The Art Market Often Works in Secret. Here’s a Look Inside.
Ever since a computer file made by the digital artist known as Beeple sold at auction in March for $69 million, observers of the art world have been fascinated and bewildered by the astronomical spike in prices for this type of work — so-called NFT-based art. These are digital creations that, because they are otherwise easily susceptible to being copied and reproduced, are sold as unique assets in the form of nonfungible tokens, or NFTs, which use blockchain technology to certify authenticity and proof of ownership. (Beeple’s piece was a collage of images that he had posted online every day since 2007.)
The market for this art has grown drastically. From April 5, 2018, through April 15 of this year, 6,158 artists sold 191,208 pieces of NFT-based art for a total of $541,378,383, according to Crypto Art, a website that tracks such sales. Roughly half of these transactions took place this past March, giving rise to one of the greatest and most sudden asset booms in history.
As an art collector, I enjoy coming across an intriguing piece of art and asking myself how much I might pay to own it. But as a network scientist, when I encounter a complex phenomenon like the art market, I am inclined to examine its hidden structure, drawing from multiple disciplines (physics, sociology, computer science) to reveal the unseen patterns of relationships that can help explain how it works and why.
From the moment I learned about the world of NFT-based artwork, I’ve been busy doing what I do best: mapping — that is, analyzing and visually representing — the patterns of ownership transactions that underlie the genre’s meteoric rise. My maps show that the market for NFT-based art is extremely insular and tightly connected, even by the standards of the art world, especially among owners who buy and sell several times. These features of the network may help explain the enormous spikes in sales prices for NFT-based art.
An NFT is a permanent and reliable public online record of ownership that can be connected to any asset. (This year the chief executive of Twitter, Jack Dorsey, sold his first-ever tweet as an NFT for $2.9 million.) In the case of NFT-based art, the NFT includes information about the “primary” market — the creator, the first collector and the sale price — as well as a record of the “secondary” market, subsequent changes in ownership and valuation over time. In the traditional art world, this kind of information is often veiled in secrecy.
It will take some time to map the entire set of ownership transactions in which an NFT-based artwork was involved, even for my lab’s cutting-edge computers. But working with the data scientist Milan Janosov, I have already discerned some interesting patterns.
We started our analysis by looking at a website called SuperRare, one of the earliest and most prominent platforms for buying and selling single-edition digital art. Using specialized algorithms, our computers traced every transaction on SuperRare in which an NFT was involved. The story began on April 5, 2018, with the posting of an artificial-intelligence-generated nude portrait by Robbie Barrat, which was purchased soon thereafter for $176 by a collector named Jason Bailey. (Mr. Barrat’s artwork resold for $112,717 on Jan. 5 this year.) By April 15 of this year, according to our analysis, 16,198 works created by 887 artists had changed ownership on SuperRare, involving 3,210 collectors and more than 23,000 transactions.
As in the traditional art market, a majority of these collectors “buy and hold” — here, the figure is over 60 percent — meaning that the digital art they purchase does not re-enter the market. But as in the traditional art market, there is also a lively secondary market for NFTs. In March 2020, the secondary-market activity accounted for 9 percent of sales on SuperRare. By March of this year, the secondary market was booming: Resales accounted for 36 percent of the art sold on the platform.
Next we had to decide which aspects of the NFT-based art market we wanted to map. We chose to look at patterns of co-ownership: We would plot every work of art as a node in the network, and we would link two artworks if they were at any time during their existence owned by the same collector.
Why focus on co-ownership? Our reasoning was that art collectors tend to specialize in certain segments of the market — a particular group of artists (say, the Hudson River School) or an artistic movement (like Impressionism), genre (like still life) or medium (like sculpture). As a result, patterns of co-ownership, we suspected, would reflect meaningful commonalities among the artworks.
PUBLISHED BY– nytimes