Prudential treatment of cryptoasset exposures

The Bank of International Settlement (the so-called Bank of Central Bank) has published a consultative paper on the prudential treatment for cryptoassets. The public has until September 10th to address comments. This is a first review, not an extensive analysis.

The impact of prudential treatment is quite important as this will limit banks to invest or use crypto assets and stablecoins. This topic was raised as one of our prospects for RWA is a big bank. As DeFi is based on price (collateral-based lending, for which the collateral price should be efficient), it is important that financial institution can put their huge balance sheet at work. Preferably using DAI than USDC.

Prudential rules 101

For banks, the main metric for prudential treatment is the Core Equity Tier 1 ratio (CET1 ratio). In Maker parlance, it’s the ratio of the surplus buffer divided by a risk-weighted sum of assets exposures. A bank is usually above 10% on that metric.

We currently disclose a CET1 ratio for MakerDAO where each crypto-backed loan is weighted 100%, ETH-B is weighted 200% and stablecoins (via crypto loans or PSM) are weighted 0%. Quite arbitrary for now, but will give you a reference for the rest of the post.

Learnings

From the document:

The first learning of the proposal is that if you hold a crypto asset like ETH, MKR or WBTC (group 2), you should give it a 1250% weight (or x12.5). With the 8% CET1, if you have only riskless assets (like Fed deposits or Treasuries) and crypto assets, you can lose all the value of the crypto asset and still be fine regarding your customer deposits/stablescoins.

For comparison, Microstrategy shares would be 300% weighted (publicly traded equity). 1250% is the maximum in the regulation.

The second learning is that if you hold a tokenized RWA (real-world asset, group 1.A) the risk weight is the one of the underlying. So a PSM-PAXG might be a very small risk (8% it seems due to the currency risk).

So PSM-ETH 1250% risk-weighted, PSM-PAXG 8%.

What about DAI?

To be in group 1 you need to have a stabilization mechanism that works at all times. This is defined as:

The difference in value must not exceed 10bp of the value of the underlying traditional asset more than three times over a one-year period."

But moreover:

Stabilisation mechanisms that: (i) reference other cryptoassets as underlying assets (including
those that reference other cryptoassets that have traditional assets as underlying); or (ii) use
protocols to increase or decrease the supply of the cryptoasset; are not considered to meet this
condition

So exit DAI and all algorithmic stablecoins.

So DAI is one of the most risked asset on earth for Basel iii.

Implications

USDC is most likely Group 1 (a or b) so costing very little capital while DAI is considered extremely risky, and therefore costly. This is obviously not good for MakerDAO competitiveness going forward. More broadly, the proposal is not satisfactory as it doesn’t make a difference between ETH and the last memecoin and doesn’t recognize governance tokens as equity of startups.

If this interpretation is correct, it would be wise to answer this call for comments and try to get more adequate regulation.

14 Likes

Curious how value is defined here. Coinbase prices? Uniswap prices (if so with respect to what)? USDC also has off-peg moments every time crypto goes a bit nuts so I’m not sure it will end up in Category 1.

As for DAI, making a persuasive argument for it to be in Category 1 is honestly going to be very hard. The complexity behind DAI is vast compared to USDC. Someone is gonna have to explain vaults, liquidations, RWAs, PSMs, emergency shutdown, governance to set rates and so on. It’s hard to see them being convinced of how all of this works to guarantee DAI’s stability.

Well for sure DAI is not cash, but we are not 4x riskier than Gamestop equity or a CCC-rated bond neither or on par with BTC (or MKR).

I don’t know for USDC, but it can be group 1a. Not sure. And quite sure they don’t care about Uniswap price.

Nice. Some work with an overlap into another client’s recent experience :wink: .

I’m in favor of responding. @SebVentures Are you thinking about a collective response from the DAO or encouraging community members to submit their own comments?

I’m not sure that pushing back on the 10bp threshold would be as effective as simply pointing out that:

  1. Dai is the industry’s leading permissionless stablecoin
  2. Dai has a stabilization mechanism that’s arguably effective but incompatible with the proposed definition of a “stabilization mechanism”
  3. Dai is collateralized by both digital and “traditional” assets,

and therefore we feel the proposed framework isn’t effective. We would prefer to see a framework that includes greater consideration for the unique characteristics Dai brings to the table.

5 Likes

So I had begun to put together some basic points for this, but there’s not just one or two reasons DAI would have a hard time getting any better treatment under these recommendations. It would require multiple heavy lifts, and it’s not just around the stabilization mechanism.

Maker isn’t housed in a legal jurisdiction, and we also don’t have external audits (though that one could be done if we really needed to check a regulatory box). Crucially, DAI is neither redeemable upon demand nor is it a “digital representation of a traditional asset” unless we want to start calling it a dollar (we shouldn’t). Additionally, no one can verify the ownership of all assets backing DAI.

There is also this condition, which I think will likely cause trouble for nearly any issuer: “Entities that execute redemptions, transfers, or settlement finality of the cryptoasset are regulated and supervised.” They note this includes the “transfer and settlement systems for the cryptoasset” which is… um… I guess Ethereum’s entire distributed ledger. Good luck with that if you’re on a public blockchain.

We should definitely leave a few comments, but DAI is excluded from a useful category pretty much six ways to Sunday, if these guidelines are adopted as they are now.

This would be an extremely heavy lift, and our time and resources are likely better spent elsewhere than an actual attempt to get these altered. If anything, this is a lesson that we need a seat at the table when these discussions are ongoing in the early stages.

4 Likes

The problem is, it kinda seems like if we don’t do something about this, it’ll really hurt us down the line. It seems like this is their way of trying to get in behind us so they can get us over the barrel. The government has one idea of what cryptocurrency is and if you don’t fit into that idea you’re gonna have a bad time; we need to make sure that idea fits us too before it’s signed into law, otherwise it will kill us in favor of USDC and the like.

2 Likes

Putting DAI in the same box as USDC may be too hard, but creating sub-boxes within group 2 may be achievable and would go a long way.

As a sidenote, I find it ironic that the Ethereum world and the BIS would classify DAI as very risky for precisely opposite reasons.

1 Like

Would say a collective response would be best, but unsure how to proceed. You probably need a strong understanding of Basel rules to write a proper response (not sure we have a Basel expert) and some good connections (I might have one or two but far from enough).

This will be a significant investment of resources and the outcome is not really in our favor but can be game-changing.

3 Likes

Perhaps get the folks over at Compound, Curve, and AAVE to join in collectively? And yes, please reach out to your one, or two connections. The community needs all the help it can get.

BTW thanks for posting this and bringing it in to focus.

3 Likes

My somewhat educated suggestion. When it comes to Basel requirements, the approach is usually not done by BIS directly to a VASP but rather through a regulatory body of the jurisdiction in case (BIS or its committees are not a regulator). When that approach comes in a “please explain” letter, rush to respond is usually not recommended. When the regulator really takes the time to send a VASP a letter, rest assured it will not be a general one.

They will take the time to educate themselves very deeply into what is MakerDAO doing and what specific points they want it clarified. Saying that from experience of a project I know well that has recently received a “please explain” letter.

If we wanted to pre-empt the regulator (which one?) to avoid the “please explain” letter then maybe there might be room for engaging a higher level committee like Basel earlier. But again, this would entail a very well prepared and structured explanation, usually not in our crypto lingo but in wording the body/entity will understand. Failing explain well in their own words, risks getting further red flags and greater number of questions asked.

1 Like