Value Fragmentation

Published in

Geek Culture

6 min readNov 21, 2021

A rising tide lifts all boats.. sometimes.

Not a day goes by on Crypto Twitter where someone doesn’t complain about the incredibly high gas fees on Ethereum. It’s easy to see how someone new to the space could view this as a failure of the grand crypto vision. Spending $50 in fees to buy a $10 coffee is absurd (for multiple reasons).

However, as most people who have been closely following Ethereum will note, this has been the plan for some time now. Ethereum intentionally limits block space to remain maximally decentralized and therefore has a limited bandwidth. As its utility and demand increases so does the cost per transaction. In other words, Ethereum has been a victim of its own success. The Ethereum roadmap has aimed to address this by moving everyday transactions to L2s* and using the L1 as the base settlement layer that is mostly used by L2s and larger players.

*For simplicity I’ll use L2s and Rollups interchangeably

Unfortunately, we find ourselves in an awkward state where L2s aren’t quite ready and Ethereum’s mainnet capacity is in extremely high demand. These high gas prices mean that only whales are able to use the network and everyone else is forced to sit on the sidelines, waiting for fees to go down. Until fees do go down, many people are entirely “locked out” of their funds since it’d cost more in transaction fees to move the funds than the funds are actually worth. It’s hard to predict how demand for Ethereum L1 will change going forward, but there will come a time where the L1 gas fees will never drop to previous rates and a large amount of funds are permanently locked. This problem is particularly unfortunate because it disproportionally affects poorer users.

How big of a problem is this?

To understand what the current value is that is lost to gas fees, I leveraged Dune Analytics to query the account token balances, the current value of these tokens, and the current gas prices (150 gwei). We look only at current account balances and not historic balances or transactions. As of writing this, the current total lost value of ERC20 tokens alone is $600,687,720.

Which projects are most impacted

We can also see which projects have lost the most value due to gas fees. Among them are stablecoins like USDT, USDC, and DAI — governance tokens like UNI and meme coins like SHIB and ELON. In the following charts, I’ve broken down loss by user type.

🦐 = Users with less than $100 in their wallet

🐟 = Users with less than $1,000 in their wallet

🐬 = Users with less than $10,000 in their wallet

🦈 = Users with less than $100,000 in their wallet

🐋 = Users $100,000 or more in their wallet

Here are some of the highest locked tokens and their distributions:

Projects focused on building a strong community (ex. Uniswap and Chainlink) may want to consider the large number of their existing community that this impacts and potential solutions to rescue their tokens. To look up how this impacts a given project, enter the token address into this dashboard.

If you’d like to look up how much value your personal address has that’s currently inaccessible, enter your address into this dashboard.

Methodology and disclaimers

To measure the total lost value, I look at the value of each token for each account and compare it to the gas cost to transfer it. If the gas cost is higher, then I add the token value, otherwise I just add the gas cost. For example, if I have $30 of $MKR and it costs me $40 to transfer it, then $30 gets added to the “lost” bucket. If I also had $50 of $OHM, the $40 (the gas fee) gets added to the “lost” bucket. So essentially $600M is total lost value due to gas fees. If we just consider tokens which are 100% unclaimable (totally underwater) the value is $77.8M.

The amount does not include ETH or NFTs. ETH was not included due to the query limitations in Dune. NFTs, which likely make up a huge percentage of lost value, were excluded due to the difficulty in deriving a market price for each unique asset — especially those that are currently inaccessible due to gas fees.

For gas estimation, I used a simple ERC20 transfer with 65,000 gas. In reality, the user would likely send funds to a bridge or an exchange contract which will cost much more gas.

It’s also possible that for some illiquid tokens, the amount locked due to gas fees significantly lowers supply and pushes up the token’s price. This would over-estimate the value locked, since all of these “unlocked” tokens could never really be sold for the actual market value.

What’s the solution

Speed up L2 adoption

We may not be able to save the L1 users already stuck in this situation but at least we can get new users to L2 faster. Exchanges and institutions are actively working on supporting withdrawal directly into an L2. Additionally, token protocols can provide stronger L2 migration incentives. Part of the reason why this is so slow to date is that no clear L2 winner has emerged. Protocols will have to be comfortable choosing an L2 and potentially migrating to a different solution in the future.

The great token rescue

That’s great for new users, but what about the hundreds of millions of dollars already stuck in L1? Unfortunately, we can’t have a “gas free day” that allows everyone to move their value to L2s while still getting to keep most of it. Each application will need to “force migrate” their users onto an L2 if they want to rescue their token holders. It’s not enough to mint L2 tokens for these users, since the gas fees on L1 may go back down at some point, giving these users twice the number of tokens that they should have. Before minting on an L2 can occur, the L1 tokens must first be burned. To ensure fungibility across tokens and compatibility remains with other contracts, it’s likely that all tokens will need to be burned on L1 and recreated on the L2. This is no simple effort, but those platforms that do perform this migration for their users will likely be rewarded with a loyal community. The goal of this post is to start the discussion. If you have other ideas for solving this problem, please share them in the comments or on Twitter.

Dune Analytics

While writing this post, I got to learn how to use a platform called Dune Analytics to query on-chain data. They pre-aggregate a handful of blockchain data streams (balances, prices, events, etc.) making it really easy to start querying on-chain data. You can also build on others’ work. If you’d like to play around with the data in this article, you can fork the query and party on. If you do, please share your findings in the comments! My only feature request for the Dune team is to allow for extended duration queries for paid users. Most of the queries that I ran timed out simply due to the number of accounts and balances that I needed to process.