NFT index
Apple proclaimed “there is an app for that” in a commercial for the iPhone. At Vinter, whenever someone talks about a trend in the market, we say “there is an index for that.” But can we build an NFT index? At first, it seems like a stupid question, however, stupid questions are sometimes interesting to linger on. In contrast to stocks and bitcoins, you can’t buy $10 more of NFT Alpha if its price is $100. Because NFTs are not fungible. Vinter’s indexes are primarily used for creating exchange-traded products, but I don’t think creating an exchange-traded product on NFTs is feasible. So what would the use case be?
Perhaps a simple index for informational purposes could be created by taking the ten largest NFTs and measuring their value. Even though the construction is simple, two questions arise: why ten and what is their value?
Firstly, ten was my first instinct because it sounds good. Index providers love numbers like 10, 100, 500, 1000. My education in statistics tells me it would be better to follow the elbow principle. It is used when selecting the number of components in a principal components analysis. The elbow principle says that we should keep adding items until there is a sharp decrease in the gain of adding the next component. Said differently, we stop adding items whenever the marginal utility of adding an extra item is too low. For example, if five NFTs are worth $100 $90 $80 $70 $10 we would only keep the first four because there is a sharp decline from $70 to $10. Viewed differently, adding the fifth is not useful because going from $350 to $360 is a 2.8% increase which is considered too small to add another item. In the elbow method, we don’t need to set a fixed number like 10. Instead, we “merely” need to agree on a threshold (like 3% or 5% or 10%) of value increase to the cumulative sum for the item to be included.
Secondly, what is their value? Is it the last price someone bought it at? If NFT Alpha and Beta were bought in January for $100, and later on NFT Alpha is sold for $200 does that mean we could infer the price of NFT Beta has increased? Probably not, as NFTs (supposedly) are works of art so the value increase in one could mean that the artist is hyped and will not have spillover effects on other NFTs.
Valuation is tricky. What happens when the market tanks? Let’s assume the prices of the five largest ones are $100 $90 $80 $70 $10 The owners of Beta, Gamma, Delta, Epsilon and Zeta all panic and sell their NFTs at a 50% loss. Alpha’s owner, however, is holding on to his NFT for dear life. Alpha was worth $100 but was never sold in the 50% crash — what is the value of Alpha now? It could be $100 or $50 or something else — who knows? Without a recent transaction, it’s almost impossible to know. Unless you’re an NFT art dealer valuation expert with a PhD in art, economics and philosophy. I searched on LinkedIn but didn’t find any.
Prices reflect information. Prices contain information. If there are no recent transactions, the information in the price is less valuable than if it was done recently. Perhaps our NFT index needs to require the NFTs to have been sold recently, for example, in the last quarter or the past month. Whether we choose a month or a quarter will likely have a big impact on the index.
An NFT can be fractionalized — split into a thousand pieces. This brings the non-fungible token slightly closer to the characteristics on fungible tokens. Could we build an index that only selects fractionalized NFTs? I don’t think so but I could be wrong. I’m just thinking out loud. This NFT index rant must stop somewhere. It ends here. For now.
This post originally appeared on the Vinter monthly industry insight newsletter.
