Unicorn Private Research
EN

Art as an asset class: what the data says about returns, correlation and art indices

Is art a good investment? The data is mixed. Repeat-sales studies find long-run art returns below US equities and above bonds, with prices that do respond to stock markets and top incomes. Correcting for the fact that only works that rose in value tend to come back to auction cut one study's average annual index return from 8.7% to 6.3% (Review of Financial Studies, 2016). The top of the market has done better than the rest since 2010, with large swings.

Key takeaways

The market in 2025

Global art sales rose 4% in 2025 to an estimated $59.6 billion, according to the Art Basel and UBS Art Market Report 2026 by Arts Economics. It was the first year of growth since 2022, and it came almost entirely from the top. Fine art lots sold above $1 million rose 21% in value; above $10 million, 30%. Below $50,000, sales fell 2%.

Even so, the $1 million-plus segment stayed below its 2019 level, and the top 50 lots of 2025 were worth about a third less than the top 50 of 2022.

The report is the reference most people quote. It is a survey and a model, built from auction data and dealer questionnaires, not a price index.

Fifteen years by price segment

The same report tracks auction sales by price band from 2010 to 2025. Only one band ended the period above where it started: works over $10 million. Every other band declined, and the steepest drop was for works between $50,000 and $250,000, down 29%.

Start the clock at the 2009 low and everything is up, but unevenly. The $10 million-plus segment gained almost 240%; the low end 8%, which is a loss after inflation.

The top also falls harder. From 2022 to 2024 the $10 million-plus segment lost 67% of its value, and sales over $1 million halved. Then 2025 brought the rebound.

Two caveats. These figures measure total sales in each band, so they move with the number of great works offered as much as with prices. And as the report itself says, the result depends on where you put the start and end points.

Repeat-sales indices: Sotheby's Mei Moses

The standard way to measure art prices is to compare the same object at two sales. If a Calder mobile sold in 1995 and again in 2025, the difference is a clean return, with no question of whether you are comparing a masterpiece with a minor work.

Jianping Mei and Michael Moses, two professors at New York University, built the most cited indices on this method. Sotheby's bought them on 26 October 2016, when the database held about 45,000 repeat sales in eight collecting categories, with roughly 4,000 objects changing hands each year (Sotheby's).

Their 2002 paper in the American Economic Review found that art beat fixed income but underperformed US stocks, and that masterpieces, the most expensive works, tended to underperform the art index. That last finding surprises people. It fits a market where the trophy buyer pays up on the night.

Selection bias: why the indices look better than reality

Paintings that have lost value tend to stay home. Owners do not consign them, so repeat-sales indices are built mostly from the works that went up.

Arthur Korteweg, Roman Kräussl and Patrick Verwijmeren measured the effect on 32,928 paintings sold more than once between 1960 and 2013. Correcting for selection cut the average annual index return from 8.7% to 6.3% and the Sharpe ratio from 0.27 to 0.11 (Review of Financial Studies, 2016). They concluded that a broad, passive portfolio of paintings is unattractive, while noting that some styles and top-selling artists may add value.

Other gaps: indices ignore the buyer's premium (28% of the hammer price up to $2 million at Sotheby's and Christie's in 2026) and the seller's commission, insurance and storage, lots that failed to sell, and every private sale. An index return is a gross number. You will earn a net one.

Correlation with stocks

Art is often sold as uncorrelated. The evidence is more careful. Using an art price index that goes back to the eighteenth century, William Goetzmann, Luc Renneboog and Christophe Spaenjers found that equity returns, both in the same year and in the year before, have a significant effect on art prices, and that art prices and top incomes move together over the long run (American Economic Review, 2011).

This makes sense. The buyers of $20 million paintings are the people whose wealth rises and falls with financial markets. When they feel rich, they bid.

What art does offer is a different timing. Prices adjust slowly, with no daily quote, and a single painting's fate depends on its own story: provenance, condition, a museum show, a fresh discovery.

What to take from the numbers

If you invest in art, the data points to a few practical habits: concentrate on works with deep international demand, hold long enough to absorb transaction costs, and judge returns net of every fee. Unicorn Art Club applies the first two by working only on 30 established masters with a planned holding period of 3 to 7 years. Unicorn Art Watch shows the recent auction history behind each name, for example artwatch.unicorn-club.com/artists/andy-warhol.

Risk

Investing involves the risk of losing capital.

Frequently asked questions

Can I buy an art index?

No. There is no fund or contract that tracks an art index the way an ETF tracks the S&P 500. Every painting is unique, so an investor owns specific works, not the average.

Why do different art indices give different results?

They use different data (auction only, or dealer surveys), different methods (repeat sales, average prices, hedonic models) and different periods. Average-price measures move with the mix of works offered, which repeat-sales indices try to avoid.

Is the Art Basel and UBS report an index of art prices?

No. It estimates the total value and volume of the market each year from auction data and surveys of dealers and collectors. A rise in total sales can come from more top works being offered, not only from higher prices.

Related guide

Sources

This page in other languages

IPO forecasts from the Unicorn Watch community

Newsletter

Research on unicorns and pre-IPO companies, by email. No spam, unsubscribe at any time.

Subscribe on Substack