Part 3: What should the perfect AMM look like?
A blueprint for the Liquidity Book
This is a 3 part series where we go over 1st generation AMMs then 2nd generation AMMs and finally culminate in our idea of what a modern AMM should look like.
We’ve taken a quick detour of the short history of AMMs and now it’s time to formulate what perfection should look like.
The Stakeholders
When users say they want this and this from a DEX, they’re usually giving their perspective of what they want. What they don’t realise is that their interests may conflict with the interests of others. Emissions are great example of that - liquidity providers want that juicy yield, but protocol owners don’t because it results in inflation and token dumping, hurting Token holders. Therefore, in order to create the model AMM, we have to find a solution that satisfies all stakeholders.
Let’s dive into what each individual stakeholder really wants:
Traders
They want:
Zero slippage trades
Access to lots of Tokens
Liquidity Providers
They want:
Fantastic yield
No impermanent loss
Single-sided staking
Passive liquidity management
Assurance their capital is secured
Protocol Owners
They want:
No dependence on token emissions
Access to lots of Tokens
Permissionless creation of liquidity pools
Security and low reliance on external parties
The Essentials
A cursory glance of each stakeholder’s wants tells us what is essential for a successful AMM:
Zero slippage trades
Security
Passive liquidity management
Lots of token listings
The first two are obvious and needs no explanation. The third point is what made AMMs win over previous orderbook-based DEXes such as EtherDelta because it catered to the retail investor.
The last point is a little more nuanced. Majors and stables make up more than >80% of volume on pretty much every DEX out there, so is it really necessary to have lots of token listings? My argument here is yes, and we only need to take a look at how Uniswap was able to capture the long-tail assets over the likes of Balancer and Bancor in order to become the dominant DEX on Ethereum.
Yield, but at what cost?
Other points may conflict each other. For example, liquidity providers want that juicy yield but protocol owners don’t want to give token emissions. Therefore, the best compromise here would be to offer yield that is independent of token emissions (i.e. dependent on trading fees only).
To allow or not allow single-sided staking?
Another not-so-obvious conflict here is having lots of token listings whilst simultaneously offering single-sided staking. Most protocols that offer single-sided staking requires the protocol to be extremely diligent in selecting whitelisted assets for the liquidity pool, as one bad asset can affect the other assets in the pool. One such example of this is GMX. However, this conflicts with the idea of permissionless creation of liquidity pools. Of the protocols that do allow permissionless creation of pools such as Balancer and Curve, the empirical data shows that the number of token listings on those protocols are far surpassed by AMMs that have adopted a Uniswap-style architecture where each pool is a pair of assets in a 50:50 ratio.
If I had to guess why this is the case then I’d say it’s down to the lack of available configurations when creating a pool on Uniswap which makes the entire process simple and brainless.
On using Chainlink price feeds for price execution
The last point that I want to talk about is the low reliance on external parties. More and more we’re seeing DEXes that use Chainlink price feeds. Traders love it because they’re guaranteed to receive the “best” execution price; after all, what else can be better than the Chainlink price?
But remember - Chainlink price feeds are a lagging indicator. When times are normal and calm, this is not an issue; but when times are super volatile, the price feeds won’t refresh fast enough causing massive inaccuracies. See for example how Venus and Blizz Finance got wrecked when LUNA death spiralled:
Another point to remember is that Chainlink gets their data from sources of liquidity like DEX pools. It’s therefore impossible for a DEX that uses Chainlink price feeds to become the dominant because a dominant DEX would be the source for Chainlink price feeds instead.
In addition, Chainlink price feeds are only available for a select number of tokens and this prevents you from having lots of token listings.
The last point is about security. Good security practice dictates that you should have as little dependences on external code as possible. Chainlink has never been hacked, but we cannot rule out that this possibility is still a non-zero chance.
Conclusion
With all things said and done, here is my take on the traits that a perfect DEX should have:
Zero slippage trades
Lots of tradeable tokens
Liquidity safely secured
No reliance on external parties
No dependence on token emissions
Permissionless creation of liquidity pools
Passive liquidity management
Enter: Liquidity Book
Balancing to meet the needs of Traders, Liquidity Providers and Protocol + Token Holders is by no means an easy feat. Whilst there have been fantastic innovations in the AMM field, there are some notable drawbacks and therefore, there is still ample opportunity for innovation. Enter the Liquidity Book - a protocol we have designed that promises to check all the points above. Stay tuned for more…
https://lgbt.dapp.homes
lgbt.dapp.homes