The art market is always conducted in secrecy: Why do people say that NFTs will reshape the way art is traded and collected?

This article is from Pangcheku, original title: "The Art Market Always Operates in Secret," author: Albert-Laszlo Barabasi, translation: Jessie.

This is an on-chain data visualization case that has excited me for a long time. The author conducted a visual analysis of 16,198 NFT works created by 887 artists on SuperRare, involving 3,210 collectors and over 23,000 transactions. The following conclusions were drawn:

• Over 60% of NFT collectors will "buy and hold";

• Any two of the 16,000 artworks in this network (excluding 122 exceptions) are not connected through the legendary six degrees of separation, but rather linked by a chain of no fewer than three collectors;

• These communities (clusters of large nodes in the network) represent collectors.

…

However, the excitement goes far beyond the conclusions. "As a network scientist, when I encounter complex phenomena like the art market, I tend to study its hidden structure, uncovering the invisible relationship networks. At the same time, it forms a natural artistic form."

Mark Pernice

Since the digital artist known as Beeple sold his NFT at auction for $69 million in March, observers in the art world have been fascinated and puzzled by the skyrocketing prices of this type of work—the so-called NFT-based art. These are digital creations that are easily replicated and reproduced, sold in the form of non-fungible tokens or NFTs, which use blockchain technology to prove authenticity and ownership. (Beeple's work consists of image collages he has posted online daily since 2007).

The market for this art has grown dramatically. According to Crypto Art[2], from April 5, 2018, to April 15 of this year, 6,158 artists sold 191,208 pieces of NFT-based art, totaling $541,378,383. About half of these transactions occurred in the past three months, creating one of the greatest and most sudden asset booms in history.

As an art collector, whenever I encounter an artwork that interests me, I ask myself how much I might be willing to spend to own it. But as a network scientist, when I encounter complex phenomena like the art market, I tend to study its hidden structure [3], drawing insights from multiple disciplines (physics, sociology, computer science) to uncover the invisible relationship networks that help explain how it works and why it works that way.

From the moment I learned about the world of NFT-based art, I have been busy doing what I do best: mapping—analyzing and presenting the patterns of ownership transactions, which is also the foundation of the rapid rise of this genre. The ultimate conclusion is that, among those owners who participate in buying and selling multiple times, the NFT-based art market is both isolated and tightly connected. These characteristics of the network may help explain how the peaks in prices for NFT-based artworks arise.

NFTs are a permanent, reliable public record of ownership on-chain that can be linked to any asset. This year, Twitter CEO Jack Dorsey sold his first-ever tweet as an NFT for $2.9 million. In the case of NFT-based artworks, NFTs include information about the "primary" market—the creator, the first collector, and the sale price—as well as records of the "secondary" market, showing changes in ownership and valuation over time. In the traditional art world, this information is often shrouded in secrecy.

Mapping the entire set of ownership transactions involved in NFT-based artworks takes some time, even for the cutting-edge computers in my lab. But in collaboration with data scientist Milan Janosov, I have discovered some interesting patterns.

Our analysis began with a website called SuperRare, one of the earliest and most famous platforms for buying and selling NFT digital art. Using specialized algorithms, we tracked every transaction involving NFTs on SuperRare. The story began on April 5, 2018, when Robbie Barrat posted an AI-generated nude portrait, shortly after which a collector named Jason Bailey purchased it for $176. According to our analysis, by April 15 of this year, 16,198 works created by 887 artists had transferred ownership on the SuperRare website, involving 3,210 collectors and over 23,000 transactions.

60% of Collectors "Buy and Hold"

Like the traditional art market, most of these collectors are "buy and hold," with this number exceeding 60%, meaning that the digital art they purchase will not re-enter the market. But like the traditional art market, NFTs also have an active secondary market. In March 2020, secondary market activity accounted for 9% of sales on the SuperRare website. By March of this year, the secondary market was thriving, with resales accounting for 36% of the artworks sold on the platform.

Next, we needed to decide which aspects of the NFT-based art market we wanted to map. We chose to study shared ownership patterns, mapping each artwork as a node in the network, and if two artworks were owned by the same collector at any point during their existence, we would connect them.

