[Poll] RWA Working Group - Centrifuge Model

Following the signal request to prioritize RWA (which will most likely be accepted on chain). The RWA Working Group have work on a model to onboard RWA, called the Centrifuge model because most of the work comes from Centrifuge with minor changes and clarifications. Centrifuge MIPs remain the truth regarding Centrifuge proposals.

Do you agree on the following model to implement RWA?

Please notice that it may not be the only model and it may not be the perfect model. If the community approve this model, it will be worked more formally with all domain teams. If the community doesn’t like this model, it might delay significantly the RWA onboarding (there are other proposals like MIP13c3-SP4 Declaration of Intent & Commercial Points - Off-Chain Asset Backed Lender to onboard Real World Assets as Collateral for a DAI loan that will be pursued in parallel in any case).

  • yes, I agree on this model
  • no, I disagree with this model
  • unsure
  • abstain

0 voters

This model was designed with the following assumptions:

  • Speed of implementation
  • No legal entity needed (a separate thread will be created)
  • The DAI hedging issue is fixed (see this thread)
  • Risk assessment will have it’s own thread. It’s mainly the process here.

Generic Centrifuge Model

The generic Centrifuge model provides an overview on how to use real world secured loans as collateral in a Maker vault to generate DAI. It is not limited to Centrifuge assets and doesn’t have to be used by all Centrifuge assets. All parameters here are only illustrative.

SPV - Special Purpose Vehicle

The SPV holds the loan made to real world actors (business) secured by real world assets (invoices, properties, …). For our purpose let’s pretend the SPV has a target of $10M loans splitted between separate 100 loans. Those loans have an interest rate of 10%. The SPV only holds assets, it doesn’t have an operating business.

To finance those loans, the SPV has a junior tranche (equity), TIN token on the blockchain, provided by the asset originator or other investors. For this example the junior tranche is $2M. The remaining needed capital is obtained by emitting debt in the form of DROP token for $8M. The DROP token accrues interest (capitalized) of 7% per year (DROP value is 100 at launch, 107 one year after, 114.49 after two years, …).


Investors can subscribe for DROP tokens on the Centrifuge website after a KYC process with a minimum amount of $10k. They can ask for redemption any time.

Maker Vault

Deposit and minting is limited to one whitelisted address, the SPV.

Liquidation ratio is 100% on the DROP that sits in the vault. The stability fees (SF) of the vault will be the same as the DROP token. So if DROP equals 100 DAI, the SPV can mint 100 DAI. After a year the vault debt will be 107 DAI and the DROP value will be 107 DAI.

Actual liquidations will be triggered on a negative risk domain team analysis on covenants defined during the onboarding process.

For instance, the SPV NAV (Net Asset Value) will have to stay above 110% of DROP amount. Also, no underlying borrower should represent more than 5% of the SPV assets.

MakerDAO governance will vote for a Target Debt Ceiling (DC) of 10M DAI. The risk team will set the initial DC at 1M (that would be the actual DC in the smart contract).

DAI Minting

At any time, the SPV can put some DROP tokens in the vault and mint DAI up to the debt ceiling (DC).

DC Increase

As specified, the vault has a target DC of 10M allowed by MakerDAO governance. Nevertheless, the risk team has set the actual DC at 1M. At any time, the SPV can ask the risk team to increase the DC. This will trigger a covenant breach analysis. If the result is good the risk team will use its mandate to increase the actual DC in the next executive vote (after a forum notification to MakerDAO governance). The DC is still limited by the target DC approved by MakerDAO governance.


At any time the SPV can redeem locked DROP tokens with DAI. The risk team might decrease the actual DC (again, forum notice then bundled in the next executive). That would lead to a covenant breach analysis before raising the DC again.


If during any risk assessment the covenants are not satisfied (at specified events and every month/quarter), the risk domain will trigger a liquidation event. Liquidation would follow MIP22 which is asking for redemptions (using future cash flows of the underlying loans as loans have a short maturity).

MakerDAO should also be able to trigger a redemption demand (to avoid being stuck in perpetuity). This would set the DC to 0 and liquidate with a 6-12 months notice. This should be done only for strategic reasons or if the overall environment is changing.

Legal Contract enforcement

Maker doesn’t have any legally binding contract with the SPV. Nevertheless, other co-investors in DROP tokens have legally binding agreements with the SPV. Some of those coinvestors will be Maker representatives who will enforce the contract (possibly with Maker contributing to legal expenses) if the SPV doesn’t comply with the contract.


I am totally behind the Centrifuge Model. @spin @_LS and the Centrifuge Team have put in a lot of hard work and dedication into making their protocol come into fruition. They have been passionate and diligent with regards to helping the Community understand the Trade Finance ecosystem and how the Centrifuge Model will work. I truly believe this will open up a new window into systemic innovation of RWA in the Maker ecosystem.

So, with that in mind–let’s get it! :slightly_smiling_face:

“Life can be much broader, once you discovered one simple fact—and that is—everything around you—that you call life—was made up by people that were no smarter than you… and you can change the world, you can influence it. You can build things that other people can use. Once you learn that one simple fact… you’ll never be the same again…" —Unknown


Is this proposal in line with what earlier has been proposed?
If there are any changes, could you please highlight these?

1 Like

It’s in line with Centrifuge proposal. We have discussed it with Centrifuge team on Tuesday. Maybe there is some minor changes but none I can remember. It’s more some additions/synthesis because we looked at it in different way.

What is new (or more detailed):

  • Maker representatives
  • DC management by the risk team
  • liquidation triggered by Maker governance with a notice.
  • more details on the vault parameters
  • exchange rate issue (not in this post)

I am voting yes because I believe that Maker should host a diversity of structures and models. In my opinion, the risk of one model over another should be handed with debt ceilings and not by creating a barrier to entry for onboarding. For example, if “Model A” seems to work at a small scale but not at a large scale, the community can demand changes before raising the debt ceiling on assets following that model type. For the time being, I think it’s important that we keep barriers to entry as low as possible and use the other levers in the system to manage the risk of various legal structures.


Completely agree with this. It’s too early for us to arm-chair general what works and what doesn’t. Let’s take a shotgun approach with reasonable debt ceilings to start and scale up from there as we learn more.


I just wanted to point out that Seb is hosting a RWA Collateral Onboarding Call today that will discuss some of these topics, if you’ve participated in the poll or would like to learn, join the call:

1 Like

Will this be recorded? I can’t make it but would love to watch later.

Yes, it has been, soon on Youtube I guess.


I’m wondering if real estate assets should be treated in a separate SPV, with a slightly modified model. As a general framework, the Centrifuge model seems workable. It would seem harder to administer, but separate redemption/liquidation/collateral levels for the underlying assets would make more sense for real estate.

@Philinje, each asset originator creates their own SPV, meaning there will be one SPV per pool/collateral type. For example, ConsolFreight (trade finance txs) created their own SPV that administers their deployment. The same is true for New Silver (real estate backed loans). Each collateral type will have its own specified risk parameters.

1 Like

Closing the poll as there is strong support for this model.

This topic was automatically closed 91 days after the last reply. New replies are no longer allowed.