Final: Arbitrum Stable Treasury Endowment Program

See the latest draft HERE

Note: This version is deprecated but kept for record purposes

This proposal has been drafted on behalf of the Arbitrum Treasury and Sustainability Working Group as a result of consultation between multiple stakeholders. It does not exclude treasury diversification proposals from outside of the proposed structure.

Title – Arbitrum’s Stable Treasury Endowment Program

Constitutional / Non-Constitutional - Non-Constitutional

Abstract - This is a framework for ArbitrumDAO to diversify 10 or 25 mn ARB into stable & liquid real world assets (treasury bills and other money market assets) that are launched natively on Arbitrum One. To maintain our position as the leader of De-Fi, we need to signal our support of the tokenized RWA projects launching on our chain. This proposal thus seeks to earn the risk free rate of return for a portion of our treasury assets and also support the budding RWA ecosystem on Arbitrum One.

The 10 or 25 million ARB is NOT a grant but an investment that the DAO can recall at any time

Motivation - The arbitrum treasury and sustainability working group was formed in September ‘23 with a mandate to discover best practices & propose best in class solutions for mitigating price impact of arb liquidations & diversifying arbitrums treasury, currently containing ~3.5 billion ARB

We seek comments on the Stable Treasury Endowment Program or STEP 1, a framework for converting 10 or 25 million ARB into tokenized real world assets (RWAs) , particularly t-bills and money market instruments, for earning yield. As this framework applies to highly liquid assets, the principal can be reclaimed by ArbitrumDAO at any time, with 3 days notice.

Proposals for utilizing interest earned from these investments (the Arbitrum Endowment Fund) can be presented to the DAO in separate proposals , giving us the capability to fund grant programs or retain staff without liquidating ARB.

STEP 1 excludes de-fi & other providers diversifying into cryptonative assets & caters to only those service providers with a tokenized treasury bill or money market RWA launched on Arbitrum; example service providers applying under the framework are included at the end. STEP 2 and STEP 3 will lay out a larger framework to expand treasury diversification to de-fi and crypto-assets.

Rationale - Why only RWA’s?

Ecosystem Growth : With STEP 1 we see an opportunity to support the new wave of financial products launching on Arbitrum that are tokenizing US treasuries & other stable RWAs. As the home of de-fi, we need to create a welcoming environment for what might be its evolution in the near term. The stable coins we use today in De-Fi are mostly backed by RWAs, so creating our own exposure can increase capital efficiency and mitigate some risk of a stable coin collapse or depeg.

Successful passage will also incentivize RWA projects to build on Arbitrum, helping maintain our TVL lead. For example, Mantle recently passed a $60 million support program for the tokenized RWAs on their chain.

Treasury Diversification: Diverse assets in our treasury can be called upon to defend the ARB price should the need ever arise. If the ARB price increases, we are in the green as it consists of 99%+ of treasury assets; if it goes down, we deploy these investments to buy-back ARB and send a positive signal to the market.

Future STEPs will look at diversifying into other crypto-native assets and De-Fi providers on Arbitrum; as a first step our WG recommends expanding into tokenized RWAs offering the risk free rate of return.

Specifications - STEP 1 gets the ball rolling with an experiment to diversify 10 or 25 million ARB into tokenized treasury / money market bills on Arbitrum to investment managers meeting the following requirements

  • To even be considered, the allocation should add to the TVL of ArbitrumDAO; accordingly, tokenized RWAs that are NOT on Arbitrum are ineligible

  • All assets need to be attested for verifiability by a 3rd party

  • A strong reputation with relevant team experience, evidence of managing assets and regulatory compliance

  • Fully liquid, secured and safe investments that can be called back by the DAO with 3 working days notice

  • Between 0.15-0.4% (15-40 bps) of total amount in fees with no bonuses, as service provider performance will be evaluated for renewal of contract.

  • Each service provider has enough votes equivalent to managing 100,000 ARB; after clearing this floor, service providers get a proportionate amount to the votes they have received.

STEP 1 can serve as the seed for the Arbitrum Endowment Fund, which spends interest earned from stable assets according to future DAO proposals ($20 million in t-bills gives roughly $1 million a year in interest).

Steps to Implement - Many delegates still have some trauma from reviewing 90+ proposals during STIP. It also yielded sub-optimal outcomes on approving a larger budget than that initially sanctioned. To overcome these issues, we propose

  • A snapshot vote approving either 10 or 25 million ARB to be diversified into tokenized t bills / money market instruments launched on Arbitrum

  • Service providers with such a product apply for consideration in a jokerace / snapshot, with our WG screening for eligibility and appeals handled by Plurality Labs.

  • Each delegate is given votes to allocate between service providers in proportion to their voting power; if they strongly believe in only one provider, they can allocate all their votes to them.

  • Service providers get to handle assets in proportion to the votes they have been allocated, provided they clear a floor of 100,000 ARB.


