Real-World Finance Core Unit Report - 2021-03

March 2021 reporting for Real-World Finance Core Unit .


MIP22/New-silver onboarding progress

We have a weekly 2-hours slot with Smart Contract team and Centrifuge to move forward on MIP22. Thanks to the work done by everyone, New Silver is now on Kovan. MIP22 was transformed to use MIP21. We have a final due diligence meeting with New Silver next Tuesday. They have already $3M of assets and are financing 10 loans.
More information will be shared next week.

No need to say how exciting it is. The first RWA investment by MakerDAO can happen in April.

Others collaterals

There is still a lot of action on other collaterals. Here is just a sample of some.

Consol Freight might follow soon New Silver (after checking that everything works and a bit of Lindy effect).

Harbor Trade Credit risk assessment should be published in April.

We had some calls with SolarX and some back-and-forth with them. It follows the trust model and we hope to onboard this one in June. Many parties are involved (landowners, equity investors, utilities) which might require strong coordination on this one.

Discussions with People’s Company are in progress to find the best way to structure the deal.

Finally, good progress on FortunaFi which is a great opportunity to diversify.

Here is the full list of what is worked on and public.

We will work on a timeline for RWA onboarding. Internally first, then see how it could be incorporated by others teams and the Collateral Onboarding Sheet.


We have found a law firm that will be able to help us analyze legal structures. Their knowledge will be helpful when we will analyze the trust model legal documents.

Additionally, for the SPV models, we had a discussion with an SPV manager that can provide independent directors services to harden security while staying cost-effective. This increases the bankruptcy remoteness of the SPV. The independent manager will also search for a backup service should the original asset originator default. This will be mandatory for any sizable investment (meaning above a few million DAI).

Research - Crypto banking 101

This paper presents our findings on how the banking framework applies to crypto.


More precisely, it draws a link between our solvency (surplus buffer) and the deposit interest rate we might need to have (DSR).

Budget dashboard

There is now a budget dashboard so you can see the link between core unit expenses, the impact on the Surplus Buffer, and MKR dilution. The MKR framework working group will provide input to have something more precise but while waiting you can make simulations.

@Aes stepped up to continue working on this topic. Exciting to see that we have experts in all fields in the community.

Data infrastructure

Coming from a data science background (and setting up data infrastructures) I put high importance on having a good data environment. For instance, in the current financials presentation I make use of Dune Analytics, but maintaining queries are painful, not scalable, and limited (it’s a great tool, but not ideal to be the data warehouse of a $3B market cap organization). It takes me hours of error-prone copy-pasting for something that should be one refresh click on Excel.

On another topic, one important question is how many DAI withdrawals can we get if farming activities are stopping. How much of the DAI usage is non-speculative?

We spent the month discussing with a lot of possible providers. There are actually way more solutions than I was expecting but the space is still super early. There is also plenty of question on what do we want to achieve. For instance, we might want to have ownership to not depend on third parties that can break your work or disappear.

Obviously, if we want something good, there will be a cost.

In summary, super early, but very good discussions. If this is an area that interests you, or if you plan to submit a Data Core Unit, feel free to reach out. My goal is to have something to show for the end of the year.

Public work

Team composition

@SebVentures - Facilitator
@williamr - Full-time contributor
@Philinje - Part-time contributor
@jameskmccall - Community member/expert for farmlands
@christiancdpetersen - Community member/expert for energy-related project finance

Previous reportings


I don’t think it’s fair to say that Dai is fractional reserve based on the USDC in the system - this would be making a quality assumption about USDC and implying that it’s higher quality money than Dai. I would compare the USDC in the system to assets held by the Fed, since Maker can execute its own monetary policy.

To highlight this point, you can’t have a “bank run” on MakerDAO. If the Protocol discards all of its USDC, it doesn’t mean that Maker is insolvent, it’s just sold some assets in order to reduce its liabilities.


It’s not a quality assumption of the entity but a qualitative assumption on the path to go back to the layer 1 money. Actually, if USDC goes bust, DAI is in a very bad situation, the opposite is not true. We should be able to manage that at some point with many PSM to diversify the counterparty risk and make each counterparty risk lower than the Surplus Buffer.

