As I write this, ethereum is moon-bound. ETH holders are making a bundle, with the price up 370% this year alone. In all this fanfare, though, you may have heard some chatter about MEV (miner extractable value).
In a nutshell, the Ethereum blockchain is written by consensus, but the content of each block is chosen by just one miner. Miners can profit from users by front-running, back-running, sandwiching and generally exploiting transactions in their block however they choose. The Flash Boys 2.0 paper, written by researchers at Cornell Tech, coined the term MEV to describe such exploits.
But is it really that bad? Isn’t it inevitable? And who cares if there’s a little extra slippage on Uniswap if by the end of the day your coins are worth 5% more regardless. We’re all making hay here. Perhaps the whole MEV thing is overblown, FUD even.
As an algorithmic trader for 12 years, I was the first to predict pre-genesis in 2014 that MEV would become an issue. So let me wind the clock forwards once again.
It’s 2035 and the ETH price has risen to $100,000, except that dollar price doesn’t mean anything anymore because no-one uses dollars, they use ETH. Mission accomplished. We’re where we want to be.
But with Ethereum at full adoption the price is no longer moving up. You can’t grow wealth just from holding ETH anymore beyond 1.6% APR in staking rewards. It’s hard to imagine now, but by this time the currency is so ubiquitous that it’s invisible to most people. Boring even.
Today, around $1.4 billion dollars of MEV is being taken from Ethereum blockchain users annually from a total decentralized finance (DeFi) market of around $50 billion. Global financial markets are worth $100 trillion.
So, in 2035, now that Ethereum is the global financial marketplace, how much MEV will that be? $20 billion extracted in MEV per year? $200 billion? No. It’s $2 trillion worth of wealth annually taken from ordinary people due to an unfixed network vulnerability (in 2021 money). That is just shy of the yearly national budget of the United States and more than China.
That’s before any other taxes, remember. So the population of the world will be charged the entire annual budget of China in a socially useless tax that goes directly to the new masters of the universe (hint: very likely the same old masters of the universe) who have no obligation to build a single hospital, school, road, wind farm, law court, library, food bank, etc., and then you still have to pay your actual taxes on top of that.
Bad, right? Yes, but that’s just the start because Ethereum isn’t just distributed finance. To our kids ether is the dominant global currency and it’s used everywhere: when you buy insurance, when you buy a train ticket, pay a restaurant bill, go to a game, buy a pizza, pay for your tuition.
You order some groceries from an open marketplace smart contract. A small local firm sees your order and can offer you the best price, including biking it over to you. But Mega-Corp has paid the MEV auction winner to censor all grocery transactions except theirs from the block. They don’t even have to compete on price. You overpay for your shopping and the local store closes down.
You manage to find an NFT [non-fungible token] ticket for the Dua Lipa comeback tour for $50 (in 2021 money) on an auction dapp. When you try to buy it, a bot sees your transaction and front-runs it for the same price. But, don’t worry, in the same block they’ve sold it back to you for your maximum bid of $100.
You want to convert your ETH into carecoins to fund your mother’s cancer treatment, but the MEV auction winner keeps moving the price away from you. This isn’t some exotic DeFi trade you’re executing here. These coins are vital for the well-being of a loved one. You can’t afford the complete course of treatment needed by your mother, your sister, your child.
Oh, but come on, we’d never let this happen, right? This is madness. We’ll have fixed all this by then.
We’d better hope so. The outcomes I have just described are artifacts of a severe data integrity issue in the Ethereum network that must be resolved. Solutions are possible, but whether they are worked on or not is down to us as a community.
Dapp designers can use more MEV-resistant design patterns in their smart contracts. Timelock, SGX or threshold encryption may be used to hide transactions from attackers. I'm collaborating on my preferred solution of a decentralized content layer with fair ordering that engages root causes. It's promising, but I really don’t care if this or some other approach gets implemented as long as it truly addresses the issue. I started the project to galvanize the community away from fatalism and towards real solutions.
Decentralization takes work. It is expensive in development time and computing resources. The payoff is that such systems can be robust and equitable. But if a decentralized solution becomes more vulnerable and less fair than a traditional centralized competitor (as well as more expensive) then it is unlikely to succeed long term.
I got involved in Ethereum in 2014 full of hope that it would offer an alternative to the corruption that was laid bare in the financial crash of 2008. I still hold that hope. I dearly want us to get there, but we're not going to while maximally exploitative MEV auctions are our primary response. These no more fix the problem of MEV than running a market selling stolen credit cards helps the victims who had their cards stolen. In fact, I have shown that MEV auctions worsen the situation by introducing exploits in which only the most wealthy can profit.
MEV is potentially more damaging in layer 2 (where transactions are made off the main chain) than in mainnet because the rollup sequencer is more powerful than any one miner. One leading scaling solution provider is openly discussing building this exploitation of users into their protocol, not by accident this time, but by design.
MEV is inevitable? Not unless we choose it to be. What needs to be inevitable is that MEV becomes history because if it doesn’t, we’d better start hoping Ethereum fails or we’re all in a lot of trouble.