This post summarizes all the recommendations I have put together to create a Yam Treasury Management and Maintenance System. There has been subjective decisions made in this process that were based on my 10+ years of trad-fi financial planning and my 1+ year of being a Yam Dao contributor. My directive while putting together the recommendations is to allow for growth of the treasury while limiting the downside risk by simplifying currently approved risk model to streamline processes so that maintenance will be easily accomplished without specific expertise that might or might not be available at a future date.
Creation of silo:
Recommendations made here, will be summarized below but for full detail refer to this link:
Recommendation #1: Change treasury targets from Beta / Correlation model into Asset Allocation Model
Primary reasons why:
Investment strategy targeting a specific beta and correlation initially was created is generally a good metric to assess risk for non-crypto portfolios. Since the creation of Yam’s treasury and investment strategy it has become apparent that this method of assessing risk is not the most optimal. A few reasons why:
Crypto assets in general are highly correlated therefor using beta and correlation with respect to ETH as a metric is not as helpful as originally intended. For example, since most assets that an on-chain treasury can acquire either very close in beta/correlation to ETH or a stablecoin, so it does not provide any quality information on which assets to hold in the treasury.
Beta/Correlation does not account for liquidity, inflation, potential revenue and etc of tokens.
There are crypto primitives that allow for hedging risk but the hedging space has not fully matured and carry additional risks beyond the financial instrument (liquidity risk, smart contract risk, ect)
Rebalancing assets becomes a hard question to answer. If the treasury needs to rebalance and there are multiple ways to do it, which assets do we sell, which do we buy?
Current Model = Beta / Correlation Target
Allocation Model = ETH / BTC / Stablecoin / DPI
Seek another alternative
Don’t do anything with treasury
Recommendation #2: If asset allocation model is selected from above
Use sushiHOUSE allocation model with Risk level of 3
Use sushiHOUSE allocation model with new Risk level TBD
Create new allocation model
Additional rules and guidelines to maintain efficiency:
Investment strategy does not include operations reserve which would be a 12 times monthly expenses.
Assets that do not fit this model will be disposed of as long as slippage doing so does not exceed 3% otherwise it will be kept in treasury but ignored.
Rebalancing trigger limits will be set at ± 10% allocation. All assets will be sold/bought back to stated allocations when triggered. Rebalancing can be done at most once every 30 days.
Recommendation #3: If sushiHOUSE allocation model with Risk level of 3 is selected from above. Treasury should adopt these yield farming strategies:
I recommend including the Yearn Vault 3 Crypto (BTC/ETH/USDT) earning 5-9% APY.
This strategy uses a convex reinvest including the LP swap fees rewards between the 3 assets. No leverage, uses well established protocols to earn fees and yield.
Applying to Allocation Strategy
If we were to adopt the new strategy of 55% ETH 20% BTC and 20% Stablecoin 5% DPI and add Yearn Vault 3 Crypto (BTC/ETH/USDT).
The assets could be allocated:
stETH = 35% ETH
3 Crypto = 20% ETH / 20% BTC / 20% Stable USDT
At a $3,000,000 treasury the allocations:
stETH = $1,050,000 earning 4-7% APY
3Crypto = $1,800,000 earning 5-9% APY Approximate APY revenue $132,000 to $235,500 a year
Update treasury to include yearn vault strategies to earn yield
Do not proceed
Recommendation #4: Protocol Owned Liquidity
Ross has written a few forum posts on protocol owned liquidity (POL):
This would be valuable to Yam because:
It would allow YAM to stop offering liquidity farming rewards.
Allows the treasury to earn trading fees.
My recommendation for the simplest solution that requires no maintenance:
Currently TVL in YAM liquidity is ~$615,000. We could use $310,000 of the treasury’s ETH and mint an equalivalent amount of YAM to be paired in a Uniswap LP pool. Whatever amount of YAM minted would need to be burned if the Protocol Owned Liquidity were ever unwound.
Discontinue liquidity farming rewards.
The POL would not be spent or utilized in other ways. It is only created for providing a liquid market for YAM.
Use treasury to create protocol owned liquidity + end liquidity rewards/inflation
This is much clearer than your other post. So thank you for that.
I voted, and have a few comments about my votes that are not captured in the vote.
