OMG Impermanent Loss!
The strategy brushed off IL and assumed ETH/YFI hodlers are less sensitive to it. This is the major flaw, I agree.
For starters, the “maximize Uniswap deposit” principle as stated in the strategy is not the optimal thing to do. Yes if the YFIETH price is just ranging, IL is low and sweet profits from Uniswap/Balancer trading fees are adding up nicely. But with sharp directional moves, it’s the capital outside Uniswap that is now more valuable because the strategy uses it to effectively take back what arbers took from us. The vault has its opinion of what the AUM should be valued at, and when it buys/sells excess ETH (residing at yETH) or YFI (at Aave), it is simply just offsetting the IL … because how else would have there been an imbalance to begin with? h/t Tarun
We want the sweet fees on Uniswap, but we also need enough fire power outside Uniswap to strike back at arbers.
So I propose these amendments:
-
split funds 50:50 between Uniswap and yETH/Aave.
-
split the rebalancing actions over small tranches of maximum $10k in order to (1) reduce slippage and frontrunning, and (2) take advantage of massive run ups in either ETH/YFI. If the vault recognizes a re-balance is needed, it would execute a maximum of $10k per tx with imposed time delay till the next rebalance. This parameter can be tweaked depending on the depth of the YFIETH pool. Currently slippage at the Uniswap’s YFIETH pool is 0.01% for a $10k worth of swap.
-
as before, the vault ignores +/-5% price changes to avoid unnecessary flip-flop rebalances during thin ranges.
Why not also deposit to Balancer and/or other AMMs
Totally possible and easy to implement. However, keep in mind that asymmetrical pools (say, 80:20 ETH:YFI pool) will definitely not attract as much volume. This is also true for Dodo and Mooniswap, who put explicit mechanisms in place to shield LPs from IL. When you take money away from arbers, they stop pointing their bots at your AMM, volume is stunted, and everybody loses.
Why not just buy an insurance policy for yETH (e.g. opyn) or buy options @ramaruro @@alexander
That costs a significant premium, while with yETH-YFI-BULL providing insurance for yETH, there is only the opportunity cost for the duration that rescue ETH remains in Maker during black swan events (which shouldn’t be too long barring a complete meltdown of Maker protocol and/or governance). Meanwhile, that safety ETH/YFI is being productive (unlike premiums which are lost capital).
Uniswap already balances
Uniswap doesn’t re-balance the IL-induced imbalance. The vault has an opinion about how much the AUM should be worth had there been no Uniswap.
Less YFI for Governance
This is an important issue but it deserves its own proposal for the community to decide whether YFI in vaults can vote or not (personally Im in favor of allowing, but it’s weakly-held).
What’s the price Oracle @thegismar
We have the Maker oracle (OSM), DEXes, and the Open Oracle initiative by Compound which is now live on-chain and aggregates/medianizes from CEXes as well (so does Maker)
People already can LP on Uniswap themselves, why do we need a strategy for that?
The gas costs of re-balancing would destroy their capital. Unless of course they’re advanced professional traders who are offsetting IL through elaborate financial wizardry on CeFi.
@madavidj
Being 50-50 with infinitesimal rebalancing means I will never get exposure to the full exponential growth if one asset leaves the other behind.
You could withdraw your funds and ride the trend, but there is no guarantee you would sell the top or close to it. The strategy does this incrementally through the re-balancing.
When considering the viability of a strategy, one should ask “is this useful for sufficiently large enough people to warrant implementing it?”. A strategy could meet that threshold and not be attractive for you personally.
@mattdw @iTo @Tiarizzi93
why not create a subsection of the yETH vault, say 10-20% of it, that is held in ETH:DAI Uniswap
I think the existing yETH strategy was sound, but would amend it to maintain a 400% ratio instead of 200%
Can’t we just take a cut from the deposited ETH in the yETH and keep it as reserve in case of a blackswan event where DAI repay does not work?
That would reduce risk indeed but it also reduces capital efficiency significantly. With yETH-YFI-BULL, the yETH vault can enjoy the capital efficiency of 200% c-ratio but at a reduced risk of liquidation. If the AUM of yETH-YFI-BULL is >= yETH’s Dai debt, that risk is actually eliminated entirely (barring and meltdown of the Maker protocol itself).
@Tiarizzi93
why not just create a pool internal to yEarn (only accessible to the token holders) where token holders provide YFI and ETH is provided by yETH vault which was over-subscribed and just keep this as reserve for a blackswan event?
Again, capital inefficient, and you’d lose out on Uniswap fees.
@CryptoOGkauai
Why does it have to be ETH ?
(1) ETH is massively liquid (2) the intersection of ETH and YFI bulls is large (3) there is no non-leveraged DeFi investment opportunities for ETH holders (low APY on lending for example) (4) We get the insurance of yETH as an invaluable side benefit
@BNR34
What if everything dumps (both YFI and ETH) causing a liquidity crunch on DAI. ?
That’s when y-ETH-YFI-BULL capital is moved to Maker to avoid liquidation.
@iTo
I would propose we ADD YFI to Maker as collateral and repeat the above yETH
That would make the strategy more leveraged since Dai would be drawn against it to make it productive capital. One of the goals of yETH-YFI-BULL strategy is be leverage-minimized. But sure we can do that if Maker agrees to add YFI, my opinion here is weakly-held.
Keep in mind that the question you ask is “is this useful for a large set of people?” not “is this a strategy I would use?” because obviously highly skilled traders can outperform, but not all people are skilled traders or can afford to hire one.
