In April 2005, at an estate sale in New Orleans, two New York dealers, Alexander Parish and Robert Simon, paid $1,175 for a badly damaged walnut panel, covered in overpaint, catalogued as "after Leonardo da Vinci".
Twelve years later, the same panel reached $450 million. In between, it had been cleaned, restored, and above all renamed.
on 15 November 2017 at Christie's, New York, after roughly nineteen minutes of bidding; $400m at the hammer, the rest in fees
The painting was not offered in an Old Masters sale, but in a post-war and contemporary evening auction, facing Warhol's Sixty Last Suppers. The price does not pay for a school: it pays for a name.
In 2013, the painting moved from $80m to $127.5m within twenty-four hours, between a middleman and his client; the affair ended in courtrooms on several continents. The 2017 price belongs to that sequence of gaps, not to an established market level.
And the buyer was never announced by the auction house. The value of a unique object is never more than an opinion, held by two people on the same evening.
The composition is not rare: around thirty painted versions by the workshop and by followers are known, along with two preparatory drawings by Leonardo held at Windsor. What is rare is the hand.
Fewer than twenty paintings are universally accepted as Leonardo's. Rarity does not rest on the subject: it rests on the author, and the author here is a hypothesis.
From almost the whole panel for its defenders down to five or twenty per cent for Matthew Landrus, who gives the rest to Bernardino Luini: three different paintings, one single price.
A Salvator Mundi appears in the inventory of the goods of Charles I, drawn up in 1649 after his execution, and again under Charles II. The trail is lost after a sale of 1763, at Buckingham House, for nearly a century and a half. The panel resurfaces around 1900 in the Cook collection, under the name of Bernardino Luini, and sells for £45 at Sotheby's in 1958, catalogued as Boltraffio.
Nothing in that chain links the 2005 panel with certainty to the one of 1649. The trajectory is plausible: it is not continuous.
2005: an overpainted panel, catalogued as "after Leonardo", bought for $1,175 at an estate sale.
Christie's was not selling a panel: the house was selling the last Leonardo in private hands. A world tour through Hong Kong, San Francisco and London, a film of visitors overcome in front of the painting, and a place in the most spectacular sale of the year, far from the Old Masters aisle.
The story was flawless, and it bore precisely on what nobody could verify. Narrative coherence did not accompany the price: it manufactured it.
No document from Leonardo's lifetime mentions a Salvator Mundi painted by his hand. The 1649 inventory attests a painting of that subject in the royal collection, not the identity of this particular panel. Examined in 2018 by the French national museums laboratory, it was ultimately not exhibited in 2019, and the volume recording those conclusions was never published.
Proven origin is not evidence here: it is a hypothesis that the price turned into a certainty.
$1,175 in 2005, $450.3m in 2017: the painted surface did not change in nature, only the standing of its attribution changed. Three axes of the grid are weak here: rarity rests entirely on the author, transmission is broken, origin is undocumented. One alone is very strong, the story. It was enough.
Since the sale, the painting has not been shown in public; in 2024, the press placed it in a Geneva warehouse. A Stradivarius can be verified by playing it; the Salvator Mundi can no longer be verified at all.