As said elsewhere, for PR we can remove the USDC part and create an internal DAI-FIAT with Paxos so we move one layer up. But that doesn’t change anything significant.

Like any bank. They are currently limited by regulation but that wasn’t always the case.

Not insolvent but lacking liquidity. Actually, we don’t make any promise on DAI besides ES. But it is my feeling that such promise is not enough in a mature and competitive market.

1 Like

I think this is a super-interesting discussion. I am not competent in this field, so mu comments below are just to try to understand.

As we discussed in the and also privately, I am also confused by your take (also of your Crypto Banking 101. This paper describes the business of… | by Sébastien Derivaux | Mar, 2021 | Medium) that MakerDAO is a fractional reserve based on USD (and it’s n-th layer derivatives like USDC).

On the one hand in your crypto-banking-101 you say that MakerDAO’s reserves are 22% (USDC in the PSM) but at the same time:

It seems we are in a paradox here, because if we had (or we go back to) 0% in PSM-USDC, we would be not exposed anymore to this risk.

A source of confusion for me is:

Usual banks borrow (from customers, from other banks, etc). We don’t. We lend and we print DAI. So I am (naively) tempted to agree with what @g_dip said.

Some vague questions:

  1. What about MakerDAO in its SCD version (just ETH)? It still had most characteristics of today’s MCD, but had 0% in USDC derivates and 100% in ETH (a layer-1 asset).

  2. Was SCD a fractional reserve with 0% reserves, in your point of view?

  3. Why people perceived DAI as a good alternative to USDT then? They had somewhat faith in our ‘reserves’.

One important point, that @SebVentures made me appreciate and that I repeat here in case it’s useful to somebody else is:

It is wrong to say that MakerDAO has crypto reserves (ETH, WBTC, etc). The reserves of MakerDAO are of two types:

  1. Loans (all the crypto vaults).
  2. Some assets (mostly USDC in PSM) → a recent addition (Nov 2020).

For (1), the loans are a good reserve as long as they are safe. And their safety is obtained through the col. ratio, and all the parameters evaluated by the risks teams.

For (2), the assets, they are good as long as they keep (or in fact, increase) their value vs DAI (=vs USD).


If this is your point of view (MakerDAO is just a bank like any other, just not regulated), should we expect regulation to come to MakerDAO (like Basel iii ?)

1 Like

But what is printing DAI if not a borrowing? If I have 10 DAI, MakerDAO owes me $10 of assets at ES. It’s a loan. It was fine to hit 1B DAI. I’m not sure that’s a good value proposition to hit 100B DAI.

Regarding the USDC. It is a risk, but it is also a feature. As long as there is 22% PSM-USDC backing you know that one DAI can be quite easily converted to cash. At the same, if you fear for USDC safety, DAI is a better solution as you would lose less by holding DAI than USDC.

It was a pure synthetic asset (fractional with 0% is correct, but probably not the best wording). And that was fine for the time. I mean when you compete with USDT being shady as hell …

Transparency is one of our best features. No fiat-backed stablecoin is transparent currently. But when they will …

Now we can also say that DAI is better because it is more censorship and regulatory proof.

Nexo and BlockFi need to pay 10% interest to have deposits and do crypto-loans (almost like MakerDAO). But they are not transparent at all and you can’t do much with their bank deposit.

But that will change. Let’s say Nexo becomes a public company where everyone can see the balance sheet and the size of the surplus buffer. They tokenize the USD deposits to create NexoUSD that yields 10% (let’s say by increasing the amount of token, not the price that is always $1) and can be redeemed at any time for USDC/BUSD/PAX/GUSD because they have a partnership and they keep $-reserves (like the PSM in a way).

And you can have exactly the same model for Aave.

Now let’s say Stripe/Visa has a payment system where you can use any safe-enough stablecoin for stuff for $1 no slippage. Do you keep USDC, DAI/CHAI, NexoUSD or aUSD in your wallet? And what if your stablecoin is not good enough to participate in this clearinghouse system?

This is the war of 2022 (or maybe more 2025). We have a headstart but we need to stay on top.