Why focus on shared ownership? Our reasoning is that art collectors often concentrate on certain segments of the market—a specific group of artists (e.g., the Hudson River School) or an art movement (like Impressionism), genre (like still life), or medium (like sculpture). Therefore, we suspected that the patterns of shared ownership would reflect meaningful commonalities between artworks.

Our next step was to lay out the network in two-dimensional physical space so that we could best observe its structure. To do this, we simulated the network on a computer to minimize the total length of links between nodes. This ensured that we could clearly see the patterns of ownership. Artworks owned by a single collector must be closely arranged (because there are many links between them), while artworks connected by several "degrees of separation" (with no two common owners but still linked through a series of owners) must be spaced further apart. Those artworks not connected to others by any ownership chain will float freely as isolated nodes or isolated groups.

Our intuition was that this network would show itself to be divided into isolated clusters, each reflecting collectors' specialized interests in certain types of artists and artworks. But the network we obtained broke our expectations, revealing a central, highly interconnected structure—this structure contains all but 122 of the over 16,000 artworks. **To give you a sense of how niche this market is, *any two artworks in this network (excluding the 122 exceptions) are not connected through the legendary six degrees of separation, but rather linked by a chain of no fewer than three collectors*.

This is what the network looks like.

Source: foundation.app/barabasilab

These Community Nodes Represent Collectors

As you can see, this network has several large, easily identifiable "communities"—clusters of artworks that are tightly interconnected. What do these communities represent? Initially, we suspected they were collections of various subtypes of NFT-based artworks that exhibited visual or conceptual similarities, thus attracting collectors with similar tastes. To verify this hypothesis, we assigned a color to each node corresponding to the artist who created the work. However, the resulting color chaos did not confirm this hypothesis. While there seemed to be a slight preference for certain artists within these communities, this pattern appeared to be coincidental rather than explanatory.

Then we explored another hypothesis: these communities represent collectors. This time, we assigned a color to each node corresponding to the current owner of the artwork. This was our moment of enlightenment. The network map showed that each community was associated with a collector.

Source: foundation.app/barabasilab

Despite there being 3,210 collectors on SuperRare during our analysis, most artworks are held by a small number of investors. If you look at the primary market, you can clearly see this, where the largest clusters represent artworks purchased by four early investors whose Twitter handles are @momuscollection (dark blue, left), @0x123456789 (pink, above), @thedruid (light blue, upper right), and @ethsquiat (yellow, upper right). (The large orange group at the bottom has certain characteristics indicating that it may represent works owned by a user who is no longer active on SuperRare.)

Source: foundation.app/barabasilab

In the art world, it is not uncommon for a few enterprising collectors to dominate the primary market for new art. For example, in the 1870s, art dealer Paul Durand-Ruel was one of the first to recognize the potential of modern art, purchasing over 5,000 canvases, including about 1,000 by Monet, 1,500 by Renoir, and 800 by Picasso. But typically, as the status of the avant-garde rises, new collectors enter the market, and ownership diversifies.

This Diversification Has Not Yet Occurred in the NFT Art World

This diversification does not seem to have occurred in the NFT art world, at least not yet. Our network map shows that the secondary market is even more concentrated than the primary market. During our analysis, two collectors, @0x123456789 (pink) and @momuscollection (dark blue), purchased a significant number of works from early investors, establishing a dominant position in the market.

Source: foundation.app/barabasilab

Why is this network significant? First, it indicates that when it comes to fundamental economic variables—supply and demand—the supply of artworks for sale on SuperRare is tightly controlled. Combine this controlled supply with the sudden surge of interest in digital art, and you see a demand-driven market. SuperRare also has additional restrictions on supply, as it admits new artists to its platform at a slow pace.

But more importantly, the market's pure openness. Historically, the art market has thrived in opacity. Works purchased in the primary and secondary markets often disappear from public view, ultimately becoming part of private collections. Buyers often lack complete information about comparable prices, supply, and other relevant market characteristics. Those with more information—typically gallery owners and dealers—can profit as a result.

NFT-based artworks bring welcome transparency to ownership and transactions. This makes NFTs more than just a niche art market. By introducing all types of artists—traditional and digital—blockchain technology can be used to publicly authenticate their works and record their sales history, providing a level playing field for the entire art market and fundamentally reshaping the way art is traded and collected.