Entropy’s stewardship of the treasury management proposal resulted from a failed proposal from Karpatkey ~6 months ago, an attempt to get things moving from Avantgarde that fell short due to community concerns around their ability to remain neutral, and general consensus from delegates that the DAO should have some sort of treasury management strategy to put its idle assets to work. Our primary goal with the GM Track has been quite simple since the beginning: Earn higher yield on idle treasury ETH than market buying an ETH LST(s), while maintaining a similar risk profile. Treasury management is, in our view, a program that optimizes for stable and safe returns with ecosystem support as an added benefit. Ironically, we do envision it being leveraged to more directly support builders over time, whereby the DAO owns $1B of yield-bearing ETH, stablecoins, and RWAs with all of the proceeds funneled back into the ecosystem to support new and existing builders on Arbitrum as a way to support perpetual ecosystem funding through a non-dilutive funding source.
Big picture aside, we want to highlight that this process was intentionally designed to protect the DAO against a situation where it felt differently about the proposed choices. No funds from the Treasury Management V1.2 proposal will be deployed until the DAO has approved the proposed solution via Snapshot with 3% of the votable token supply in favor/abstain. There is, per the constitution, a 1-week customary waiting period between the initial forum post and Snapshot, meaning the earliest date this proposal can move to Snapshot is Thursday, March 6th (yes, we can and should discuss this in ETH Denver). If the community sentiment is overwhelmingly negative, we are happy to alter our selections, wait an additional week, and post to Snapshot at the appropriate time. This is why the original proposal stipulated that the preferred choices of the GMC would be posted to the forum in the first place. We are happy to see that the DAO is fulfilling its function as an authority in this decision and look forward to altering the recommendation in line with the DAO’s desires.
With all of this in mind, we would like to provide more context into how the GMC arrived at the conclusion it did with the initial “preferred choices”. We spent countless hours on phone calls with teams submitting applications/answering their questions, back-and-forth over telegram with various applicants fielding questions, and many long calls amongst the GMC members reviewing each application together. Every protocol application contained a ton of nuance and required in-depth conversation, as one would expect. The GMC did enter this process with a mindset to reserve a portion of the funds for Arbitrum-aligned/native protocols, which in itself is a hard category to define, but after further discussion, opted against it with all factors considered. The applications that we identified as the 4 most promising proposals within this category were GMX, Dinero, Camelot/Jones DAO, and Dolomite (all proposals can be found here).
GMX:
TL;DR of the proposal: Deposit a portion of the 7,500 ETH into GMX’s gmETH single-sided liquidity pool, which earns trading/borrow/spread fees paid by traders. Additional options included BonsaiDAO/Umami’s ETH/USDC backed GM pools with automated hedging against USDC exposure, and dHedge/Toros’ vault that LPs in the WETH/USDC GMX GM pool and supplies ETH on Aave while also borrowing stables on Aave with automatic rebalancing. The submission noted that multi-collateral pools that support both WETH and wstETH are expected to be supported by the end of Q2.
Dinero:
TL;DR of proposal: Mint and hold 500 orbETH, LP 500 ETH worth of orbETH/WETH on Balancer.
Camelot/Jones:
TL;DR of proposal: Deploy 1k ETH in LST/LRT Camelot pools such as wstETH/wETH, weETH/wETH, ezETH/wETH, rsETH/wETH through the automatic management of Jones’ vaults.
Dolomite:
TL;DR of proposal: Supply 500-1k ETH to Dolomite borrowers, and an undefined amount of ETH to lever up against GMX’s gmETH/USD on Dolomite by borrowing USDC against the underlying to buy more gmETH/USD, thus achieving 1x ETH exposure in practice.
Why didn’t the GMC select any of the aforementioned, and where did it go wrong?
Without singling the influencing factors of any particular application out, each of these proposals included some form of the following elements that made the GMC arrive at the list of preferred allocations:
- The DAO would have made up a vast majority (>90%, in some cases) of the TVL of certain pools/vaults involved in the proposed strategies
- The strategy’s yield was paid out in illiquid tokens, making it difficult for the DAO to inform the Arbitrum Foundation how to handle said rewards. Do they actively vote if these tokens have voting power? If there is an unlock period, at what point does the Foundation initiate this and what is the process for doing so? If there is revenue share with these locked tokens, is the Foundation expected to claim them, and if so, how often, and what do the rewards tokens get converted into and how? Etc.
- Fees related to vault operations.
- Direct price exposure to LRTs.
- Lack of traction/PMF of the product on Arbitrum or other ecosystems
- Exposure to factors such as loss of principal, trader PnL, borrow rates that are out of the DAO’s control with strategies that involve leverage, and liquidation risks (not maintaining ETH exposure).