  • December 2023 - incorporate changes to the STEP framework from DAO delegates, community members and RWA service providers

  • January 2024 - release of the post consultation version followed by snapshot vote

  • February 2024 - onchain tally vote & applications from service providers

  • March 2024 - vote for service providers followed by KYC approvals , logistics

Investments can begin by April 2024

Overall Cost - The options put before the DAO for a snapshot vote are

  • Diversify 10 million ARB into tokenized, liquid and risk free rate of return RWAs on Arbitrum

  • Diversify 25 million ARB into tokenized, liquid and risk free rate of return RWAs on Arbitrum

  • Do not support the proposal

  • Abstain

As our Working Group receives remuneration from the Plurality Labs multi-sig (AIP-3), we do not estimate costs over and above what is needed for diversification

Frequently Asked Questions

  1. Why 10 or 25 million ARB?

We do not recommend more than 25 million ARB as this amount is ample for signaling our support to this nascent and emerging space. Moreover, RWA diversification beyond 25 million ARB should happen only after the DAO decides on a larger investment framework.

We know of at least 3-5 providers applying under this framework. 10 million ARB divided among them is sufficient for gaining credibility from ArbitrumDAO that they can use in getting other clients

  1. Won’t diversification affect my bags of $ARB?

STEP-1 will show we can take major decisions on prudent treasury management. When we start earning yield, we develop an endowment that can fund grant programs or employees without any sell pressure on ARB

  1. I get that, but in the short term how will you deal with the sell pressure from diversification?
  • Selected service providers will get a roster of OTC providers that buy the $ARB from them off the market, minimizing price impact.

  • Our working group also did a price impact analysis on the depth of $ARB limit orders in centralized exchanges to absorb liquidations without impact on price.

  • Moreover, a simple snapshot vote can demand that these investments be reconverted back to $ARB for defending its price.

  1. How do we evaluate performance of service providers?

We can quantify the cost effectiveness of each service provider based on risk-adjusted returns, comparing the yield they earn for us (net return) versus their risk profile. We can also benchmark performance by tracking to money market monitoring tools like morningstar

We also measure % of their holdings from ArbitrumDAO, which should go down over time as they start getting other clients.

  1. What do you personally gain from this framework passing or failing?

A major lesson from STIP was the importance of having delegates vote on frameworks rather than evaluate individual proposals not coming under any pre-approved framework. For example, only 20-30 applicants took part in STIP consultations but over 90 applied once the rules of the road had been laid down.

This means we need working groups who draft tenders for delegates to vote on and service providers to apply under. Our WG performs this role, paid from the AIP-3 Plurality Labs program

  1. Risks from passing this proposal
  • Regulatory issues around holding real world assets. In STEP-1, this is partially mitigated by using the Foundation for KYC checks and custody but it could still be confiscated (this risk also exists for USDC and is overall quite low).

  • The service provider is a centralized point of failure who may go bust and/or hide their actual holdings. Therefore service providers with onchain verifiability and sound legal structures to mitigate these risks are preferred, such as being bankruptcy remote.

  • The type of risks involved; such as counterparty (i.e. issuer), service providers and liquidity

  • Price of $ARB increases, causing a net loss from diversification

  • Lower returns compared to de-fi ; treasury assets only provide about 4% yield, compared to as high as 16% for USDC on compound. However these asset classes have different risk profiles and the focus in this proposal is risk free rate of return

The foundation will act as the legal entity and face the service provider during the KYC process. The foundation and our working group will have authority to withdraw assets from any service provider should any risks be discovered.

Overall, whether this proposal passes or fails, we hope this helps ArbitrumDAO start thinking about treasury diversification and how to maintain our TVL lead.


Here are 5 investment managers (in alphabetical order) who intend to apply under this framework

This section has been written by each of their teams and not our working group, which only consulted with them during creation of this RFP

  1. Backed Finance :

We are a Swiss-based company that has developed a compliant solution for issuing freely transferable tokenized versions of tradable securities, such as stocks, bonds, and ETFs. Assets are issued by our subsidiary, Backed Assets GmbH. This is a dedicated special-purpose vehicle that benefits from Switzerland’s innovation-friendly regulations, such as the Distributed Ledger Technology (DLT) Act, which allows the existence and transfer of digital asset securities on permissionless venues.

Key points about Backed and bTokens:

  • Backed offers tokenized structured products under Swiss law. bTokens are tracker certificates pegged to the underlying asset’s value and fully backed by it.

  • All underlying assets are held by regulated Swiss custodian banks.
    Backed integrated with Chainlink’s Proof of Reserve (PoR) to provide users with a transparent and trust-minimized means to confirm the collateralization of our tokenized assets.

  • Chainlink Price Feeds are available for Backed’s T-bill token.

  • Backed and its products are trusted by leading industry players such as Aragon, Gnosis, Morpho, Ribbon, and Angle.

Backed Finance natively supports Arbitrum since July 2023. We welcome such an initiative to grow TVL & activity for RWAs on Arbitrum.

