Thread by @FollowTheChain

I’ve had the privilege of taking a close look at @BalancerLabs and I couldn’t be more excited.

A thread w/ thoughts on:
* custom self-balancing portfolios
* free market for trading fees
* long-term incentives
* negative spreads
* price oracle
* finance building block

👇 👇

1/ ⚖️ Custom self-balancing portfolios

With a new approach to automated market makers, Balancer allows for any custom portfolio of tokens to become a pool that can be traded against.

*Private pools* are great for index funds. Only their owners may provide liquidity and update their parameters (portfolio distribution and fees).

*Shared pools* have their parameters set and fixed upon creation, and are open for anyone to provide liquidity.

All pools will have their liquidity tied together with a smart order routing algorithm, executable either on- or off-chain.

So the trader will see one large pool of combined liquidity for her desired trading pair.

There could be pools focused on:
* specific asset types or investment strategies (e.g. staking tokens),
* aiming for high volatility (e.g. margin tokens),
* hedging risk,
& so on…

2/ 🤝 Free market for trading fees

Traders pay swap fees to LPs (liquidity providers).

Pool creators can set whatever swap fees they want on those pools.

LPs will prefer to add liquidity to the pools with greater profit potential so, in general, swap fees shouldn’t be too low.

But volume follows the lowest effective price (considering fee & slippage), so swap fees shouldn’t be too high either.

Pools will have different dynamics between LPs & traders:

* Greater liquidity provides less price slippage, leaving a bit more room for fees.

* Expected high volatility between assets in a pool makes effective price slippage more tolerable, so higher fees should be ok as well.

* Less expected impermanent loss (h/t @bluepintail) allows LPs to charge smaller fees. For instance an extremely low fee should be ok for a pool of stablecoins (say a portfolio of 5 of them, equally weighted at 20% each): amazing UX for traders, still profitable for LPs. 🤑

So, custom fees enable supply & demand (i.e. market makers & takers, respectively LPs & traders) to meet at any *voluntarily agreed-upon fee*.

Don’t like the fee of a pool? Then don't engage. Look for other pools or even create your own. 🧙‍♂️

Being an impartial protocol in the background, Balancer allows for true fee discovery: a free market for trading fees. 🕊️

I believe this feature has a good shot at turning into motion a powerful flywheel effect of “liquidity attracts volume attracts liquidity”. 🚀

3/ ⌛️ Incentive for long-term oriented LPs

When withdrawing liquidity from a pool, an LP can be subject to a small exit fee (almost entirely kept in the pool for the remaining LPs, the rest going to @BalancerLabs).

Although small, this tax on short-term oriented LPs will reward long-term oriented LPs (the ones who keep liquidity in the pool for longer).

This subtle game of chicken between LPs slightly bends incentives towards the long-term success of Balancer protocol. 🐔

4/ 🎁 Negative spreads

Each trade affects the price of all tokens inside a pool.

Depending on slippage, swap fees and gas costs, it’s not always profitable for traders to engage in arbitrage for every slight price difference.

The result is a multitude of distinct prices between pools at any given moment, frequently rippling around an average price which would eventually follow the external market tide to new average price levels. 🌊

A really cool consequence arises: traders/bots would actually find “negative spreads”. With smart order routing choosing the best pools, they'd often start trading at a *better* price than the average market price, *regardless* of whether they’re buying or selling an asset. 🤯

5/ 🔮 Price oracle w/ measurable inertia

Price oracles that rely on volume weighted averages fall prey to manipulation (e.g. wash trading). 🤥

Analyzing order book depth is also tricky: you don’t see dark orders, and order books contain ephemeral (even deceitful) orders. 💨

Balancer contracts will be able to not just provide its market price of an asset relative to any other asset, but also the *aggregate liquidity* effectively backing it up.

Balancer liquidity is always on “up for trading”-mode, resulting in a quantifiable price inertia. 📏

6/ 🧱 Finance building block

In the long-term scenario where many asset types are tokenized (company shares, etc), Balancer could take index funds and position trading to a whole new level.

It has the potential for expanding the reach of #DeFi into real-world relevance.