Quote:
Originally Posted by stinkypete
This is indeed a point that many smart people don't understand. Yes, an AMM and any market maker that captures the spread or a fee will perform better in a choppy or mean reverting market. But it's not +EV in an 1. efficient market. It's simply a bet on mean reversion, and an inefficient one at that, 2. particularly in high gas fee environments. If you have an edge forecasting mean reversion, there are much better ways to bet on it with better returns and lower risk than throwing your money in an LP pool 3. that will get destroyed in a trending market.
1. And, what does that have to do with the current crypto market?
2. Arbitrage primarily goes through Flashbots/mev-geth
3. Show your work.
Regarding better ways to bet, other types of edges require tangential skillsets, access to certain platforms, capital/margin requirements, in some cases being awake/alert/at the computer. It's not as simple as, "if you can do x, you'd be better off doing y". There are different tradeoffs involved.
I would agree with you though that putting money in an AMM with no math/data science work, no additional incentives, and no "game selection" will, if not already, quickly become -EV in v3. You no longer have the benefits of v2 where all [lazy*] LPs were receiving pro rata payoffs and auto-compounding with each trade.
*seen some quite advanced mempool-watching alpha in v2 =)
Also worth mentioning, you can use uniswap v3 to buy/sell at prices you already want to buy/sell at. In that way, you will always outperform yourself as you receive fees when your orders get executed as opposed to paying them to exchanges. Maximizing net fees (gross fees + incentives[if any] - IL) may be the goal of a linear-wealth-preference profit maximalist, or professional MMer, but that is not most people.
Quote:
Originally Posted by Clayton
are you talking on a relative basis? because i fail to see how the guy tossing 3.7M into ETH/USDC on a range of 600 to 20k is gonna lose. sure it's probably suboptimal considering your relative share is lower on the more active ranges, but that guy i reckon is just straight up not getting IL and can just ignore it for all they care.
This is very similar to a uniswap v2 range, which would go from 0 to ∞. v2 still leaks to arbitrageurs. In a prolonged down market you'll get destroyed (both because you're long ETH *AND* short gamma). In a flat / ranging market, you'll do fine. In an up market, you'll make money but may not outrun IL.
I'm interested in a DEX protocol that can:
1) dynamically adjust fees based on market conditions
2) discriminate spread based on sharpness of flow
I think we will continue to see the space evolve.
Last edited by Two SHAE; 05-26-2021 at 07:01 PM.