- Required the Arbitrum Foundation to actively manage a large number of individual positions, which would increase operational complexity/risk.
This is where the GMC went wrong. While we did very clearly outline in the application template that risk would be the most important factor, we told applicants who asked, that low-risk strategies should be included alongside medium/high-risk strategies if they wanted to submit those, and that we felt it was implied that safe strategies were preferred given the fact this is a DAO treasury RFP, we should have provided feedback to applicants based on their submissions and communicated the programs intentions more clearly.
Additionally, the GMC did not believe its “place” was to judge what “Arbitrum Alignment” meant. For example, if a protocol applicant pursues a multi-chain strategy, is it still considered “Arbitrum Aligned”? If a protocol applicant announces that its token is launching on a chain not named Arbitrum, is this protocol “Arbitrum Aligned”? The GMC decided that these questions were not for it to answer, but rather for the community to signal via the forum post period, and for alterations to the choices to be made in accordance to this feedback. The GMC simply put forth its most objective “preferred choices” from the perspective of a risk-minimized treasury management strategy that can be built upon.
Aave is highly conservative, and a regular choice for idle asset deployment for many other DAOs. It is also Arbitrum’s largest protocol by a large margin and brings significant value to the chain and ecosystem, and comes with service providers focused on risk, a strong treasury, and a safety module to protect against adverse events. Fluid is a rising star on Ethereum mainnet that has yet to solidify its L2 home base - which we believe can be Arbitrum. Please read the “Justification” section above for more information. These two protocols, together with Lido, perfectly achieve the original intention of the GM Track stated above “Earn higher yield on idle treasury ETH than market buying an ETH LST(s), while maintaining a similar risk profile” with the added benefit of fostering ecosystem growth by deepening lending market liquidity, wstETH liquidity, a reward share with the Lido DAO that can grow into something larger, welcoming an Ethereum native application that has proven high-growth in nature to its new L2 home, etc.
We have also seen a lot of feedback on X/Twitter and on telegram in regards to GMC member COIs and pay. To the latter point, our initial proposal had the pay for GMC/TMC members at 30K ARB per member with a 2-year vesting period in an effort to achieve long-term alignment with appointed service providers while lowering overhead. However, numerous community members gave these payment terms backlash and demanded higher pay with no vests. With the implemented payment structure, no GMC or TMC members earn any of their earmarked payments until a recommendation has been passed via Snapshot (this is the first milestone for each member’s first $20k payment). In terms of COIs, Entropy has none. LlamaRisk is one (of two) paid risk providers for Aave, but holds no other COIs in relation to GM Track applicants. Callen from Wintermute in his personal capacity has no COIs however Wintermute holds AAVE and LDO tokens alongside being a public delegate for both DAOs. Lastly, an open governance call was hosted at the end of November where any community member had the opportunity to ask the appointed members questions or raise concerns. During this call, Callen (Wintermute) and River (Llama Risk) clearly state their wider involvement with other ecosystems, applications, CEXs, etc. thus disclosing their COIs prior to confirmation via Tally.
Lastly, we feel as though the DAO finally moving on treasury management is a good thing. We firmly believe in the GMC’s ability, with the DAO’s input, to get the right allocations in place that optimize for safe/sensible treasury management while providing ecosystem support as an added benefit. At the end of the day, this is the DAO’s treasury. The GMC is simply subservient to the DAO and looks forward to making any changes that are clearly consensus before moving to Snapshot. Let’s continue the good faith discussions on the forum and work through these qualms in a manner that supports the Arbitrum ecosystem/brand.
Next Steps:
Gather a better understanding of what the DAO believes the GM Track should be focusing on: Is it only supporting Arbitrum Native/Aligned protocols, and if so, how do we define “Arbitrum Native/Aligned”? Is the primary goal to build a strong foundation for a conservative treasury management program, and if so, do the original choices of the GMC reflect that? Should it be a mix of both direct builder support and a conservative treasury management foundation, and if so, how much of the 7,500 ETH should be allocated toward the former?
Please leave your feedback in the comments below, and we look forward to talking this over in Denver this week.
As mentioned, there will be no Snapshot vote until March 6th at the earliest, but it may be postponed further in order to alter "preferred choices” per the community’s feedback and to work with existing applicants to modify their applications accordingly.
Reminder:
This is just the first round of many treasury allocations to be made in the future. When a future proposal is made, we aim to put a heavier emphasis on Arbitrum native/aligned protocols. Also, as mentioned in the beginning of this post, we envision a world where all of the DAO’s revenue can be funneled to builders as a sustainable funding source.