Allocation vs beta/correlation
Strong agree with the allocation model, but that question should not assume the allocation model is ETH/BTC/stable/DPI. We are either changing it or we aren’t. If we aren’t then nothing happens. If we are then the next question is to determine the allocation (which you do, but the poll doesn’t reflect that).
Voted “new allocation model” but I would be happy with the SushiHouse #3 risk but without the DPI. I’m fine with that risk factor, but don’t see the benefit of DPI in there. If we are changing the allocation to something new there is no reason it has to be exactly the sushihouse model. Instead we can turn that 5% into more stETH, earn more yield and get very similar price action/exposure while simplifying rebalances. DPI has underperformed and I see little reason why it won’t continue to underperform ETH in the future, and 5% provides minimal diversification.
We should also commit some funding to making it so we can sell all assets in our treasury. Simply ignoring the funds that we can’t sell is short sighted and is a solvable problem.
Agree here assuming the above votes pass. Again, even better without DPI since that is the only outlier not earning yield.
This all seems reasonable to me. As I have mentioned before I hope that in the future we can move to an 80/20 model and potentially pay contributors partially in that, but it is a separate question and shouldn’t stop this initiative.
Where would the line be drawn if we are changing asset allocations? Why are those lines drawn that way and not another way. While I see your point, your headed towards very subjective decision making in crypto that does not have good long term established investment rules. sushiHOUSE risk 3 and allocation model has had more eyeballs and discussion vs any of the other models discussed. This does not mean that it is any better or worse than anything you or I can come up, it’s closer to the idea of the Lindy effect.
Here are some reasons that I think we should keep DPI in the treasury:
I think of it as a potential place holder for another investment “like” in the future.
Past performance doesn’t equal future performance, it could be an outperformer.
My response from above.
DPI itself is a super interesting experiment into a crypto native Defi Index fund, in this respect it has a similar ethos to Yam.
the line would be drawn by the governance process. We would discuss what assets the DAO should hold in this model and then there would be a vote. This isn’t rocket science. Yes, there are subjective decisions. That is why governance exists. To decide upon them. Your role, as an advisor, is to recommend something (this is subjective too!). Token holders are then able to ask for changes and vote on that.
You keep implying that sushihouse is some tried and true model that is better than anything else but in reality it was used once by a protocol that literally forgot it existed for pretty much all the time they had it. There is no lindy effect for sushi-house. It was just some numbers that you and krugman put together and everyone thought they were good enough. Did it work reasonably well? Sure, I guess so. But so did my strategy for trading the last bull market. That doesn’t mean that I won’t adjust that strategy in the next one.
So with that said… finally you are engaging with the actual issue and giving reasons for holding DPI.
There is no benefit to doing this. Nowhere does your model imply that it requires a placeholder to add new assets.
I can’t disagree with this. It could be an out-performer. Or it could continue to under-perform, which is more in line with the data that we have. We could say the same for every asset that we hold so it is a silly argument. We either decide if we thing it will outperform or it will give some other benefit. One benefit would be diversification, but we have learned that it provides little of that in a portfolio mainly of ETH and BTC. The other would be what you state in your point 4.
While Index Coop may be a cool DAO doing good things, I have a hard time seeing this as a good reason to add an asset to a re-balancing treasury if it doesn’t fit the goals of the treasury. There are lots of DAO doing cool things with similar ethos to YAM, but we don’t have them in the treasury. This is a slippery slope and exactly why I wrote about separating out “partner” assets from the investment portfolio.
If we don’t include DPI we have a re-balancing portion of the treasury that in simpler (3 vs 4 assets), all earning yield (in Yearn), with an almost equal risk and volatility profile.
Removing DPI is just taking that one extra step to optimally simplify the thing we are trying to simplify. We are using the lessons from the past to update the treasury on other ways. It just makes sense to also update it in this way.
In this case it was good that they forgot about it and is now working to utilize the funds during the down turn in crypto. The sushi tokens they sold to create sushiHOUSE is worth significantly less than current value of the portfolio.
When I was a trad-fi financial advisor, when it came to subjective decisions, it was always better to lean on a decision that has history vs something that in theory sounds good but is minimal in effect. In this case, there is some history with using the sushiHOUSE model vs removing the 5% DPI. As it is not that significant, I lean towards keeping it vs removing it.
Your points are valid, but so are mine, this is the problem with subjective decisions. If you want to vote to remove the DPI from the portfolio, feel free to.