One of the main motivations behind Balancer Protocol is to allow AMMs (automated market makers) to fully utilize idle capital even if it is not evenly spread across different tokens. Today anyone can provide liquidity using Uniswap pools, with the requirement that the value provided with the ERC20 token is the same as the value provided with ETH. This means that all Uniswap pools are even, “50/50” pools.
Some really clever ways have been devised to combine Uniswap with other DeFi tools to allow for more customized exposure strategies (see DeFiZap). However, we at Balancer Labs believe there is significant demand waiting on the side lines for a native way of providing liquidity with full flexibility.
Potential Use Cases
Liquidity Bootstrapping Pools
Liquidity Bootstrapping Pools (LBPs) are smart Balancer pools that allow projects to optimally bootstrap liquidity for their tokens. This is possible because projects can provide only a small fraction of the total pool value in ETH and/or DAI (e.g. 10% or 20%) and the rest in their own project tokens, usually an abundant resource for them.
As time goes by, the gradual decrease in the project token weight also means that the project is slowly distributing their tokens in return for ETH and/or DAI.
While some investors like to diversify and distribute their wealth evenly across a number of tokens, many like to make big bets in specific projects they like the most. These investors can find in Balancer a way to put their whole holdings to use: by providing liquidity to uneven pools.
Founders and Advisors
Projects usually give founders and key employees skin in the game through equity, and often tokens. These token holders will therefore probably hold the lion’s share of their portfolio in their project tokens. Balancer allows them to earn trading fees and at the same time help their projects by increasing the market’s liquidity.
A closer look at Uneven Pools
If this is the first time you are seeing this term, please refer to this article by @pintail. In summary, impermanent loss is the loss in value when investing liquidity in a pool compared to just holding tokens.
The following chart shows the impermanent loss for three different Balancer pools: 50/50, 80/20 and 95/5. The x axis represents the relative price change between the two assets in the pool as measured at the moments of adding and removing liquidity — 1 represents no change — and the y axis represents change in value of the pool compared to holding — 1 means the pool is worth the same as holding. Notice that we did not consider any trading fees for simplicity’s sake.
With a 5x change in price, the impermanent loss for a standard 50/50 pool would be 25.4% whereas in a 95/5 pool it would be only 3.88%, over 6.5 times smaller.
With Balancer, if a token holder believes in a token, they can keep strong exposure to that token while at the same time earning trading fees through balancer pools.
Imagine you are bullish on ETH and want to provide $1,000 worth of liquidity with ETH and DAI. Let’s consider 3 different pools, again with different ETH/DAI weights respectively: 50/50, 80/20 and 95/5. The chart below shows how the pool value would change with a changing ETH price assuming an initial ETH price of $250:
Notice how uneven pools allow for a much tighter correlation to only holding ETH.
Slippage and APR
In terms of slippage, the most efficient weights for a 2 token pool are 50/50. This means that for a pool with say $200,000 worth of assets, having each asset represent $100,000 of value will minimize slippage for trades.
The chart below shows the slippage in % that a pool with a total of $200,000 worth of assets incurs. Assume the pool has tokens A and B. The chart shows the slippage for varying weights A and a trade of $1,000. The chart is symmetric around 50% because the sum of weight A and weight B is always 100%.
We have seen that Balancer pools allow for a novel area of experimentation around weights choice. While allowing for a higher value correlation to selected tokens, uneven Balancer pools incur in higher slippage, which reduces trading volume and APR.
Only examples with two tokens have been covered in this article. However the same logic applies for pools with 3 or more tokens. We then consider the ratio of weights between two tokens, since any trade will involve two — and only two — tokens at a time. For example, a pool with four tokens and weights 40/40/10/10 would behave like a 50/50 pool when analyzing trades between the first two tokens but like an 80/20 pool when analyzing trades between the first and the last tokens (40/10 = 80/20).
If you want to learn more about Balancer follow us on twitter @BalancerLabs and join our discord channel .