Strategist rewards are way too low at the moment and it is unlikely that new strategists would even cover the costs of testing.
Strategists currently receive 0.5% of the profit from their strategy. Imagine a strategy that averaged $1m AUM at 10% APR and lasted for 6 months. The strategist would only earn $250 ($1m * 10% * 0.5% / 2). Which is less than what they would have spent.
This may not have been a problem for V1 as AUM and APR were very high per strategy. But going into V2 there will be multiple strategies per vault sharing the AUM and APRs are settling.
From a growth point of view: We want to attract the best strategists. I believe that Yearn’s edge is going to be that it will employ complex strategies that people won’t trust anon quick fork developers to have safely implemented. But building those are difficult and so strategist compensation needs to make it worthwhile. Especially when we get to the point that strategists are teams and not just individuals.
From a safety point of view: Underpaying the people responsible for creating a safe strategy for millions AUM is dangerous. Even disregarding bad actor risk, if the only way a strategist can make money is to churn out new strategies quickly they will take shortcuts in testing.
From a selfish point of view: I am a strategist building the new leveraged DAI comp farming strategy and will probably lose money on it so want to change that.
Increase strategist compensation. V2 vaults haven’t been completely finished yet so it can come from multiple areas. It can come from increasing overall fees. Or sharing some of the AUM fee. Or increasing strategist share of profits above 0.5%. This should be a discussion.
For: Prosperity for all
Against: Lose all our strategists to other projects
I propose a mixed fee structure, where a fixed bounty is accrued through fees more aggressively with higher fee structure and a lower fee (e.g 1%) is payed out afterwards.
This compensates the work put into the strategy fairly quickly and also rewards longtime thinking.
The longtime fees should be very low to keep vaults competitive to other platforms.
I love this. Besides being the life of yearns vaults…it is a valuable incentive, crucial for talent attraction and having well-funded strategies in general (from a financial and intellectual capital perspective).
$1m AUM is not really a realistic figure for Yearn strategies or vaults. Whatever gets put up will likely have much higher numbers, the current yDAI vault has $36m in there. I’m familiar with a levered DAI strategy as I’m doing it myself, and its about 17-20% APY. I can say with almost certainty that no strategist will lose money on Yearn vaults/strategies if they end up being approved.
I think you should be more specific regarding exactly what changes you want to propose for the fee structure.
Stopped reading at $1M AUM.
Yep pretty much
If a strategy only got 1m aum it ain’t that good to begin with
$1m was just meant to be descriptive and a starting place to do your own maths. How much the strategist is being paid per million invested. I don’t know what the average will be and hope we’ll see more deposits as v2 gets released. But don’t just use overall vault figures for your guesses. there will be multiple strategies per vault.
With regards to the 10% over 6 months think that is pretty accurate for a strategies lifetime. Starts higher and then slowly drops until it is turned off. Some will be higher and burn out quicker. Some will be lower and last longer.
As to if any strategists lose money, we’ll see. Gas is very expensive and deployment and testing costs a lot. Once the first few strategies are live we will have better figures on that. Hopefully those costs will go down as we get more experience with v2 and more efficient with making strategies.
After developing a self-sustainable harvest bot and a couple of strategies I have a few observations.
Remember “subsidized gas”? Harvesting should be paid for from the performance fee. This was never solved, so be it. All my strategist rewards for 3pool go towards keeping the strategy running. 0.5% is not sustainable to keep strategists happy.
Other point is developing a strategy. It is very expensive to deploy and test contracts, I’ve spent over 15 ether so far and I’m not finished yet. The $250 from the example wouldn’t even pay for the Vault + Strategy deployment if you miraculously manage to nail it from the first attempt.
I think the initial split was a huge mistake and a 10x bump of strategist reward is warranted.
I think the harvesting costs should be paid out from the protocol, and if the Strategist pays them, they should be reimbursed. Is this not the case?
Also in terms of testing/deploying, if the strategy is approved it should also be reimbursed by the protocol. There is a way to make this more frictionless for developers and inclusive for new strategies.
We need to open up the discussion channels more to understand each part so we can come up with a long-term solution. Everything should be aimed at making it easy for developers to add value, without being constrained financially.