  1. Centrifuge - Anemoy Liquid Treasury Fund

Centrifuge has been working with Offchain Labs since June to bring access to RWA yields to users directly on Arbitrum. We have also spent time with the Arbitrum Foundation and they have confirmed that they will be able to facilitate this proposal. We have approached our launch holistically as a way of developing Arbitrum’s RWA thesis and strategy - with the goal to both grow TVL on Arbitrum as well as to help capital on Arbitrum diversify in RWA yields. To accomplish this, we have deployed Centrifuge liquidity pools natively on Arbitrum, worked to educate the Arbitrum community on RWAs through webinars and research reports, and have also custom-designed a legal structure for Arbitrum to use for a safe and sustainable RWA strategy.

Centrifuge is natively deployed on Arbitrum, provides direct access to T-Bill yields in the most legally robust, battle-tested structure, and is able to help Arbitrum diversify into other RWAs in the long-term. The Centrifuge platform and ecosystem can support a variety of assets, from liquid asset funds to structured credit deals. Treasury Bills are just the first step in an RWA allocation, and Centrifuge wants to contribute to growing the Arbitrum treasury by providing access to higher yields across a diversified portfolio of RWA. Through Centrifuge Prime, a managed services offering that can be used as a contributor model in other communities, Centrifuge has deep experience working with DAOs such as Maker, Aave, and Gnosis, and has partnerships with managers and service providers such as Karpatkey and Steakhouse Financial, to bring a full-spectrum service offering to Arbitrum.

  1. Mountain Protocol

Mountain Protocol is a regulated and permissionless yield-bearing stablecoin.

The company is a regulated financial institution based in Bermuda (one of the most sophisticated digital asset regulators, regulating companies like Coinbase and Circle), under license #202302512 (

Most recently, USDM has been rated B+ by Bluechip, the same risk rating held by USDC.

USDM is fully backed by USDM Reserves composed of short-term (<60 day avg maturity) US Treasury assets, attested monthly by a 3rd party accounting firm.

The accounting firm also performs a full transactional audit, a first in the industry, ensuring that USDM Reserves are managed as laid out in the Terms and Conditions.

USDM also offers the best liquidity, with the ability to redeem USDM for USDC at scale in the primary market, powered by a partnership with Wintermute

USDM tokens are freely transferable, akin to other stablecoins like USDC/USDT, which can be tried it out by swapping into one of the liquidity venues: Curve 3pool/USDM (Ethereum), Curve sDAI/USDM (Ethereum), Curve USDC/USDM (Polygon POS).

Mountain Protocol’s products, including the USDM token, the platform or any other services, are not available to U.S. Persons. For more information, refer to Terms and Conditions.

Mountain Protocol is unique in that it’s the only permissionless asset that has demonstrated trading volume and the ability to grow holders and transfers. Despite being in the market for a fraction of the time and TVL of most RWAs, USDM has more holders and transfers than most other players.This is a requirement if Arbitrum expects for this move to be seeding the use of RWAs in the ecosystem.

For more information about Mountain Protocol and USDM, please refer to:
Security Center (powered by OpenZeppelin):
Token: 0x59d9356e565ab3a36dd77763fc0d87feaf85508c
Open source code: Mountain Protocol · GitHub

  1. Ondo Finance - Ondo US Dollar Yield Token (USDY)

Ondo Finance provides institutional-grade, blockchain-enabled investment products and services. Ondo is the market leader in tokenized securities with over $175 million assets. Ondo’s current live products include OUSG, the first tokenized US Treasuries product (for accredited investors), and USDY, the first freely transferable US Treasuries product (for retail and institutions alike). We also developed Flux Finance, the first lending protocol supporting tokenized securities as collateral.

For most applications in DeFi, USDY is a superior alternative to conventional stablecoins since it pays a yield. USDY can be integrated broadly into the Arbitrum ecosystem to offer protocols a competitive edge over protocols in ecosystems without liquid access to USDY.

Here are a few potential integrations for USDY:

  • Use as collateral in perps protocols, such as GMX: In TradFi, collateral for derivatives is often US Treasuries, rather than 0% yielding assets like conventional stablecoins. GMX and other perps protocols on Arbitrum can increase their competitive edge by adopting yield-bearing collateral.

  • Use as the “cash leg” in AMMs: One of the downsides of AMMs is their requirement to lock up large amounts of capital in liquidity pools. This downside is mitigated by utilizing yield-bearing collateral.

  • Use as collateral in lending protocols: Users wishing to borrow crypto on DeFi lending protocols can do so in a more capital efficient manner by posting USDY collateral. We have already passed a snapshot to add in fUSDC, an asset over-collateralized entirely by Ondo’s $OUSG (tokenized treasuries product), as collateral on Aave, and we are working on the technical integration. We also developed a lending protocol, Flux, which is exploring moving to Arbitrum.

As a brief technical note, there are two versions of USDY: one that is rebasing, with a mint/redeem price that stays at $1, and one that is accumulating, with a mint/redeem price that goes up by a small increment each day. Users will be able to swap between rebasing and accumulating versions via a simple contract call, very similar to swapping between ETH and stETH. We anticipate that most protocols will integrate the accumulating version given technical constraints, while most users holding USDY for savings purposes or using it as a means of payment will use the rebasing version given its accounting simplicity.

USDY is not yet deployed on Arbitrum, however we would deploy USDY natively on Arbitrum (rather than simply wrap it in a decentralized bridge token). We would integrate Arbitrum into our native asset bridge powered by Axelar.

