Commerce has played a significant role throughout human history and has continually developed and evolved to reflect the changing needs and capabilities of society. The practice of trade, initially shaped by bartering, was the bedrock upon which early civilizations thrived. This ancient system involved the direct exchange of goods or services, dependent entirely on the relative needs or surpluses of individual parties. The intricacy of this system led to the emergence of a more efficient approach: currency.
With the invention of currency, trade transcended the inherent constraints of bartering, establishing a universal store of value that could be accumulated and/or exchanged. Coins, gold standards, and paper money each marked a significant evolution in the architecture of commerce.
In the 2010s, the digital form of commerce, e-commerce, experienced immense growth and adoption. Transactions transcended physical boundaries and enabled instantaneous trade all around the world. The likes of Amazon, eBay, and Alibaba dominated and proved to everyone the power of a marketplace with no geographical limitations, where a product could be sold from one corner of the globe to another with the click of a button.
Today, we stand at a crossroads in the NFT world. We have the Amazon-esque marketplaces, but there hasn’t been enough innovation in the approach to pricing NFTs. As we venture into this new frontier of commerce with blockchain and cryptocurrency, we should explore novel mechanisms for pricing, particularly ones that were not possible with past systems but are possible today.
Over the past couple of years, we’ve witnessed the development of a few different mechanisms by which an NFT can be sold. The following text revisits some of these systems in an attempt to understand where they each thrive and where they fall short.
For this mechanism, it is up to the seller to take their best guess at one price when it comes to listing their NFT(s) on the market. Listing price is usually influenced by the current floor price of a collection, which is the lowest priced item in a collection at a certain point in time. The floor price serves as a reference point: in liquid collections, one can be fairly assured an NFT priced at or near the floor will sell. However, for almost every other NFT, there are no such reference points. Rare NFTs or 1/1 pieces are meant to sell for a premium, not the floor price. Certain metadata traits of NFTs within a collection may have their own floor price, usually higher than the collection’s floor price, but it is extremely unreliable and fluctuates even more drastically than the collection’s floor price. Without a way to determine fair prices, the market is extremely inefficient: there is no price discovery.
A few months into the popularization of the Amazon-esque storefront experience, NFT marketplaces began adding additional functionality, like bidding. This is where a buyer offers what they would be willing to pay for a specific NFT, whether it is listed or not listed on the market. The seller is notified of the offer and can then choose whether they want to accept it or ignore it. This unlocked more liquidity for NFTs, especially non-floor ones.
This is the most well-known type of auction. It is also sometimes referred to as an open ascending auction, where potential buyers bid on a certain NFT up until a set expiration date. Once the expiration date is reached, the highest bid wins the auction, and the NFT is transferred to the buyer.
Another innovative approach to selling an NFT is putting it up for raffle. This mechanism unlocked a level of price discovery that did not exist previously. Because the barrier to entry for a raffle is naturally a lot lower than a collection’s floor price, raffles for desirable NFTs usually generate more in sales than a traditional listing or an English auction. Suppose one were to put up a CryptoPunk with a floor price of 50 ETH for raffle, with an unlimited number of raffle tickets priced at 0.1 ETH each. This CryptoPunk raffle Is likely to generate more than 50 ETH in raffle ticket sales, simply because there are more people willing to risk 0.1 ETH for a chance to win a CryptoPunk than there are people who have 50 ETH to buy a CryptoPunk up front.
NFT fractionalization permits fractional ownership of NFTs. Essentially, an NFT that is listed on a fractionalization protocol is broken up into a set number of fractions. From there, buyers can acquire these fractions to get a proportional number of rights to govern the original listed NFT. Let’s go back to our previous example of the CryptoPunk with a floor of 50 ETH. Suppose it was fractionalized into 100 fractions, with each costing 0.5 ETH. Fraction owners get a say in the governance of the original CryptoPunk. The more fractions one holds, the more weight their voice holds in governance. This model has proven to be inefficient for listing and/or selling NFTs since the listed NFT ends up being locked in the protocol after fractions are sold. This problem arises because governance consensus among the fraction holders is required to decide what to do next with the fractionalized NFT (transfer, trade, sell… etc), which introduces a lot of friction to the process.
This is a mechanism most often used in primary NFT sales and much less frequently in secondary sales. Teams that hold a mint for their collection will typically structure their mint with a gradual decreasing Dutch auction. The mint will start off with a relatively high price (e.g. 3 ETH) and gradually decrease by a set amount (e.g. 0.1 ETH) every set number of minutes (or hours) until the collection mints out entirely. Dutch auctions guarantee that collections sell out entirely since the price eventually gets low enough to the point where buyers mint it out.
This model is now over a year old. It takes away the non-fungible nature of NFTs and treats them as equals. Within one NFT AMM pool, NFTs can either all be listed at the same price or the seller can adjust the price curve by which the price of the NFTs within his/her pool increases by a specific amount every time one of these NFTs sells. The innovation in NFT AMMs is the fact they allowed users to earn trading fees from setting up their own pools.
After carefully studying the models above, we came to the conclusion that there is no perfect pricing mechanism currently present for NFTs. Existing models have their strengths but also have their pitfalls. This prompted us to create Waterfall, a paradigm-shifting trading and pricing protocol for NFTs. We combined all the good parts of present mechanisms to come up with a new, robust protocol capable of pricing any NFT, even 1/1 pieces with no prior activity.
Here’s how it works:
An NFT is listed on Waterfall
The NFT is fractionalized into a set number of NFT shards (ERC-1155 tokens)
Shard traders can step in to make predictions on the future value of the NFT by trading shards
A buyer can step in at any time to acquire the listed NFT by paying the current price, which is the sum of prices of all shards
The Waterfall protocol enforces all shards to be listed at all times, thereby solving the issue that plagued legacy fractionalization protocols (where a listed NFT would get locked in the protocol because It is extremely difficult to achieve governance consensus)
Every shard has two pieces of financial metadata attached to it: the predict price & the expiration date. A shard trader is immediately prompted to input both as soon as they acquire their shard(s)
If a shard reaches its set expiration date, it enters a linearly decreasing Dutch auction until a buyer is found
The protocol imposes a fee based on the parameters set for shard relist price & expiration date. The higher a relist price and the further into the future an expiration date are set, the higher the fee. Protocol fees are split between the corresponding shard seller and Waterfall.