Yield farming is moving at light speed right now, and I expect there will always be new strategies and optimizations that can be made even after things slow down, making YFI a valuable meta-yield protocol to govern. However, I noticed the main way value accrues to the YFI token is fees from the yield (feel free to correct me if I’m wrong here).
One issue that may occur is parasitic farms. Take this image as a dumbed-down version of how I’m perceiving v2 to work:
What’s preventing someone from taking their own forked vault that doesn’t take any fee and pointing it at the same controller/strategy? Since all of this is happening on-chain, depending on the design of yearnV2, someone could trivially design a parasitic farm that rides on the strategies that YFI governance comes up with without paying the associated fees into the YFI system. I believe even in a worst-case scenario someone could manually fork all of the contracts for themselves (including having no fee).
There’s a similar issue that was discussed in the prediction market community about parasitic markets [1], or oracles in general.
I want to raise this issue to pose these questions for consideration as we move forward:
What can/is being done to make parasitic farming harder?
This can be things like making the controller contract not reveal which strategies it’s delegating to on-chain except in cases necessary for the system’s function. This at least makes a parasitic farm have to manually follow along to keep up with new strategies.
What can we do to entrench YFI/yearnV2’s value proposition in a way that parasitic farming is actually unappealing relative to just using a yVault?
This could be things like protocols having special incentives for the yearn pool that a parasitic farm wouldn’t get, using an exchange contract that doesn’t accept non-yearn liquidity, or anything similar that makes the gain in yield from pooling in specifically yearnV2 relative to a parasitic farm greater than the fee on the yield.
Gas costs should inhibit parasitic farms from attracting liquidity. They would need to attract sufficiently large TVL from sufficiently large set of users such that:
(1) the gas cost of rebalancing/rotating their funds < yEarn gas cost + YFI fees, and
(2) their farm has enough interactions from users to constantly trigger the rebalncing/rotation and achieve the same optimality as yEarn.
Recall: yEarn contracts are “asleep”, they don’t just rebalance/rotate on their own … they only do so when a user interact with them in some way by depositing/withdrawing etc.
You read my mind– the past few days I’ve been thinking a LOT about the fact that this is exactly what protocols like YFII or YFV are likely to do– they copy-paste strategies/contracts from yearn, then advertise them at 0.1% withdrawal rates or something silly like that. Realistically, I’m unsure if companies like Akropolis (Delphi) will start to do similar things as well.
There’s a part of me that is hopeful that yield farmers would respect the work that the yearn team and community is putting in to optimize yield protocols, but I know there will definitely be those who will only go where the maximal yield is. I would appreciate those with more technical knowledge weighing in here as well.
There’s also a competing part of me who believes that Andre created this to share with the community while at the same time optimizing his yield– so I’m not sure whether treating this as intellectual property of yearn would really fit in the ethos of yearn in general. But I definitely think this is something we should be discussing.
In addition to the liquidity and large userbase moats, yEarn will probably gain privileged access to other DeFi legos (which I like to call “hooks”). Here is an example of YFI planting a hook in Maker. A fork of yEarn would have to convince the Maker governance to grant them the same which is unlikely if they’re just a clone or parasitic farm. This privilege is tied to yEarn’s contract address.
Another advantage of large liquidity, is that it makes it economically viable to harvest more often. This leads to a greater compounding effect over time, all though there are diminishing returns to harvest frequency.
In fact, this has not happened so far. Rather the opposite in fact. With the exception of Curve, YFII’s vaults use different strategies not used by yearn, recently targeting “fad” coins like YAM and ZOMBIE. They have also now raised fees on profits to 10% (when accounting for all fees including token burn), so they are not competing on price at all.
That said, I don’t think these fears or concerns are unfounded and still think we should continue being proactive in the approaches that have been outlined above. It’s a matter of when not if, as the TVL game is zero-sum. It would definitely be of great value to ideate various different moats to ensure we can minimize the effect of predatory competition.