Our tokenized cash equivalents bring the same 24/7 global transferability and compatibility with smart contracts and DeFi that stablecoins offer while paying their holders a yield and offering superior investor protections. The combination of high “real world” interest rates, low on-chain yields, a large quantity of capital in stablecoins, and regulatory and structural risks in stablecoins makes this a compelling immediate opportunity.

Project Links:

  1. OpenEden TBILL

OpenEden will be natively deployed on Arbitrum over the next few months. The firm has built the first TBILL Tokenized Vault that offers a DeFI user experience while offering the most institutional-grade direct access to RWA assets such as T-Bills. The OpenEden team is confident in bringing best-in-class on-chain RWA assets on Arbitrum because it has extensive commercial B2B partnerships with regulated institutions such as Standard Chartered Bank backed Zodia Custody, Siam Commercial Bank-backed Rakkar Digital and etc, and has worked with best-in-class service providers such as Ernst and Young, KMPG, London Stock Exchange Group (LSEG), StoneX alongside its sub-custodian BNP Paribas.

The tokenized RWA Vault to offer 24/7, instant and direct access to U.S. Treasury Bills (T-Bills). OpenEden’s “TBILL” tokens are issued by a BVI-regulated mutual fund established under the British Virgin Islands Securities and Investment Business Act 2010. The fund is managed by a registered fund management company regulated by the Monetary Authority of Singapore (MAS) with an investment mandate for short-dated US Treasury Bills. The portfolio of T-Bills are custodised in a regulated custodian in segregated accounts. The two key Portfolio Managers have collectively 30 years of experience in capital markets across large Investment Banks such as Goldman Sachs, Citibank and Morgan Stanley.

TBILL tokens can be minted instantly 24/7 on-chain via OpenEden’s first tokenized RWA Vault, the TBILL Vault ( upon KYC/whitelisting and held in the investor’s self-custodial wallet. OpenEden’s TBILL Vault is currently the market’s only platform that offers real-time, instant on-chain minting of tokens backed by T-Bills, thus removing settlement risks.

As the Token Issuer is a regulated mutual fund vehicle managed by a regulated fund manager, an independent fund administrator is appointed to (i) provide an independent valuation of the TBILL token’s NAV and (ii) administer multi-sig for all fund transfer transactions for investor protection. In addition to generating >5.3% yield, TBILL tokens will also help to hedge against stablecoin de-peg risk as underlying assets backing TBILL tokens are in USD-denominated T-Bills and excess liquidity is held in USD fiat. There is no lock-up for redemption thus providing liquidity if required.

The following are the key features of the TBILL Vault which sets it apart from its peers:

  1. Regulatory compliance: TBILL tokens are issued by a professional fund regulated by the BVI Financial Services Commission. Token holders’ contractual rights are documented in the Private Placement Memorandum and Subscription Agreement. Furthermore, there are independent service providers engaged by the fund such as Fund Administrator, Fund Auditor, Legal Counsels, Regulated Custodians and prime brokers. The underlying assets of the TBILL Vault are held via a bankruptcy-remote special purpose vehicle, with fully segregated accounts in qualified custodians which are regulated. Assets held in segregated accounts cannot be used by the custodian to pay its own debt or obligations by regulation, and it must keep those assets separate to protect client interests. The portfolio of T-Bills are managed by a regulated investment manager in Singapore. TBILL tokens are available to all accredited investors, including to U.S Accredited Investors through SEC Reg D 506c.

  2. Instant settlement on-chain: Interactions with the TBILL Vault are atomic. Subscription processes happen instantaneously in a DeFi-like transaction on-chain through on-chain price oracle providing live NAV of TBILL token price. By remaining aligned with how DeFi transactions happen, the TBILL Vault product offers familiarity and ease-of-use to DeFi-native investors. Redemptions can be same day settlement and latest T+ 1 business day settlement.

  3. Highly transparent: Transparency is at the core of the TBILL Vault. The Vault smart contract has been formally audited by Hacken. On the off-chain side, daily portfolio reports from the T-Bills’ custodian, monthly attestations of the BVI fund’s (Token Issuer) balance sheet conducted by a regulated public accounting corporation and monthly NAV report by the Fund Administrator are all uploaded to the portal for investors to view. A Chainlink Proof-of-Reserves for real-time verification is also currently being integrated into the Vault to provide another reliable source of on-chain and off-chain audit of the Vault’s assets.

  4. Direct T-Bills exposure without intermediaries: The underlying assets of the TBILL Vault consist mainly of T-Bills that are purchased directly through a regulated and qualified prime broker. This approach offers holders of the TBILL tokens direct exposure to T-Bills, as opposed to ETFs. There are several benefits of having direct exposure to T-Bills. Investors would not incur additional charges like ETF management fees. There will also be no tracking error that may arise from ETFs. These additional costs from ETFs will reduce the net yield received by investors. Furthermore, direct T-Bills exposure means having access to the highly-liquid T-Bills secondary market, which has an average daily volume of $150 to $200 billion as compared to Bond ETFs which trades less than $400 million a day.

