I am referring to the first two figures in the original post.
The first one:
In the linked document it says:
Gnosis has by far the most assets earning yield. It’s yield assets are valued at
more than half of all other DAO yield assets put together. It is also the DAO with
the largest allocation of its portfolio earning yield: over 80%.
So the figure gives some information on how Karpatkey uses funds in the treasury: it invests them into yield farming assets. If I understand correctly, it does not tell us how profitable the strategy is. Yield farming assets can also change value or generate little yield.
The second figure is:
It seems to depict the revenue of Karpatkey. It shows that even during the first half of the year 2022, Karpatkey profited from managing treasuries. If I understand correctly, it does not tell us how profitable it is for DAOs to have their funds managed.
True. I am sure you can come up with some other quantitative argument to justify that you achieved your goal stated in the beginning:
Thanks for your interest in our proposal and for the sharp questions. We’ve updated the chart image with the rest of 2022 in the proposal.
Regarding the PFPs, we preferred its aesthetic over pictures. We have nothing to hide, but having an org chart with actual photos felt too Web2. We’re currently 18 people at Karpatkey (some of them like to be anons), and we’re growing our team, since we’re convinced that bear markets are the best moment to build.
With regard to the fees, here are the formulas (we’ve added them to the proposal as well):
Management fee = Weekly portfolio balance * 0.5% / 52
Performance fee = (Weekly liquidity mining rewards - trades slippage - deposit/withdrawal fees) * 10%
We wouldn’t mind charging the performance fee in ETH instead of COW tokens, we proposed charging it in COW to minimise the burden on CowDAO’s finances and as a token of our long term strategic alignment.
Thanks for taking the time to put together a detailed reply!
The pie chart illustrates the proportion of yield-earning assets of the 33 DAOs with significant treasuries that were obtaining a yield. Your understanding is correct; Karpatkey advised investing the tokens in strategies of diverse complexity that you can find in our weekly reports, where you can visualise how profitable they were.
The second chart shows the accumulated weekly revenues in millions of $ that GnosisDAO perceived in 2022 (we’ve just updated it), as well as the average APR obtained each week as a result of the investment strategies we suggested to them. They’re the main DAO with which we worked in terms of treasury size, and in this chart, you can visualise how profitable our strategies were to them at an aggregated level.
That second chart helps justify how we achieved the goal of sustainably growing funds with low exposure to risk. We have curated a selection of risk-adjusted investment opportunities in reputable protocols, putting together at least one strategy per week, which we executed diligently to allow the DAOs we’ve worked with face all their financial obligations on time while helping them grow their treasury. Also, we carried out 6 emergency meetings to de-risk positions and prevented liquidations of collateralised loans during sudden market downturns.
Thanks for the fast responses and for clearing up a misunderstanding on my side!
Would it be fair to say that the chart
essentially shows yields. And that it does not take into account that yield farming assets can also decrease in value resulting in a net reduction of the value of assets under management, even without any operational expenses. If that is the case, it might make sense to have some metric in place to monitor that value as well during the cooperation with CoW DAO. If I am wrong again, congratulations to an exceptional performance of zero weeks with losses during crypto winter.
Yes, that chart essentially shows yields. It shows the aggregated performance of the entire portfolio, which is disaggregated in our weekly reports. We’re working to report on impermanent loss in a mathematically precise way.
The math behind its calculation is more complex than one would think, not only because positions sometimes receive several deposits and several withdrawals at different points in time, but also because it differs across protocols and specific cases within protocols. The basis of its calculation is the following formula, which calculates the IL per LP token for any AMM with constant swap fees:
I think his question was not related to impermanent loss.
Let’s say there would be an asset that has a constant yield (on single asset yield farming, so no IL) of 100% APR. Karpatkey would buy that asset and generate very high yields and reinvest it weekly. But the asset over a year slowly goes to 0.
Is it correct that the chart would show very high yields (decreasing growth rate over time, but still), although it was a horrible investment?
Thanks for the clarification. We take token prices into consideration in yield calculations, so the reported yield of an asset with a very high APR and a steep price reduction would be low.
The price reduction of any given token will reduce the yield shown in all positions which have that token, since we express it in USD.
So let’s assume this:
A DAO holds $100k worth of ETH and nothing else.
Then the DAO gets managed and the management sells all ETH and buys token FARM for it.
Token FARM is then staked and yields 10% APR per week (the 10% remain the same for the entire year).
Over the year, ETH price remains constant, while FARM loses 50% of its value.
So after 1 year, the DAOs original strategy would have resulted in a DAO value of $100k.
The new strategy results in $50k (value of FARM) + yields (which is around 10%*(100k+50k)/2=7.5k), so $57.5k total.
However, the DAO management received
management fee - around 0.5% of 75k (bit more actually)
performance fee - roughly 10% of 7.5k
So although the management strategy underperformed the original DAO strategy, there is still a somewhat significant fee being paid in this example.
So the price development of the assets held is not truly reflected or is it?
I mean, it should not be reflected vs. $ in my opinion anyways, as the DAO would not have held $ anyways. But it should be reflected vs. the portfolio the DAO would have held by doing what it has been doing so far IMO
Assuming a sustained linear price decrease, our management fee would also decrease week after week (its current formula incentivises us to grow the $ value of the total portfolio). This is aligned with the calculation of the performance fee, which also rewards greater yields in $ (and would constantly decrease in your example). A scenario like the one you describe would therefore result in a decreasing income for Karpatkey.
Reflecting the price development against the original portfolio of the DAO would be tricky, since it would depend on the token price, which is an exogenous variable defined by the market as a whole. In your example, Karpatkey’s performance would be poor, but if the day after the payment of the fee, the FARM token price rises 200% while ETH’s price remains unchanged, Karpatkey’s performance would have been outstanding, yet not rewarded. Therefore, we believe that our fee structure favours a win-win balance between us and CoW DAO.
I understand your concern. One of the decision drivers behind having the treasury professionally handled vs. doing it in-house is the expertise and experience of the team that takes care of it. We’ve learnt many lessons in the last couple years that have improved our management capabilities, we constantly update our risk management matrix and have a dedicated research team. Achieving the same synergy with a single contributor or small in-house team is hard to accomplish.