 million. Christie’s, which changed its schedule in September 2025, charges 27% up to $1.5 million, 22% to $8 million and 15% above. Phillips charges 29% up to $2 million. A work hammered at $1 million therefore costs $1.27–1.29 million before shipping and any import VAT, which is 5% in the UK, 5.5% in France and, since 2025, 5% in Italy. Gallery purchases carry no visible premium but a dealer’s margin of 50% of the retail price is standard on primary-market work.
Holding. Fine-art insurance typically costs 0.1–2% of insured value a year, according to insurer and broker guidance compiled by Masterworks’ research arm, with large, well-secured collections in low-risk locations priced below 0.5%. Climate-controlled storage in a freeport or specialist warehouse runs into thousands of dollars a year for a single large canvas. Insurers such as Chubb require formal appraisals for works above $500,000 and recommend refreshing them every three to five years, at a cost of several thousand dollars each time. Conservation is irregular but can be large; a single relining or cleaning of a major painting costs tens of thousands.
Selling. Seller’s commissions are negotiable and, for major consignments, often waived, but the typical private seller of a mid-value work pays 5–15% commission plus insurance, photography, catalogue and sometimes a withdrawal fee. If the work fails to sell, it is “burned”: Apollo magazine cites the standard finding that works bought in at auction lose an average of 32% of their value on resale. The alternative, a private sale through a dealer, typically costs 10–20% of the price. Taxes come on top; in the US, gains on art held over a year are taxed as collectibles at up to 28% federal rate, higher than the 20% rate on shares.
| Cost item | Typical range | On a $1m work | Source / note |
|---|---|---|---|
| Buyer’s premium (auction) | 15–29% of hammer | $270,000–290,000 | Sotheby’s 28% / Christie’s 27% / Phillips 29% on first tier, 2025–26 schedules |
| Import VAT (if cross-border) | 5–5.5% (UK, France, Italy) | $50,000–55,000 | Reduced rates for art; US has no federal import VAT |
| Insurance | 0.1–2% of value a year | $1,000–20,000 a year | Chubb, AXA XL, Hiscox guidance; under 0.5% for large secured collections |
| Storage and handling | $2,000–15,000 a year | $5,000 (typical) | Climate-controlled specialist storage; higher in freeports |
| Appraisal | Every 3–5 years | $2,000–5,000 each | Required by insurers above $500,000 (Chubb) and by lenders annually |
| Seller’s commission and fees | 0–20% of hammer | $50,000–150,000 | Waived for major consignments; dealers charge 10–20% on private sales |
| Capital gains tax (US) | Up to 28% federal | Depends on gain | Collectibles rate exceeds the 20% top rate on shares |
How liquid is art, and what happens in a forced sale?
Art is one of the least liquid asset classes in common use. Roughly 32% of fine art lots offered at auction in 2025 went unsold, the sale calendar allows two or three serious opportunities a year per category, and a forced sale typically realises 20–30% below appraised fair value, according to lender guidance summarised by Deloitte and specialist lenders.
The stock-to-flow arithmetic explains the problem. Deloitte estimates $2.56 trillion of art and collectibles in ultra-high-net-worth hands; the whole market turns over $59.6 billion a year, and the fine art auction market only $11.1 billion. Even a modest wave of estate sales can overwhelm demand in a given category. The Art Basel & UBS report notes that the average dealer served only 57 unique buyers in 2025, the fewest since 2021. For a specific artist, the number of serious bidders at any price level above $1 million may be in single digits.
Timing is the second constraint. Consigning to a major evening sale means committing four to six months ahead, agreeing an estimate, and accepting that if the work fails the public record will show it. Masterworks’ filings show the time from acquisition to sale for its exited works ranging from 2.7 to 58.5 months, median about two years, and its circular tells investors to expect holds of three to ten years. Sotheby’s Financial Services says a loan secured on art can close in about 30 days, which is why borrowing against art, covered in our guide to art-secured lending, has become the standard way to raise cash without selling.
Secondary markets for fractional shares were supposed to solve this and have not. Masterworks’ share trading moved to an alternative trading system in January 2025, and in October 2026 that arrangement ended early with trading migrating to an internal bulletin board that does not match orders; the company’s own filings describe expected volume as “extremely low”. Liquidity in art is a function of the physical object and the handful of people who want it, not of the wrapper.
Which segments of the art market have behaved differently?
Returns vary enormously by segment, period and artist, and the segment that is fashionable when an investor buys is usually the one that disappoints. In 2025 Impressionist and Old Master values rose 47% and 30% at auction while contemporary was flat and mid-career “ultra-contemporary” names were still trading 20–40% below their 2021–2022 peaks.
The ultra-contemporary boom of 2021–2022 is the clearest recent example. Works by artists under 40 were flipped at auction within months of purchase for multiples of their gallery price; Masterworks’ two best exits, at 788.9% and 351.8% annualised, were short holds of a Cecily Brown and a Joan Mitchell resold within months, which shows how a quick flip inflates an annualised figure. By 2024 the same names were being bought in or sold at steep discounts, and the Art Basel & UBS report records contemporary dealers as stagnant in 2025 while modern dealers grew 11% and Old Master dealers 9%. Deloitte’s survey found only 38% of collectors and 27% of wealth managers reported high trust in art market data, which is partly a reaction to that cycle.
Blue-chip names with deep markets (Picasso, with 3,729 auction results in 2025 according to Artprice; Warhol; Richter; Basquiat) offer the most reliable pricing but not the best returns: Masterworks’ Warhol, Basquiat, Hockney and Richter exits came in at 4.1%, 6.3%, 6.0% and 6.6% a year respectively. The trade-off is familiar from every asset class: liquidity costs return. Deloitte’s figure that the global mid-market ($50,000 to $1 million) was about $8 billion of auction sales in 2024, only 4% of lots sold, shows how thin the pricing is between the mass market and the trophies.