  5. Validation by Regulated Institutional Clients: OpenEden’s TBILL token has been live since May and emerged from beta in November. Since then, several large regulated digital assets custodians and platforms had established partnerships with OpenEden to offer TBILL tokens to their own users. One example is Standard Chartered Bank-owned Zodia Custody who after months of due diligence selected OpenEden’s TBILL as the only tokenised T-Bill product to be offered to its own institutional clients. Standard Chartered-backed Zodia Custody to offer yield on crypto holdings | The Block

More information on the TBILL Vault can be found at the documentation site.


Hi, Parv from Estate Protocol here. I’m generally in agreement with this proposal, but I do not believe the foundation should be picking and choosing the beneficiaries and winners arbitrarily, and instead come up with some criteria to qualify.

Also, the scope should expand beyond just tbills, as there’s not much innovation that can happen there, so the signalling value is not high. At the same time, interest rates are widely expected to start coming down in 2024, so the benefit to the foundation treasury would be limited.

Avalanche Vista is a program we should take inspiration from. High quality assets that may or may not be liquid, but the foundation helps create liquidity as long as the quality of assets is high.

Correction: I now understand the DAO would vote on the providers, so no arbitrary winners.

1 Like

We will comment in more detail later, but we want to commend the working group for getting this framework together. GFX has had the benefit/frustration of seeing different implementations up close for similar plans elsewhere.

At first glance, this framework looks good. We would suggest a couple minor changes. The fee offered is above market and unnecessarily high. A cap of even 20 bps would be above what is required to attract applicants.

Governance should have veto power over the third party chosen for attestation. Even better, a dozen or so community members should be given direct, read-only access to the accounts. There are no trade secrets to protect with a tbill strategy. If using only attestations, have a grace period and hard line that automatically triggers unwinding without a DAO vote if they are tardy.

Specify up front a schedule for returning interest on-chain. Include a grace period and hard line that terminates the agreement and triggers unwinding (e.g. three strikes and you’re out).

Specify up front a narrow investment mandate. This could be no maturities beyond a certain threshold, and should also explicitly state if the manager has discretion about how to weight the portfolio or if it’s strictly an even-weighted ladder.

Specify up front who bears unexpected expenses directly related to managing Arbitrum’s money. An example would be not planning for a tax to move money through a jurisdiction.

Specify up front whether the manager has a custodial relationship with Arbitrum, if Arbitrum is merely a creditor, or any other arrangement.

Three business days seems excessive. Whatever the time frame, hold at least one unannounced fire drill. It is worth any minor costs associated with liquidating some tbills to find out soon if the manager is blocked. It will also flush out some AML delays if the transaction is flagged.

Obsess over the legal and financial terms. This requires Arbitrum to get its own legal review of the structure, governance’s control over the structure, etc. This is traditionally a major failing when DAOs look at tbills.

Require a real, actual audit at six months and annually thereafter. The DAO should pay for this so the auditor is beholden to the DAO.

The managers should have sharply delineated ways to move funds into and out of the vehicle. If funds travel in an unexpected manner despite contractual obligations, there must be a full postmortem within 30 days or the investment should be unwound. This is a common weakness at some tbill arrangements.

Make this process open to other providers. There are high quality financial professionals that would bid on this. Having a token shouldn’t be a requirement, since the key goal is tbill management, not building a market for a token. If a token is needed, it’s easier to find experienced financial institutions and help them make a token than it is to take developers and get them up to par with a financial institution.

We will have more detailed thoughts as the proposal and bidding process gets underway. This is an exciting opportunity for Arbitrum, and we strongly want to contribute to this process so Arbitrum can learn from the past success and mistakes elsewhere


This is great work. The time and attention to detail shows. Thank you.

At first, I didn’t understand why this was scoped down to just a tbill strategy. Once I discussed it with @thedevanshmehta I realized that this is a first step. It makes sense to try and pass one small strategy at a time and then get bigger as opposed to going for one giant omnibus treasury plan vote.

I also realize that starting here is acutely intentional as there is the ability to signal to all of defi that Arbitrum is supporting RWA protocols.

I kinda agree here, however, talking with Devansh made me realize that there is intention behind the smaller scope.

This whole post is great feedback. Thanks for your time and attention.


Hi ParvEP,

Just to clarify, the Arbitrum Foundation will not be in a position to pick and choose the beneficiaries / winners. That is up to the ArbitrumDAO / whoever is running this program. As mentioned in the proposal, we will take on tasks to operationally support and enable the proposal.


Hi @thedevanshmehta!
I think it is a good idea to make money work.
However, there are several questions:

  1. How will the income be returned back to the DAO? Will the income be sent to the treasury, will the ARB be redeemed with the received stablecoins from the market? Or another option?
  2. In what parts and within what time frame will the profit be returned? Month, year or what?

Really great notes @GFXlabs


Thanks for putting this proposal together @thedevanshmehta

I love the idea of creating an endowment to fund grant operations. To be clear, the goal here to sell $20m arb for stables and deploy those stables into in t-bills / MMFs to generate $1m a year in stables that can fund grants/ grant ops in perpetuity, yes?

