Pickle as a Uni V3 active manager - what is the hypothesis?

The governance call mentioned that a bounty for active Uni V3 management has been accepted.

What’s the hypothesis here?
What are we trying to achieve?
Where is the alpha/opportunity in active LP management?

Uni V3 liquidity provision is an extremely competitive and tough field. I estimate that 95%+ percent of LPs are dissatisfied. I estimate that a few entities running wide market making operations on CEXes and DEXes capture majority of the net liquidity provision profits.

I think Uni V3 has shipped code suitable for liquidity mining on V3. I could agree with a hypothesis that pickle could aim to farm that rewards, although I’m skeptical.

I don’t think I’ll accept a hypothesis that Pickle finance has a secret sauce that will make users money. My hypothesis is that Pickle’s uni v3 active liquidity manager will significantly underperform 50-50 pools when accounting for liquidity mining rewards.

Setting the range very wide largely obviates the need for a manager.
Setting the range tight means that we’ll be forced to buy high and sell low on each rebalance.

I’m not familiar with the V3 active managers that operate currently, but I think they are largely underperforming.
This comes from Banteg (about uni v3):
“All of the active liquidity managers have shown worse performance than simple holding.”

I’m generally happy to take a bet that any strategy you’ll post here will underperform a 50-50 pool with liquidity mining rewards.

I personally think active V3 management is a lost cause and the effort should be abandoned. V3 suits Alameda, CMS and Jump trading alike. They have big pots and they can manage V3 liquidity as a part of a very big picture. Having a small pot and using it all to provide liquidity in a tight range is a prescription to lose money.

Even a design with 33% of the pot in stETH, 33% in Uni V3 USDC-ETH, and 33% of the pot in MIM3cvx (stable) following the 50-50 curve is a bad idea.

Building such a tool adds a lot of complexity and attack surface.

If you think you have alpha that can’t be shared, happy to hear it on priv, pinky promise I’ll not sell it to Harvest Finance :slight_smile:

Here you can track the active managers underperforming HODL:

Happy to explain maths why this is happening and point out flaws in liquidity management designs.