A perspective shift on automated market makers through MEV

It takes time for new ideas to be properly understood. Often the way we first formulate an idea, is in hindsight not the simplest or the most elegant way, and a change of perspective can lead to new understanding. I think we are in this situation with automated market makers right now, and MEV (miner/maximum extractable value) is forcing a perspective shift.

Automated market makers (AMMs for short) are a relatively recent development, and although the math behind swaps is well understood, I don’t think that this is the full picture. Even if the swap itself is well understood, the mechanism can be flawed.

Flash liquidity in Uniswap v3 is a clear example of this: it’s a mechanism that admits a profitable strategy which was not intended by the developers, and this profit comes at the expense of the intended users (regular liquidity providers). In a sense, it’s a game-theoretic vulnerability in the protocol. Sometimes these flaws can be fixed at the protocol level, but sometimes they indicate a flaw in our understanding.

Swaps are limit orders, not market orders

My original understanding of AMM swaps was that they are like market orders: you trade asset X for asset Y, and you pay whatever the pool’s current price is at the time the swap executes. When the swap executes, the pool price may be different from the price at the time when you entered the order, and to protect against unexpected expenses, you can enter a maximum slippage percentage. This sets the maximum price that you are willing to pay for the asset. If the pool price is above your maximum when the swap executes, the transaction fails. I think the intent of the max slippage was just that: to limit slippage at busy times, when the price can change due to other swaps executing before yours.

All was well for a while, until miners discovered miner extractable value (now also called maximal extractable value), or MEV for short. Extractors started sandwiching swaps, and the effect of that is that you always pay your maximum price, not the pool price.

In a sense, in the presence of sandwiching, a swap behaves more like a limit order than a market order: if it executes at all, it executes at the price that you set, not at the market price. Like with a limit order, the price you pick is a trade-off: at lower prices your order might not be filled (your transaction may fail), at higher prices you risk overpaying.

If you accept sandwiching as a fact of life, then our understanding of AMMs is wrong: the pool price is not the market price, it’s a minimum price. We’ve got our user interfaces backwards: we shouldn’t display the pool price and then hide the max slippage under advanced settings, we should own up to it and include the selected slippage in the price.

Make it a feature, not a bug

If you look at it from this point of view, isn’t it crazy that we let MEV extractors take the difference between the pool price and the maximum price? Isn’t that a flaw in the AMM’s design? A game-theoretic vulnerability?

So here is a proposal: let the AMM take the difference instead. Users who swap always pay their maximum price, and the difference between the pool price and the maximum price goes to liquidity providers, as part of the swap fee. For users, nothing changes. They pay their maximum price either way. Liquidity providers benefit, and the MEV opportunity goes away.

Conclusion

When software contains a critical vulnerability, we frown upon the attackers who exploit it, but we also recognize that the only way to address the issue is by patching the software. What is really bad, is a software vendor who leaves a vulnerability unpatched when it is being exploited in the wild.

I think we should approach sandwiching — and unintended MEV opportunities more generally — the same way: as a flaw that needs to be patched. I frown upon the MEV extractors who sandwich. But at the same time, I think the real blame is with AMMs for not doing anything about it. This is not obvious if you think of AMM swaps as market orders. But MEV is forcing us to change our view, and recognize that AMM swaps really do have a limit price.