I think as a standalone experiment in creating an endowment it’s super interesting. I do think the framing of it as a diversification of the treasury muddies things up a bit since the overall diversification and “price support” mentioned is relatively small to the overall treasury size.

That being said, it seems super interesting and I love the inclusion of some awesome Arbitrum native investment managers. Not sure if their input has been added yet, but I know Karpatkey has done work with ENS for their endowment. I’d be curious what their thoughts are on this.


Thanks Pat, I’ve added the correction to my original comment.


Thanks for all the feedback and discussion, really appreciate the critical comments.

Our WG is responsible for defining eligibility requirements and approving investment managers that can participate. After that, it is up to the delegates for allocating to the investment managers they think are competent. For example, if 10 million ARB is approved and a total of 100 million ARB is voted with in the jokerace, each vote gives a provider 0.1 ARB to manage.

We definitely do not want a council, working group or any centralized body to select providers as that is too easily corruptible and more generally against the spirit of DAOs and decentralization.

Any large company usually gets tbills on its balance sheet before venturing into other asset classes. Also worth noting that Maker, Circle and any stablecoin innovation is tbill backed, so I would pushback on no innovation happening with these instruments. In particular, my WG co-lead @sids2000 is cooking up some interesting proposals related to CDPs and collateralization with tbills & other stable assets to prevent grantees from having to liquidate the ARB they receive.

It is possible to get a lock-in at the prevailing interest rate, right? I would again go back to the fact that ArbitrumDAO should have a nonzero amount of t-bills on its books, with the exact amount being higher or lower depending on the interest rate at the time.

Would love to have your participation in STEP 2 , which would look at a more holistic vision of diversification. You can join in the discussions here -

we expect outputs from our rapid research grants to Aera, Avant Garde and Karpatkey to be delivered by January 15th, after which we have a base to discuss what larger treasury diversification for Arbitrum looks like.

1 Like

I like your suggestions on read-only access for community members, an auditor hired by the DAO to verify holdings, triggering unwinding without a DAO vote if there is a failure on 3rd party attestations or returning interest earned on time, holding fire drills withdrawing assets to clear out delays so we are safer if a crisis does actually arise and flagging unusual movement of funds.

I kept this ceiling based on discussions with , which charges 20 bps for entry and 20 bps for exit or else 35 bps AUM per year.

Would we want our WG to deny eligibility to any provider over 20 bps, or keep a higher ceiling & let delegates give less votes to the more expensive providers ?

Same question on your concern with 3 business days being excessive, is the role of a RFP to keep a moderately high end and the role of delegates to vote for the providers that give the best offer?

I’d also like more details on this - would governance here mean delegates, our WG or the Foundation ?

Thankfully the Foundation has agreed to play a slightly more hands-on role compared to usual, so we can use their lawyers and due diligence teams to strike out providers that are not bankruptcy remote (which ensures assets are ours even if the provider is out of business) or have poor governance structures, even if they have got an allocation from delegates.

We can also unilaterally liquidate positions from an investment manager if we see any cause for concern; we would of course post an explanation on the forum if we ever invoked this power.

That’s a good point, the framework is lacking the maturity period of tbills under consideration.

My initial instinct is to disallow t-bills of more than 6 month duration, what are your thoughts here?

Such a great point, we can always assist financial institutions to launch on Arbitrum and its easier to do that than work with developers on the right financial safeguards.

My initial thinking was keeping it restricted to those launched on Arbitrum to increase our TVL, add an incentive for RWA to launch on our chain and also narrow the pool of eligible providers to make it easier for delegates to select and vote.

If we open it up more generally, I do worry we will see a flood of applications and land up in a situation similar to STIP, with delegates having to review a ton of providers.

My 2 questions are, how do we restrict the pool of providers delegates can vote on to a manageable number and how many providers do we ideally want to work with?

Really appreciate the meaty comments, I can already see some major changes required to this draft before putting it up for any sort of vote.


This is a good question; utilization of interest earned feels like it would require a separate vote as there are divergent and valid views here.

The options are; set it up as an AMM to buyback ARB from interest earned; fund grant programs or personnel in a more nimble manner than proposals before the treasury which usually require frameworks ; let it compound

Keen to hear more ideas on how we can employ interest earned, we would likely need a separate vote for consensus on this question.

it would be returned upon maturity of the t-bill, the maximum allowable of which is 6 months. @GFXlabs had some great comments to make in this regard, on unwinding all our assets from a provider if they do not return interest on time.


We think this is very high. However, if this is a competitive process, then there’s probably no harm in allowing them to pitch it. 10-20 bps is likely to emerge as the competitive rate. So perhaps it’s not worth denying eligibility on the speculation that other providers will be much lower. If they are, nothing is lost by letting one more applicant apply.

Same response as above. Probably no need to exclude, upon reflection. 1-2 business days are likely to emerge as the winners anyway.

Excellent question. Probably one or more of the latter two? Delegates probably don’t want to deal with evaluating an accounting firm.

