I’ve spent the last few months working with Josh Stark at the Ethereum Foundation, specifically focusing on supporting projects like The Year in Ethereum 2021 and a handful of other smaller research pieces.
One of the projects I’ve been tinkering with recently is figuring out how to calculate the total market capitalization of assets on Ethereum, or in other words, the Ethereum ecosystem’s total market cap. This is the total value of all assets on Ethereum including ETH, ERC-20s, NFTs, and synthetic/derivative assets.
This is a relevant and useful metric. The entire crypto ecosystem is made up of blockchains, and these blockchains enable the creation of new types of assets, as well as new methods of utilizing and trading the assets. One of the most obvious questions to ask about crypto is “what is the value of all of the assets on Blockchain X?”, and for a huge blockchain ecosystem like Ethereum’s, the only sensible way to answer that question is to include the value of all assets that live on Ethereum, and not just the value of the blockchain’s native token (ETH, in this case).
Despite the various data providers and analytics solutions available for Ethereum, a single source of information for this data isn’t very easy to find. Most publicly available estimates for “total market cap of Ethereum assets” usually just refer to the value of ETH or just to the sum value of all ERC-20s. But with the rise of NFTs, most of which are transacted on Ethereum, and the large amounts of other blockchain’s native tokens that live on Ethereum as wrapped tokens, there’s a larger scope of assets that could be included in the calculation and compiled into one simple number.
Why is this a difficult problem? For one, the various data providers sometimes have different data about the same thing. For example, here are two different NFT data feeds - Crypto.Com and NFTGo - with significantly different values for the market cap of different NFT collections. On top of that, some metrics provided by sites like CoinMarketCap don’t correspond to the data that I am intending to capture in this analysis. USDT is a good example here; CoinMarketCap lists USDT’s market cap at ~$80B, but that number includes the token contract deployed on non-Ethereum chains like Solana and Avalanche. EtherScan shows that the USDT balance *just on Ethereum* is closer to $40B.
With all of this in mind, I set out to make a rough v1 dashboard for the total Ethereum ecosystem market capitalization calculation. My hope is that by poking at the problem and then open-sourcing my work, I might inspire someone to build a dynamic dashboard with permanence. If you’re interested in doing this, DM me!
I broke down the Ethereum-based assets that I wanted to include in this calculation into four major categories; ETH, ERC-20s, ‘synthetic assets’ and NFTs. The market cap of Ether alone is simple; $310 billion as of January 26, 2022.
For ERC-20s, I used CoinMarketCap to check which blockchains individual tokens were deployed to, and then I verified the reported market cap number versus Etherscan contracts. In situations where the token contract is deployed on several chains, like Tether described above, I only used the supply of tokens that was deployed specifically to Ethereum. Of course, this was a tedious task for v1, but going through the motions here helped me think through the nuances that need to be considered when building out a more dynamic dashboard.
In total, the ERC-20 calculation added an additional $240 billion in Ethereum ecosystem market capitalization. While ETH itself is not an ERC-20 token, wrapped ETH is an ERC-20-conforming token, so it is included in the calculation with other ERC-20s. I capped the analysis at ERC-20s with a total asset market cap of $300M or more as of 1/26/2022, but as a rough estimate, I believe that the ‘long-tail’ of ERC-20s adds another $20-40 billion in market cap. This is not included in this calculation, in an effort to conservatively underestimate the true number.
I also wanted to capture the value of what I’m calling ‘synthetic assets’; basically, derivatives of cryptoassets that are traded in separate Ethereum-based liquid markets from the original asset.
The fact that many assets are now cross-chain introduces additional complexity. Some assets on Ethereum were “imported” from another blockchain. This means that they were created on a different chain, but then migrated over and “wrapped” inside an Ethereum ERC-20, which serves as an “IOU” that can be redeemed on the original chain. For example, wBTC is “wrapped Bitcoin” which can be used like an Ethereum token. We’ll call these “inbound wrapped assets”. There are also Ethereum-native assets that are migrated over to other chains, where they are ‘wrapped’ into another blockchain’s token format. This can be referred to as an ‘outbound’ wrapped asset.
An example of an outbound wrapped token is surprisingly ENS, the governance token for the Ethereum Name Service protocol, which would naturally seem like an Ethereum-only asset. However, the ENS token’s CoinMarketCap page shows that the ENS contract has also been deployed to Solana. If the Solana contract is opened, it’s clear that this “Solana-deployed ENS” is just a ‘wrapped’ version of ENS that has been sent over to Solana. Counting this Solana-wrapped ENS in this calculation would double count the true value of ENS, since the Ethereum-based ENS market cap already includes the Ethereum-native ENS tokens that remain on Ethereum, where they can be “redeemed” using the Solana-based IOU. These situations are not counted in my calculation to avoid double-counting assets.
Some DeFi positions are considered synthetic assets for the sake of this calculation. Uniswap v3 turned liquidity provider positions into NFTs, and other derivatives like stETH and xSUSHI act similarly - a financial instrument is created using another asset, and then the created financial instrument is traded in its own liquid market and utilized in its own ways, separately from the initial asset used to create the instrument. A reasonable argument against including these types of synthetic assets is that inclusion would seem to double count some values of assets on Ethereum, like the staked ETH in a staking pool. However, my thought process is that these derivatives have their own value and their own risk profiles (for example, consider the risks of holding stETH vs the risks of holding just ETH), so I chose to count them in this calculation.
I included all synthetic assets with a total asset market cap of $300M or more as of 1/26/2022. Synthetic assets accounted for roughly $25 billion in assets that live on Ethereum.
Finally, NFTs. Like I mentioned in the introduction, finding the right NFT data is difficult. The biggest reason for this difficulty is the fact that most markets for NFTs are infrequently traded and illiquid, which makes calculating a market cap figure difficult. On top of that, the methodology used by two different firms can be significantly different, while both remaining completely reasonable. I used NFTGo’s data feed because I felt that their methodology was the most transparent and reasonable of the NFT data providers. For example, NFTGo counts NFTs that have previously traded at $X but are currently listed at a higher price $Y at the value of their last trade ($X) rather than at the inflated listed price ($Y).
NFTs added another $9B to the total market capitalization figure, which might be surprising given their outsized influence in ‘web3 culture’ relative to their cumulative market capitalization. As the NFT digital art and collectibles category continues to grow and evolve, I’m confident that this data will become more standardized and accessible, and I’m hopeful that NFTGo’s soon-to-be-released API will make it easy to pull this data into a dynamic dashboard.
So that’s it! During a static moment in time during the last week of January 2022, the Ethereum ecosystem’s total market cap was roughly $585 billion. This “static moment in time” approach is crude, but it serves a few purposes, like helping me think through the process, surfacing potential issues to solve in a future iteration of the dashboard, and providing a useful ballpark estimate for the true figure.
Hopefully, this project inspires people to experiment with creating dynamic versions of this dashboard or approaching the problem from a different angle. Especially if 2022 becomes the year of the L2, the inclusion of Layer 2 assets into this calculation will add a whole new dimension of complexity to the problem. If you’re interested in building something like this out or if you have any feedback, please reach out via Twitter or e-mail!
Calculation and sources are documented here.