The AMM Wars

Control for decentralized market making has just begun.

Canoe Capital
4 min readFeb 20, 2022


A refresher on how AMMs work. (source)
A refresher on how AMMs work. (source)

Nobody in TradFi is market making to lose money.

Back in 2017, everyone assumed large banks and institutions would buy up Bitcoin and we were all going to Lamboland. In reality that didn’t happen, and we ended up in a ~2 year bear market. In retrospect, institutions were right not to buy Bitcoin, due to the low liquidity and high slippage on many centralized exchanges. Institutions don’t like losing money. This was later resolved by improved infrastructure such as deeper liquidity and reliable over the counter (OTC) desks and now companies and institutions can be orange-pilled.

DeFi is at that earlier stage of Bitcoin now. To us crypto natives, it is awesome! To outsiders, particularly intuitions, it is murky and full of risks. Automated market makers (AMMs) are a key DeFi primitive where assets can be exchanged. The benefits of AMMs are that they provide liquidity in a decentralized manner, allow users to collect trading fees and yield tokens and allow for permission-less trading of assets. While the threat of smart contract exploits are rapidly decreasing through better auditing, key structural problems of AMMs still pose risks scaring off liquidity providers (LP). Impermanent loss (IL; demonstrated below) is an almost guaranteed risk for LPs, especially without LP incentives or higher fee collection. This begins a cycle where MMs take away liquidity, which lowers total value locked (TVL) and leads to further IL for those left in illiquid pools. In TradFi no one is market making to lose money, so why should people deal with it in DeFi?

Demonstration of impermanent loss. As one asset increases in price (X-axis) at a a more rapid rate than the other, total liquidity value decreases (Source).

Note: I am talking about AMM DEXs. There are DEXs with private MMs who make money via traditional means.

War is coming.

While people can opine about all the problems of AMMs, this is a space of innovation and solutions. The Curve Wars has been probably one of the fascinating stories in DeFi. Curve (and other AMMs, notably Bancor) has been innovating to lower these structural risks in AMMs, such as using like-like pools (lowering IL), algorithmically concentrating liquidity (lower slippage and efficient fee collection) and tying liquidity rewards rate to staking and governance. These strategies are strengthened by their huge TVL (>$19B at time of writing) and has made CRV token a hot commodity. Convex Finance has ultimately won this war by controlling 86% of CRV tokens. They now control through the DAO which pools get the highest CRV rewards and also a large part of the huge TVL under Curve. In turn, Convex provides stability to Curve by permanently locking up assets and preventing bank runs from selfish whales. From the user perspective they earn higher yield from Convex’s established staking and extra CVX tokens.

Other AMMs have also introduced similar risk mitigation strategies and incentives to increase liquidity. SushiSwap increased their liquidity through incentives paid in the Sushi token, leading to a vampire attack on Uniswap, though Sushi really hasn’t decreased some of the AMM-associated risks. Bancor provides liquidity rewards, single-sided liquidity and impermanent loss protection, all of which makes it a promising AMM for both retail and potential institutional MMs. However, due to the space limitations in Bancor protocol, their TVL has stunted, though they do plan to resolve this issue in their V3. Platypus on Avalanche is a curve like AMM, with features such as time-based reward boosts and single-sided liquidity which has been previously seen in Bancor.

Control of these other AMMs have already begun. A number of protocols are already bidding for control of Platypus’ token and liquidity (ie: Echidna Finance, Vector Finance and Yield Yak) . Frog nation under Dani Sestagalli launched a take over of Sushi not too long ago. And some on the sidelines are waiting for Bancor V3, perhaps to offer yield optimization in exchange for control of LP tokens, incentives and governance.


In conclusion, I see high demand for AMMs specifically focused on: 1) mitigating risk, 2) building liquidity, and 3) protocol controlled incentives. The combination of these critical points provide fertile ground for a potential war to control these AMMs.

The net earnings for last quarter for Coinbase was about $1B. According to Grayscale, DeFi is still quite early, with less than 5% of crypto users using a DeFi protocol. As AMMs become lower in risk and secure higher TVL, these AMMs become a major force to be reckoned with, something whales, institutions and decentralized autonomous organizations (DAOs) are taking note of.

Disclosures: I have financial interests in Curve, Convex, Bancor, Platypus and other AMMs. I have not taken payment of any kind for writing this piece and opinions are my own.

Comment here or on my twitter @CanuckLink

Originally published at on February 20, 2022.