TL;DR: the spread between SF and DSR will fall except if we are super good.

1 Like

What does ‘super good’ mean operationally?

I’m being convinced by @iammeeoh’s questions that the ‘just a bank’ descriptor creates an inaccurate picture (in part due to misconceptions about the word bank, but words are made to communicate, not pay respect to the dictionary) which is to our detriment. For instance bank-run situations while similar at first glance are actually pretty different.

Maker is a unique mix of features from different existing entities, plus some unique ones. It should be described as a new object.

1 Like

Not sure, but Nexo has 60 developers, BlockFi and Aave probably more, we have 5 (something like that). Even if everyone is worth 10 it’s not enough to compete. Let’s say I’m humble. If we are better we will win anyway so useless to spend too much time on this case.

Which unique ones? That’s a key point. If we have something unique we should push on it.

From Wikipedia: “A bank is a financial institution that accepts deposits (USDC) from the public and creates a demand deposit (DAI) while simultaneously making loans (crypto-backed loans).”

On the marketing side, I fully agree. I’m not a marketing guy. If the marketing team finds a better term that is still somewhat correct, happy to use it. I’m okay to bend the reality, I’m not okay to live in a fairy tale.

1 Like

This is an archaic description of the way a bank operates. Banks source deposits ex post facto to meet the capital requirements outlined by the central bank and regulators. These requirements exist, at least allegedly, because the banks are printing dollars that are considered fungible within the Federal Reserve’s payment network. If they did not want this privilege of printing money that can be used in the Fed’s network, at least in theory they would not need depositors at all. Rather, they would source holders of their notes in order to ensure that they don’t suffer from inflation versus whichever benchmark they are targeting (like Dai with the USD). I’m aware the outcome is semantic, but the difference in concept is important.

This is why it doesn’t make sense to say that Dai has “reserves” of any sort. Reserves for what? If you’re saying that it’s for liquidity, I think you’ve taken the opposite perspective of what’s actually happening. The Protocol purchases USDC because users need Dai liquidity, not the other way around. I’m afraid I just don’t understand the premise of Dai being “third tier” money as this assumes we seek fungibility with Federal Reserve Notes. We seek stability versus these notes, not acceptance as fungible within their payment network. I’d argue the counter, that being outside of the legacy banking system is our core competitive advantage, and we should seek to sever ties with existing banks (USDC) as soon as possible in favor of real world assets pledged directly by borrowers.


This is what I see. Especially as collateral continues to be diversified, any dramatic collapse in DAI’s peg to the downside would be hard to hedge against anyways. Either the dollar collapsed — not strictly a problem for DAI’s mission of stability vs the dollar as long as liquidity remains in the system — or so many different assets collapsed at once that there’s not much you could have done without the ability to engage in hedging.

You can think of DAI like a worldwide pawn shop — it provides liquidity rather than soaking it up. In a crisis, you actually want to be that provider. See 12 months ago as reference. As DAI becomes bigger, that upside relative to the dollar should become less pronounced. You can already see the smoothing of price fluctuations for DAI as it has grown since last summer.

Not sure about your point. Sourcing deposits decrease the capital ratio. The definition works fine for free banking experiments outside of central banks and regulators.

You can have money first or credit first, it doesn’t change anything really.

Not sure what is the difference here. USDC is not fungible with Fed reserves. But we want high stability. Not eventual stability.

You should be able to pay on Amazon (or whatever) with DAI where DAI = $ at 100% of the time.

I leave the crypto-dollar outside of the legacy system like eurodollar and I hope for the same success that eurodollars got. That or shadow banking, as you want. But if our core competitive advantage is regulatory arbitrage … that’s a bit sad.

Reducing USDC exposure is important. But cutting the link to the US Dollar system will lead to instability due to the absence of risk-free arbitrage (of last resort if you want).

MakerDAO should own tradable liquid assets, not only DAI loans to manage DAI price.

Or there was a game theory that there wasn’t enough DAI to repay all loans. Anyone buying DAI above peg is not wanting liquidity, it’s either a vault owner in a rush to close a vault or a yield farmer that doesn’t care.

