from Osmocon Video Part 1: 02:42:37 - 02:56:40
@tarunchitra
A regular on Zero Knowledge podcast and more recently, a regular segment, The Chopping Block, on Laura Shin’s Unchained podcast
tl;dr
Attracting liquidity makes DeFi protocols useful, but managing incentives to attract and retain that liquidity can be difficult. Underpay and have little liquidity. Overpay and you inflate away your token value and create feedback loops where the rich get richer and liquidity does not flow where needed. Olympus bonds provide one interesting example of feedback control (systems that dynamically adjust to on-going real-world data based on certain parameters, which themselves may be dynamic as well) in liquidity management, where, in this case, having many different durations of bonds that are priced along a pseudo-yield curve reduce the volatility of the underlying asset, in this case, OHM. This feedback control theory can be generalized to include more parameters, like trading volume, fees, percent of OSMO staked, limiting inflation, sizing different pools differently, capping pools, making sure all or most pools have at least some incentives, etc. With more parameters, the intuitions stay clear, but the math gets more difficult.
[Gauntlet and others (including our own incentives working group) are working to research and implement these types of feedback control into Osmosis incentives]
—
Osmosis incentives are complicated: staking rewards, liquidity mining rewards, superfluid
-how to reason about these things?
Who are the real users of DeFi?
Liq. Providers - have assets, seeking yield — risk takers
Traders / Asset Managers - long-term financial goals
-e.g. they’re a DAO that needs to pay contributors
Keepers - users who insure that prices are sync’d across mkts and chains
What is DeFi Liquidity Management
-the act of attracting liq. providers with incentives
fees, governance tokens
-protocols usually have to distribute to a lot of pools, but a fixed supply
So, how to allocate?
-if they protocols overpay, token ownership becomes concentrated and price goes down
-if they underpay, they have trouble attracting users, and the protocol is not useful
-if you don’t optimize your liquidity?
-activist investments: Convex w. Curve governance (Aura about to do the same with Balancer)
-in Cosmos, people love sovereignty, so you probably don’t like that
-protocols die if they don’t properly do liquidity management
Why is it hard?
-lots of pools == expensive
-pool incentives have to be dynamic based on mkt. conditions
-ex. high vol. in an asset, but your protocol has lowest incentives, you need to increase them to attract liquidity and make trading in that asset a good exp. for your users
-incentives interact: can create feedback loops
ex. removing incentives from one pool and putting them in another
-all the volume in the protocol goes through the best incentivized pools
Renting, Leasing, and Buying Liquidity
Rent - providing incentives without locks (might be mercenary and leave)
Lease - need certain liquidity over a duration, (i.e. incentives per unit of time)
Buying - DAO earns fees, has gov. tokens, has stables — it provides liq. itself
How to Reason about these Things?
-Olympus DAO was the first to experiment with Protocol-Owned (or Controlled) Liquidity (or Value)
(PCL, PCV, POL, POV)
-ignore their Ponzi aspect: 5m% staking yields
-the bonds were the interesting part: locked people up for a long time, but people could redeem over time
-these bonds enabled the DAO to modulate how much they were buying, renting, and leasing