On the faq page (https://docs.yearn.finance/faq) it states that the performance fee for a v1 vault 5% is And the crveurs vault is a v1 vault. But when reading the contract variables on etherscan of the strategist contract for the vault you can see that both the treasuryFee and strategistReward are set to 10%. Which is a performanc fee of 20% which is 4 times higher than the advertised fee.
@jhack, thanks for pointing this out! Your question made me dig a bit deeper trying to understand what was going on here.
As you correctly point out, a v1 vault is having a 20% perf fee set, while the docs say it should be 5%.
What happened? The short answer is that v2 vaults took longer than expected to roll out, and in the meanwhile we updated fee structures on v1 vaults in accordance with YIP-52, but failed to reflect this in the docs.
YIP-51 set a 2% mgmt fee and 20% performance fee for v2 vaults specifically.
YIP-52 set a new performance fee structure in general, where 10% would go to the Strategist and 10% would go to the Treasury. YIP-52 builds on YIP-51, but is not referring to v2 vaults specifically.
At the time of writing, v2 was âweeks awayâ and as a co-author of both YIPs, I naively thought v2 would be rolled out much faster, and that all v1 vaults would transition to the v2 vaults immediately.
In reality, this transition is taking much longer, and is more gradual. Weâre now expecting this to play out over the coming months.
In the meanwhile, new v1 vaults were rolled out, for additional CRV gauges. The strategists responsible have followed YIP-52, which has the perf fee set to 20%. Older vaults were also updated to the new fee structure. This however has not been accurately reflected in our documentaiton, it was missed completely. The docs still make the distinction between v1 and v2 vaults.
I propose we take two immediate actions:
- Update docs to capture the above accurately.
- Display the fee structure for each vault on the website, to avoid any confusion.
I welcome other ideas and thoughts as well. Once again, thank you for flagging this, was in our blindspot and Iâm not sure we would have noticed it had it not been for your post!
Update:
- Display the fee structure for each vault on the website, to avoid any confusion.
This is now live on website, as a tooltip over the vault version number. It also shows how much CRV each vault locks into yveCRV (backscratcher).
As a User that deposited a very substantial amounts to the Vaults V1.
This is not ok.
Firstly changes in pricing should be notified to the users so they can make informed decisions.
Your own Team/Support was providing users with information about 5% fees. If not for that I would probably not deposit into the Curve Vaults since they follow simplistic strategy that does not change (that was the argument we were provided anyway regarding the difference in fees between v1 and v2 vaults).
Now we learn that all v1 got fees increased basically 4 times without informing anyone for vaults that follow simplistic Curve strategy that most of the users could execute themselves without paying those fees.
I get you want to charge more for USDC or DAI vs 2 vaults since they follow complex strategies that probably need to be constantly tweaked . But do not see this reasoning for v1 curve LP strategies. (since v1 vaults strategies do not change anyway)
What does stop Yearn from sneakily increasing the fees again since most of us myself included donât constantly change the vault contracts values. This just destroys any trust you had with users that you really need after recent DAI vault hack.
I understand your feeling about
sneakily increasing the fees
Fee and settings are monitored 24/24 and 7/7 by quant desk whales , i guess a discord / tweet bot api oracle notification tool can be use or fund .
It is not about feelings but very real financial lost. It cost me couple of k $ total to deposit into the vaults now if I wanted to withdrawal from strategy this would cost me much much more.
Why should I be the one paying it? Since I just followed Yearn provided information?
And do you think with all the scamming around it is reasonable for Yearn users to try and to build some bots etc only to monitor the fees to make sure that Yearn did not sneakily increased them?
On top of the defi risk and hacking risk as per Dai vault hack?
I mean take all of this risk for what exactly?
Some users like me want to deposit and not constantly track the information. Every fund you invest into have this transparent. Besides you should not increase the fees on existing contract ever. You should inform the users and create new vaults with the new fees while disabling the deposits for the old vault so new users would not be able to use it. I would consider it fair.
But not the situation we just learned about
It always makes me chuckle how people find those curve strategies as just simple contracts that deposit and thatâs all when they are complex as hell (the joke that talks about how only 2 people in curve actually know how curve math works isnât actually a joke). If itâs so simple, anyone could just write them and charge less fees to eat yearn alive but the true is that the competence opted for just forking yearn contracts, charge less fees and, even after that, they have problems to offer similar returns to yearnâs even after adding additional tokens rewards. Incentives to strategists are important to keep innovation high (or you can just wait for other protocol to innovate and fork it without any shame), using a fraction of the rewards to keep the rates offered competitive is a must ( or competitors can just fork, add shitty tokens as rewards inflating the returns and eat yearn alive) and of course yearn needs sources of income because it doesnât run on thin air ( the maintenance of the entire protocol and of course the insane amount of gas paid by yearn so users donât have to care about that).
Totally agree on the outdated documentation, hope yearn can incentivize more the documentation writers and other community members because we also have to eat everyday and staying updated about all the yearn changes is very time consuming.
Update
- Update docs to capture the above accurately.
Pull request submitted, should fix this and a bunch of other stuff that was identified in the same go.
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.