1 Like

I think maybe we’re saying the same thing? Both of those are someone wanting DAI to hold for some period long or short, and needing to find them from somewhere. So the asset appreciated in a kind of currency short squeeze. But as DAI grows in market cap, that kind of swing should become less possible.

1 Like

The deposit is created simultaneous to the loan. If you come to “Greg’s bank” and borrow $1,000, the $1000 I print for you becomes the deposit. Monitoring a capital ratio (hopefully we are using the word the same, I am saying that we’d monitor this ratio to ensure Dai is always sufficiently collateralized) does not require maintaining reserves. This is the core premise of your argument that I’m disagreeing with. Dai is not a eurodollar because it does not (and should not) promise its holders USD on demand, nor does it treat any kind of liquidity obligation as a measure of its solvency. You would surely admit that a foreign bank is insolvent if they cannot provide their depositors USD on demand in a dollar-denominated checking account. This is clearly not the case for Dai. The only promise that Dai makes is that there’s always >$1 of collateral for every 1 Dai.

USDC is fungible with commercial bank deposits which are practically fungible with Federal Reserve Notes (so long as the bank is in the Fed’s network and of a certain size). Banks have historically used reserves to ensure that there’s enough liquidity to give the illusion of fungibility in what is technically a non-fungible asset, but the changes in policy made since '08 have made this requirement all but obsolete. In practice the Fed will never allow a TBTF bank to have “insufficient reserves” to meet their obligations and will buy their assets to “provide liquidity” (enforce fungibility where there should not be fungibility). The tradeoff of this “total solvency” being “total subordination.” The key difference between this system and ours is that a bank will fail if it cannot deliver Fed Notes to its depositors when they are demanded, MakerDAO will not fail as it has no obligation to deliver USDC, or any asset, to Dai holders. This is why it does not make sense to say that the USDC in the Protocol are reserves. It is just an asset on the balance sheet. To make this kind of promise would implicitly subordinate Dai to USDC.

The concept of maintaining reserves is only relevant if you (as a subordinated institution) are trying to meet an obligation to deliver base money, or in the case of USDC commercial bank money, to the holders of your liabilities. This is not a relevant comparison for Dai because it is base money. Sure we want ample liquidity with fiat currency, but we cannot commit to fungibility, or at least any level of liquidity that mimics fungibility, without subordinating ourselves to the legacy system (like USDC). To illustrate this point with an example, you wouldn’t call US Treasuries purchased by the Fed “reserves” when they use open market operations to maintain the price level of the US dollar. The same applies to the Protocol purchasing USDC. The Protocol can’t have reserves for the same reason the Fed itself can’t have reserves, because both create base money in their respective ecosystem. Neither entity can default on its obligations because it has not promised depositors anything for holding its liability. These are just assets on the balance sheet. Dai holders may have recourse to the balance sheet through an ES, but this is up to MKR holders and is not done at the demand of Dai holders - the same way that the Fed does not owe holders of Fed Notes access to its collection of US Treasuries. This might be easier to conceptualize if you imagine that Dai was pegged to CPI instead of the USD - what would the reserves be in this situation?

I fully agree with this an think we’re saying the same thing. I am saying that we should sever our reliance on bank liabilities and try to source productive assets from the legacy system (the same assets banks are competing for in order to generate their own liabilities). I don’t care if the assets are readily tradable/liquid so long as this risk is correctly priced.


I Iike this a lot, and hopefully the tokenization of income-producing assets will hurry up. Because that would provide what you’re describing. I suspect it would also be less correlated with what is already used as collateral.

1 Like

@g_dip Sure we use different words but we agree on almost everything.

While we don’t have any “contractual” obligation besides $1 of stuff per DAI at ES, market forces will impose something more challenging on us.

Going from a bank deposit to DAI and back should be done with no slippage, no exchange rate different than 1:1. Otherwise, you are adding friction and a product with friction is just not good enough. Even USDT is able to deliver on that.

I know this is a bit early. I mean, $FEI is punishing their users if they want to sell and people have bought anyway.

1 Like

Yes, we can shake on this :slight_smile:

Now let’s go find some high quality RWA


7 posts were split to a new topic: Indexed Universal Life Insurance - High Quality RWA

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.