Where the +4% did not come from
−2%
Post-War & Contemporary, the market's growth engine for a decade, fell to $4.5bn. Works under $50,000 slipped.
In March 2026, the Art Basel & UBS Global Art Market Report announced a return to growth. The figure travelled fast and uncritically. Read it twice — once as reported, once as structured.
"After two years of decline, the global art market returned to growth." The headline is true. It is also the least informative sentence available about the state of the market.
A 4% gain assembled almost entirely from the top of the market and the deepest art history is not a recovering market. It is a correcting market with a loud exception — and the exception is being reported as the rule.
The market peaked at $67.8bn in 2022, fell to $65.3bn in 2023, then to $57.5bn in 2024 — a 12% drop. The 2025 figure of $59.6bn remains 12% below the peak. Two years of contraction were not reversed; they were partially, and unevenly, papered over.
The single most important fact about 2025 is not the total. It is the spread between segments. The auction market did not rise together — it split in two.
And the weakness is not cyclical. The middle of the market — works between $50,000 and $250,000, where most serious collecting actually happens — has been hollowing out for fifteen years.
Three readings confirm the rebound sits on a deteriorating base. A market in genuine recovery does not produce these numbers.
"A recovery built on a handful of exceptional lots is not a recovery in any meaningful structural sense."
— Independent analysis of the Art Basel & UBS 2026 ReportThis is the finding the Index can stand behind without qualification: the blue-chip results are not evidence against a correction. They are its signature. When uncertainty rises, capital does not leave the art market — it retreats into the narrowest band of scarce, established, museum-validated work. A rising top end and a hollow middle are the same event.
ArtTactic's Latin American Art Market Report 2025 reports record auction sales for the region. True — and, on its own, misleading. On 20 November 2025, Sotheby's sold Frida Kahlo's El sueño (La cama) for $54.7m: a record for the artist, for any woman artist, and for any Latin American artist at auction. Now remove it.
A market whose record is delivered by a single 1940 canvas has not re-rated. It has had one transaction. Strip the trophy and the picture is sober: Latin American auction volume in the most recent full year was, by Artnet's count, lower than it was a decade earlier — roughly $245m against $270m.
For the Index this is the hinge. The hollow record is not the bad news. It is the opportunity, stated precisely: the institutional case for Latin American art has moved faster than its prices. The gap between the two has a name.
The claim of a discount needs a benchmark, or it is only an opinion. The Index uses one: institutional standing measured against clearing price. Where an artist's place in the canon — permanent museum collections, retrospectives, scholarship, biennial inclusion — is settled, but the auction record is not, the gap is a discount. Latin America's foundational figures show that gap at its widest.
Read the floor of that chart. Hélio Oiticica and Lygia Clark co-founded Neo-Concretism; they hold permanent displays at Tate Modern and works in MoMA. Jesús Rafael Soto and Carlos Cruz-Diez are foundational to kinetic art worldwide. Their auction records sit in the five-to-six-figure range — the band the Art Basel/UBS data identifies as the market's weakest. Artists this thoroughly canonised do not, in a rational market, clear at mid-market prices.
"Established artists from the region remain undervalued against their international peers."
— Recurring finding across post-Biennale market analysis, 2024–2026The 2024 Venice Biennale, Foreigners Everywhere, was directed by Adriano Pedrosa — the first Latin American curator in the event's 130-year history. Of 331 artists, more than 80 held Latin American links: ~24% of the show, against ~11% in 2022.
Anna Maria Maiolino took the Golden Lion for Lifetime Achievement. Museums are actively reassessing Neo-Concretism and kinetic art; Hauser & Wirth, Lisson, and David Zwirner now carry South American abstraction. Validation is not pending. It is on the record.
Against a softening global market, dealer sales in South America rose — driven by a 21% year-on-year increase by Brazilian galleries in 2025 (Art Basel/UBS).
And the forward signal is the strongest in the world: 72% of collectors in Brazil intend to buy in the next twelve months — the highest purchase intention of any market surveyed. The Surrealist re-rating, much of it Mexico-based women artists, saw the category's global market share rise from 9.3% to 16.8% between 2018 and 2024.
The global market has spent three years teaching collectors a single lesson: in conditions of uncertainty, the only reliable store of value is art that is scarce, established, and institutionally validated. That is precisely why the top end rose while the middle fell.
Latin American art now satisfies every term of that definition — a deep, coherent twentieth-century art history; settled museum standing; an accelerating institutional cycle from Venice outward — and it is still priced in the band the market has abandoned. That is not a contradiction. It is a mispricing, and mispricings are directional.
Every figure in this instrument is drawn from a named, published source. Nothing here is modelled or proprietary. The Index's standing depends on a reader being able to check it — so this is where the checking starts.
This is an argued exhibit, not a closed proof. A global correction cannot be "proven" as a forecast — the defensible claim, and the one made here, is structural: the 2025 rebound is narrow and top-heavy, sitting on a base that is still contracting. That claim is fully sourced.
The Latin American discount is presented against an explicit benchmark — institutional standing versus clearing price — and supported with named comparables. It is a strong argument, not a single statistic. A rigorous, defensible discount figure requires the metrics beneath the headline totals: sell-through rate, median price with trophy lots stripped out, mid-tier volume, bought-in rate, and market breadth — how many artists actually transact. ArtTactic holds that segmentation. The Index does not yet. Acquiring it is the necessary next step before this exhibit's central number is published as fact.
One deliberate omission: certain regional market-share figures circulating in secondary coverage could not be reconciled across sources and were excluded rather than published unverified.