If the main objective is to maintain ready access to the funds, shorter durations are better. It also has the benefit of limiting the investment mandate. It’s easier to expand this to longer maturities or other assets than it is to shrink it. Given this is an initial trial, a max duration of not more than a year would allow for unwinding and also lower exposure to interest rate risk (if rates rise, longer dated bills and notes would be more affected in market price). We do not have specific advice on where the max maturity should be, other than “less than a year” allows a maximum of a year to get all the money back if selling on the market would create a loss. It also limits the skills required to successfully execute upon an investment mandate. You want a strategy any dummy can run, because sooner or later, any dummy probably will.

This is an open question. Perhaps require it to be an entity that is already in business (so no start ups without a track record)? Or require a de minimis deposit of ARB that will be returned to the applicant after evaluation? Another option is to have a very detailed RFP that simply can’t be answered in 5 minutes by a rando on the internet. Incomplete applications can then be excluded.

There could be value in stating that up to two providers will be selected. If one performs poorly or violates terms, assets can be returned and routed to the other provider, preventing interruption of the program. From experience, it may also be that having two providers provides some assurance that money can be recalled in a timely manner even if one structure gets funds stuck due to a bank, brokerage, or manager error. We would recommend that the two providers not be carbon copies of each other – much of the value in having two providers would be lost if they both utilized the same banking and custodial partners.


Kudos to everyone who contributed to drafting this proposal. It’s a promising step forward.

We think the DAO would benefit from starting with broader and overarching decisions before delving into specifics. A robust treasury framework can be established by first reaching a consensus on the DAO’s long-term strategy. This framework would guide every financial decision moving forward, clarifying roles, investment horizons, risk tolerance, capital requirements, etc.

The ENS Endowment is a testament to the success of this execution. Its goal to sustain the ENS DAO indefinitely dictates long investment horizons and a conservative risk approach. The Endowment supports the DAO by covering expenses and generating extra resources to advance ENS as a public good. We think a similar strategy would benefit the Arbitrum DAO before making specific funding or development decisions within the ecosystem.

Nevertheless, as @thedevanshmehta pointed out, this STEP 1 is a test run, helping the DAO to break the ice and start moving in the right direction. Beyond financial gains, this proposal promises valuable experience and insights for those executing it, leading to better decision-making. We’re all for this approach.

The proposal has already garnered insightful feedback. We’d like to additionally suggest the following:

  1. Clarify requirements to prevent an influx of misaligned proposals. For instance, specify viable treasury maturities and the preferred methods for exposure to selected instruments. Are options like ETFs (e.g., IB01), direct purchases, or lending to accredited investors equally viable, or are there limitations?
  2. Set explicit criteria for provider applications. Clear benchmarks (like TVL) should be established to attract top-tier providers with proven track records. This also helps streamline the review process for delegates.
  3. Define success and failure for providers. Establish clear KPIs and outline steps for discontinuing underperforming providers.
  4. Lay out precise reporting guidelines. This ensures effective monitoring of investments and maintains transparency throughout the process.

We at Centrifuge really commend this framework approach and how quickly the working group was able to incorporate feedback into this proposal. Managing the treasury is a journey, and we see this proposal as an important first step.

We think this approach helps the Arbitrum DAO cut down on noise as well as start to think of its treasury diversification as a holistic strategy that can have service providers - like Centrifuge - come in and support an approach that the community agrees on. This makes it easier for the DAO to streamline its decision making process, letting Arb delegates focus on the bigger picture decisions without needing to micromanage the details, and for teams like Centrifuge to deliver on the needs of Arbitrum DAO.

As the leading RWA project in the industry - and having now officially launched on Arbitrum - we are excited to propose Centrifuge as a legally robust and battle-tested option natively on Arb for Arbitrum’s Stable Treasury Endowment Program to provide its treasury access to reliable yield if this initial proposal passes.

We think Centrifuge Prime is a fantastic option for this proposal, and we will be running as a candidate to be a service provider for this STEP program. Centrifuge Prime is a flexible product designed for ease of use by DAOs, allowing this allocation to evolve and take advantage of the best opportunities in RWAs as things change over time. We look forward to questions from the community and upcoming next steps should this proposal pass.

Other Benefits to Consider:

  • Monitoring is easy.The Centrifuge Credit Group, a DAO entity setup to provide risk and credit analysis, can be utilized to provide general-purpose reporting for the assets invested. This provides an additional level of ongoing monitoring and independent analysis that Arbitrum can utilize as part of its risk-management framework

  • Other investment options beyond US treasuries; the Centrifuge Prime setup would also allow access to the following asset classes:

Bucket Focus Description Target Returns Liquidity Thresholds
Liquidity Lowest risk while maintaining US Treasuries, MMF, AA+ rated short term bonds 5-6% Daily
Preservation Investments in senior secured loans such as real estate, corporate bonds etc. 6-8% 3-6 months
Acceleration Trade finance, non-bank originators, emerging markets 8-20% 3-24mo
  • Diversifying responsibility beyond the Foundation and making legal recourse available; KYC can be done through an established legal conduit set up exclusively for Arbitrum. Centrifuge can help Arbitrum DAO establish this legal structure in the future and define rules and procedures that only allow the legal conduit to receive and follow orders from a DAO governance process. The legal entity does not represent the DAO generally but simply serves as a tool that the DAO can use to engage with the traditional legal system of RWA markets.