Price point matters as much as period. Artprice reports 80% of lots in 2025 sold for under $3,160; that market is enormous in number, driven by decorative demand and younger online buyers, and nearly useless as an investment because transaction costs exceed any plausible appreciation. The investable segment, works above roughly $250,000 by artists with institutional recognition and an established auction record, is small, expensive to enter and dominated by a few hundred collectors and dealers.
How do wealth managers and family offices actually treat art?
Most now treat art as a wealth-planning problem rather than an investment product. Deloitte’s 2025 survey found the share of wealth managers offering art-related services fell from 63% in 2023 to 51%, while the share saying clients ask for help with art rose from 44% to 65%. The emphasis has shifted to estate planning, lending and collection management.
The numbers behind the shift are generational. Deloitte expects up to $992 billion of art and collectibles (roughly $100 billion a year) to pass between generations over the next decade, on the assumption that 5% of a projected $31 trillion wealth transfer is held in collectibles. Family offices reflect this: 67% now cite estate planning as a focus for their art holdings, up from 41% in 2023, and 51% cite philanthropy, up from 23%. Family offices surveyed allocate about 13.4% of wealth to art and collectibles against 8.6% for private-bank clients. The service demanded is custody, valuation, succession and tax structuring rather than a return target.
Art-secured lending is the one financial product that has grown consistently. Deloitte puts outstanding art loans at $33.9–40 billion in 2025, up from about $8 billion in 2015, and projects $42–50 billion by 2027. Bank of America’s art-lending commitments grew 14% year on year in the first half of 2025. Tellingly, 73% of wealth managers offering such loans say clients use the proceeds to fund other businesses and investments, not to buy more art. The collection has become collateral for everything else.
The intellectual case has also narrowed. Deloitte reports that “art as an asset class” fell as a stated reason for offering services from 60% of wealth managers in 2023 to 52% in 2025, while demand for collection management rose from 52% to 63%. The industry has, in effect, conceded that art’s financial return is secondary and that its job is to manage the asset the client already has.
Where does art fit in a portfolio, and what is the Türkiye picture?
For investors who already own art, the sensible frame is to treat it as a non-yielding, illiquid real asset held mainly for enjoyment and legacy, to budget its costs, and to use its collateral value when liquidity is needed. For investors who do not own art, the evidence does not support buying it for financial return.
The portfolio arithmetic is unforgiving. A 3.2% long-run index return minus 1–2% a year in holding costs leaves a real return near zero before transaction costs, which on a ten-year hold amortise at a further 2–4% a year. To justify that, an investor needs either non-financial return (which is real and is the main reason the ultra-wealthy own $2.56 trillion of the stuff), or a structural edge: access to primary-market allocations at galleries, which are sold at a fraction of auction value to favoured clients, or the ability to act as a guarantor or lender rather than an owner. Our posts on auction economics and on the Investment Analysis hub cover how to evaluate those positions.
Türkiye illustrates the local version of every point above. The national market is estimated at over $200 million a year by Antik A.Ş. chairman Turgay Artam, with the record for a Turkish work still the roughly $4.5 million paid in 2016 for an Osman Hamdi Bey. Prices for modern Turkish painters such as Burhan Doğançay, Fahrelnissa Zeid and Mübin Orhon are set largely in lira at Istanbul auctions and intermittently in dollars at Sotheby’s and Christie’s, so holders carry currency and liquidity risk at once; lira depreciation has produced nominal records that are flat or negative in dollar terms. Contemporary Istanbul’s founder Ali Güreli is pushing for a cut in art VAT comparable to Italy’s move to 5%, which would reduce one layer of transaction cost; until then the round-trip cost of buying and selling in Türkiye is higher than in Paris or London, and the buyer base narrower.
Frequently Asked Questions
Does art outperform the stock market?
Not on the broad evidence. Deloitte’s Art & Finance Report 2025 cites the Artnet Fine Art Top 100 index at 3.2% a year over 20 years and −2.9% over 10 years, against 10.4% and 13.3% for the S&P 500. Narrow indices built from the best-performing artists show higher numbers but suffer from survivorship bias, and all index figures exclude transaction and holding costs that can exceed 30% of a work’s price over a holding period.
What percentage of a portfolio should be in art?
There is no evidence-based target. Deloitte’s 2025 survey found wealth-management clients hold an average of 10.4% of their wealth in art and collectibles, and family offices about 13.4%, but these reflect inherited and passion holdings rather than allocation decisions. Given costs of 1–2% a year and round-trip transaction costs of 25–45%, most advisers treat art as a lifestyle asset to be managed, not a return-seeking allocation.
Is art a hedge against inflation?
Only weakly. Art is a real asset priced in nominal dollars and has held value over very long periods, but its short-run behaviour follows interest rates and the wealth of top collectors rather than consumer prices. Global sales fell 12% in 2024 during a high-inflation, high-rate period and recovered 4% in 2025 as rates eased. Gold, inflation-linked bonds and real estate have far more reliable inflation records.
How long does it take to sell a painting?
Typically three to nine months through a major auction house, because consignment deadlines fall four to six months before the sale and the main sales occur only in May, June, October and November. Private sales through a dealer can be faster but at a 10–20% cost. About 32% of fine art lots offered at auction in 2025 found no buyer, and a failed work typically resells 32% lower later.
Discover more from Kurums | Business Intelligence
Subscribe to get the latest posts sent to your email.
