The crypto trade publication Protos ran an article with a rather alarming headline, at least for NFT investors, last week: 

"The NFT market bubble has popped and we’ve got the charts to prove it"

Anyone can claim a bubble has popped, but when the empirical evidence is trotted out, that’s when you know it’s serious. 

The thrust of the Protos piece, which relied on an analysis of data from the site NonFungible.com, is that the market in NFTs has fallen over 90% from a peak in early May. The volume of NFTs traded fell from $170 million at the peak to $19 million or so at the end of May. 

Protos’ analysis went viral. But then NonFungibles.com itself, whose data Protos used for its analysis, refuted the publication’s claim. This was followed by a further analysis by the art-world trade publisher ArtNet. Both of these responses centred on the semantics of Protos’ claims: it’s true the NFT markets have declined, they say, but it doesn’t amount to a bubble being popped. 

At the heart of the NFT bubble kerfuffle is a question of the language and data: What constitutes a bubble, and how do we know if we were even in one?

First, let’s try to understand the claims. Protos’ main argument is that the market peaked in early May and has fallen precipitously since then. However, the market peaked on that day for a reason: the Larva Labs drop of Meebits, which accounted for over $90 million in sales in a single day. This is the assertion of the Artnet analysis. 

To complicate matters, the NonFungibles data that Protos uses doesn’t include a bunch of important things, such as off-chain sales, like any auction house activity; and activity on chains like Flow, whose NBA Top Shot collectible has been doing roaring business. That means the market might look anemic when it isn’t. 

Indeed, NonFungibles themselves write that the market is still up massively since the start of the year: from about $1 million in January to $70 million at the end of May. The trend line isn’t weakening, NonFungibles claims. 

What all the parties do agree on is that sales were way up, and now are significantly down. Almost everyone also agrees this is a good thing: there was no way NFT prices could sustain their manic price trajectory. Where they disagree is whether what took place was a bubble bursting. 

At this stage, a timely intervention by the Nobel prize-winning economist Robert Schiller is illuminating. Schiller has written a lot about bubbles, and it turns out, has an opinion on whether NFTs were in a bubble. His take? He told Bloomberg’s John Authers that they are probably not a bubble, and that bubble is probably the wrong term for lots of these frothy markets anyway. He prefers to call it a “fad” or rather more ominously, an “epidemic.” He told Bloomberg, “The narrative emphasizes the all-or-nothing crash. We learned it wasn’t the case with Bitcoin, it’s on its third-wave, like disease epidemics that go in waves.”

So what’s in a name? Both Schiller and his interlocutor Authers say that bubbles pop and they disappear—it was supported by thin air. But fads or viruses can mutate and come back in slightly modified forms, often in waves. This describes bitcoin and NFTs but also gold, Schiller says. 

Perhaps more importantly, fads and waves of viral adoption often accompany the introduction of genuinely transformative technologies: railways, canals and the automobile, Authers says. It was simply difficult to judge the value of these innovations, thus prices could swing between extremes. 

“There is a burgeoning literature in defending bubbles as a by-product of healthy excitement over investing in new technologies—but it’s impossible to support a true “bubble” like the South Sea Bubble or tulipmania. Defending investment fads or epidemics makes far more sense,” Authers writes.  

Okay, so maybe the NFT “fad” has died down a little, but by how much? And why? For the answers I turned to an analyst at research firm Messari who follows the market closely, Mason Nystrom. He says the NFT market is still growing, but at a slower pace. The collectibles and cryptoart segments in particular have been hit hard—even stalwarts like CryptoPunks are down 50% over the last month. 

Major NFT marketplaces like OpenSea and Rarible are still registering some of their best months for sales volume ever. High-quality projects like Axie Infinity, a game involving cute creatures that battle one another, are seeing growth rates of 400% and tens of millions of dollars in volume.  

“Overall, the NFT market experienced a healthy correction that will result in projects with solid business models and communities gaining traction over projects solely relying on sentiment and speculation,” Nystrom says. 

So a bubble bursting it was not. But are market-watchers to make accurate assessments? It’s tough, Nystrom concedes. He points to the challenging data environment for NFTs: It’s fragmented and difficult to parse. NonFungible tracks a limited number of projects; you have to scour a plethora of project-specific sites to manually aggregate data for a fuller picture. “NFT data is sparse compared to DeFi data,” Nystrom tells me. 

The fragmented data means that drawing inaccurate conclusions is easy to do. The plethora of categories and disparate data sources means that it’s easy to be alarmed by a sharply declining chart. And there is peril in examining data trends in a quickly emerging category, since many of these boundaries are still sorting themselves out. After all, “NFTs are simply a file format that can be applied broadly to a variety of assets,” Nystrom points out.