A few additional thoughts/feedback:

In deciding on the amount to allocate:
We recommend starting with a $25M allocation to be able to allocate at least $10M to each of the two chosen service providers. That amount enables capital deployment in a more efficient way, considering possible up front costs from each provider and the time it would take to recoup the initial costs. Generally, the larger the investment, the faster Arbitrum will be able to diversify the treasury and start capturing yield on idle stablecoins. Since US Treasury Bills are considered the lowest-risk asset class, our recommendation is for the Arbitrum DAO to maximize this opportunity. As well, given the high liquidity of $ARB it does not raise concerns for the effect of a $25M allocation on the price of $ARB.

On the feedback in comments to not restrict the pool of potential service providers to only those natively on Arbitrum:
Generally we would agree the Treasury should look for the best strategy regardless of where it is deployed - and we do think Centrifuge is the best option (though we did just launch natively on Arbitrum!) - but there should also be an ecosystem incentive as well for projects to launch natively on Arbitrum and this type of framework is a great way to push that forward. So we see preference for projects that have launched on Arb to be ideal but long term this doesn’t need to be a strict rule. However, for this first proposal given the great options available that are already natively on Arb, we see no strong need to move beyond the options that already exist natively.

Take a look at some of our resources and proposals at other DAOs:


Thanks for your explanation!
However, I believe that the purpose/distribution of the resulting profits is the essence of this proposal. The goal of this whole idea is to try to make money on locked tokens, which, in fact, idle and do not work.
Therefore, I consider it wrong to postpone the process of profit distribution to another proposal.

  1. I believe that profit should be useful for all Arbitrum users.
  2. Therefore, I think that issuing grants with this profit is incorrect, since the community cannot manage these funds and bad decisions can be made without the participation of ordinary users. There are separate proposals for grants and a large treasury for this.
  3. As an option, I ask you to consider buying back the ARB from the market. This option will support the rate of the ARB token, which is held by users who vote, among other things, for this proposal.
    Any other proposals aimed at the general benefit are also considered.

Thank you for all the detailed comments. I see several additional modifications to STEP 1, namely

  • Creation of a screening committee to restrict entry into the selection round to compliant, low risk providers only.

STEP is unique in that the failure of even one provider means our program has failed, as the DAO has lost money when the whole purpose is earning it. Rule number 1 and 2 is to NOT lose money because of this program.

A heavy-handed committee screening out untested, higher risk applicants ensures that assets are protected whoever the delegates end up voting for; this also addresses @karpatkey’s feedback on streamlining the review process for delegates and attracting top-tier providers only.

  • Hiring only 2 or at most 3 providers for managing assets

Related to the above point, minimising our attack window by keeping the number of providers low. Selection procedure via ranked choice voting on snapshot will suffice; weighted voting is more appropriate when there are a high number of providers selected in a program.

  • Streaming interest earned back into the treasury to increase our USDC holdings

We agree with @cp0x comments on interest earned benefiting all Arbitrum users; accordingly the path of least resistance is simply transferring interest earned back to the treasury where the DAO can decide on its allocation following the usual process.

This can also help with @karpatkey’s suggestion on KPIs for providers (amount & promptness in returning interest earned to the treasury)

As a major prong of this proposal is asset diversification, we think it makes more sense for interest earned to increase the treasury’s USDC holdings (currently at 10 cents). This does not preclude the DAO from converting its USDC holdings into ARB at any point of time.

  • restricting this RFP to assets of not more than 1 year duration

We appreciate @GFXlabs specific feedback on restricting to assets of at most 1 year maturity period and think this is an important addition to the RFP. This also partially addresses @karpatkey’s suggestion on better clarifying requirements.

  • keeping the requirement of having launched on Arbitrum as giving bonus points rather than a hard requirement

Thanks to @cassidy_centrifuge for weighing in on this question. We don’t as yet know the quality of providers that will apply; accordingly we can keep the RFP more open and leave it up to the delegates over whether to give brownie points to those that have launched on Arbitrum

Thank you to all who engaged in the discussions for improving STEP ; we will keep the current version open to feedback for 1-2 weeks more before releasing an updated version after the new year.


I think this set of responses is an example of how firestarters grantees should approach their efforts. They need to be inquisitive. They drive the conversation forward, aggregate the results, hold context for the rest of us, and clearly show why they are making the decisions they are in how the proposal is drafted.

It’s a fine line between driving the conversation forward and making nuanced decisions so we don’t get stuck voting 15 times in the drafting of a proposal vs bulldozing forward without regard for critical thinking.


Thanks Joe for the kind response !

This is possible in large part because our WG funding comes from the Plurality Labs multisig and is not tied to the passage of this proposal. That lets us draft neutral frameworks which are interesting for the DAO to consider without a vested financial interest in getting it passed.

More generally I also think fixed pay allows for more creative work compared to retroactive funding as your wages/survival is de-linked from achieving any short term outcome