Considering the current fee structure, and also other projects targeting 20-30% performance fees (and 20% being a traditional performance fee), I think we can easily bump this up. Obviously, these fees should scale based on difficultly and the work done, I would suggest 4 tiers as follows:
- 5% - simple copy-paste jobs, no effort
- 10% - remixed from another strategy with little change
- 15% - moderate work, including simulation and analysis
- 20% - very complex design, including simulation, analysis, and extensive on chain testing
We could also make the governance performance fee a flat 5% to account for gas costs for keeper solutions.
If Yearn has the best strategies and a veCRV advantage (see my other thread) we can make the whole show %2 of AUM and 20% of profits and compete on excellence instead of price.
Or maybe 1 and 10 as probably few would agree with me we should go so far.
- Professional and in line with the world’s best hedge funds.
- That income if AUM are sustained will lead to massive $YFI appreciation - and that is important.
- Provide strategists with say 2% of profits. Or sliding scale with AUM (5% for small pools down to 1% for big pools). Remaining 8% to YFI holders.
If we have the best income for investors in vaults, paying 10% or 20% is no big deal at all. Time to compete on excellence instead of price.
Note that governance still has to approve Strategies to the Vault, so it’s not like these fees are set in stone. But it is nice create a set of guidelines moving forwards.
I am obviously for.
Just something crossed my mind : A vaults strategy market where users could simply choose they vault and anyone could publish a vault, setting their own fee.
Users would obviously choose whichever vault provide them the most value (return minus fee).
It would basically become an open market showcased by YFI where everyone is incentivized to build the best strategy and can set an higher fee if they think their strategy is the best.
I said from get go should be at minimum 1% of the profits. I’m open to any percent that makes sense for both strategists and the users. I like the idea of earnings from profits as there’s incentive to make high profit generating strategies which is good for users, yfi holders, and the strat creators.
Charging a larger performance fee makes sense. Hedge funds, CTAs, and alternative investment professionals typically use a 2/20 fee; call the 2% a management or anything else you like but it still produces diminished return for end customer on a long enough timeline. I would recommend we do away with the management fee used traditionally and implement a performance fee.
Within the performance fee we could do a bit of innovating: sliding fee’s based on complexity, fee vesting, etc. The most important thing to get right is the flexibility to distribute the funds from the fee. If a strategy does markedly better than others, whether through pure AUM or higher Sharpe Ratio (risk-adjusted performance) we would want to incentivize strategists to bring that to Yearn through higher incentives. It would be possible in these cases to allocate a higher percentage of funds that would have gone to treasury towards strategists as the pie would be bigger while operational expenditures would likely not grow.
It strikes me that it will be important to map capital and operational expenses to their end result, this is one of the value adds of crypto, transparency, right?
If anyone has compiled books for Yearn or know of anyone who is working on something like that please let me know
Perhaps an initial one off payment to repay gas fees spent during development for an approved strategy then an on going %?
The tier approach makes sense as not all strategies are created equals. We want to encourage high yield and high quality strategies. Though, I would modify the performance fee slightly.
1% - simple copy-paste jobs, no effort
5% - remixed from another strategy with little change
10% - moderate work, including simulation and analysis
20% - very complex design, including simulation, analysis, and extensive on chain testing
As Klim suggested, any discussion about fees should start with data. I’m working on preparing a dataset so people can easily simulate different fee structures.
I’ve modeled a new fee schedule trying to compensate for the removal of the withdrawal fee. I’ve backtested on yUSD vault since it is one of the largest and oldest product with more than 3 months of historical data. Surprisingly the traditional 2/20 from the traditional world matches almost perfectly.
- 5% performance fee: 466,822 (17% of total)
- 0.5% withdrawal fee: 2,243,078 (83% of total)
- Total fees: 2,709,901, which will be our benchmark
- 20% performance fee: 1,867,288 (73% of total)
- 2% management fee: 699,237 (27% of total)
- Total fees: 2,566,525 (95% of target)
The split of performance fee between treasury and strategists is left as an exercise to the reader, but I suggest an equal split.
I’ve shared the full dataset here so you can play with other models.
Another idea is to consider Warren Buffett style model where there is no management fee and only performance fee. I believe his structure early in the years were something like 0% management fee + 25/30% performance fee above 6% annualized return.
The reason to remove mgmt fee is to avoid rewarding managers who sit on large AUM earning mgmt fees yet don’t really generate alpha. So while managers are taking the “risk” of less compensation when performing below expectation, the reward/ compensation when alpha is generated is higher.
Perhaps reasonable annualized return hurdle can be thought of as median of stablecoin yields on yEarn products.
Separately, should there be a consideration of vesting of rewards to strategists?