[Signal Request] tin/tout parameters for new PSMs

Overview

The Signal Request about the introduction of new PSMs led by @Aaron_Bartsch ended last week. I want to follow up here to keep this nice initiative running.

First of all: sorry for the delay; we had some discussions around this topic within the MOMC and in the end we agreed not to poll on all the different parameters (tin - the fee we take for every swap stablecoin, tout - the fee we take for every swap DAI → stablecoin, line - the Debt Ceiling - for BUSD, PAX and GUSD) but instead propose a plan on how to move ahead here:

As soon as we have PSMs for PAX, BUSD and GUSD in place we would like

  • to incentivize minting DAI from stablecoins on those new PSMs instead of the existing USDC-PSM to diversify our stablecoin exposure
  • encourage cross stablecoin swaps

To achieve this, we propose to

  • set tin to 0.1% for PAX, BUSD, GUSD and
  • increase the tin we have on the USDC-PSM to 0.2% and
  • set tout to PAX, BUSD and GUSD-PSM to the same level we have on USDC-PSM: 0%

As for debt ceilings, it is all about counterparty risk. Both BUSD and PAX are managed by Paxos. BUSD is a bit more liquid on-chain than PAX (see slippage curves BUSD PAX GUSD), but in essence on-chain liquidity isn’t as important factor as for other collaterals since there tokens are redeemable for USD and arbitrage should assure close to 0 slippage when potentially offboarding these tokens in the future. Yet, we must decide on the counterparty risk level we are willing to take towards Paxos which manages both PAX and BUSD. Same goes for GUSD where Gemini manages it.

Do you support the tin/tout parameters proposed by MOMC?

  • No (please add a comment why it is a bad idea)
  • Yes
  • Abstain

0 voters

How high should we set the Debt Ceiling for PAX-PSM?

Please vote for all options you would support in an onchain poll

  • 100 MM
  • 300 MM
  • 500 MM
  • 1 B
  • Abstain

0 voters

How high should we set the Debt Ceiling for BUSD-PSM?

Please vote for all options you would support in an onchain poll

  • 100 MM
  • 300 MM
  • 500 MM
  • 1 B
  • Abstain

0 voters

How high should we set the Debt Ceiling for GUSD-PSM?

Please vote for all options you would support in an onchain poll

  • 100 MM
  • 300 MM
  • 500 MM
  • 1 B
  • Abstain

0 voters

Next Steps

The Polls will run until 2021-06-24T13:00:00Z; their outcomes will either result in on-chain-polls assuming the outcome of the polls deems it necessary or its intermediate results are going to be taken to another initiative - another signal, onchain-poll or urgency executive.

6 Likes

Do you or anyone from @Risk-Core-Unit have recommendations about the debt ceilings?

The range is pretty large and it would be good to know how we view Binance, Paxos and Gemini vs Circle, in terms of our confidence that they are backed by dollars.

4 Likes

Can we delegate this setup to the MOMC?

1 Like

Echoing what @AstronautThis commented I’ve voted for lower DCs for all options for now so we can see how they perform, unless @Risk-Core-Unit offer any alternative advice.

I agree the tin on USDC PSM needs to be increased. I actually would propose an automated bend don’t break (somewhat elliptical) tin/tout on USDC based upon a sliding scale of DAI demand from the PSM and DAI market value that could go up to 0.5% on either side. Won’t give all my reasons here, but an automated one would give confidence to the market and DAI holders, permit recognition at this time that DAI is stronger than USDC or USD and not hand over all that value to 3rd party arbitrage, and still keep DAI tightly pegged to USD.

2 Likes

I voted no because I don’t think we should set tout to 0. This float of stablecoins is unproductive right now, but it’s not without value, and if someone wants to pay us to to hold IOUs we think are of equivalent value to our own, we should not be in a hurry to unwind that. Work is being done to try to find a way to generate yield on these in a way that does not increase our risk, and as soon as we do, we may regret trying so hard to offload billions in free float.

4 Likes

Agree. I voted yes to get the tin up on USDC, accepted the zero tout on the new stablecoins as a compromise. if I were writing the rules tout on the new stablecoins would be something more than zero - 0.01% perhaps.

2 Likes

If we complain about too much stablecoin exposure a tout of 0% for sure will help on the slightest down-peg. Imho an accelerated offboarding of stablecoins is far more important than getting yield out of it.

2 Likes

Point taken. But we don’t have that problem until we’re sitting on a disturbing amount of a particular stablecoin, and we can act then. Initially and so long as a particular stablecoin PSM is running well we should take something for the exchange service provided.

2 Likes

Right. so we could just keep the tout to 0 for USDC and have it a bit higher for non-USDC.

BUT: knowing that the tout is 0 will hopefully make the non-USDC-PSMs more appealing - and that’s the major motivation for having those new PSMs in the first place: to diversify the stable coin risk.

I think having a tout > 0 for the new PSMs is a bit like repeating the mistake we did with the USDT-A ilk: it did not have an appealing setup, so it never got adoption.

We can always readjust along the way. But we should start with something not totally off.

1 Like

I am a no for multiple reasons:

  1. I don’t like bundling all the liquidity into a single price band as I have commented on previously.
  2. Personally would rather see CUs explore options to get us out of this PSM mess as a PEG management mechanism. I have yet to even see a comment from risk or anyone that has looked at doing this and decided there are no viable alternative approaches.
  3. If legislation comes down it is probably going to hammer us across the board here.
  4. I don’t see that these other coins offer much liquidity. We have already eaten up a significant chunk of the USDC outstanding. Exposure is not going down.
  5. Would like to see some movement on RWA and institutional borrowers because it is my impression if we get bigger players on board here we might be able to do away with the PSM approach to PEG management and utilize other approaches.
  6. Back to ideas of using a treasury so that we can use our own funds with special internal vaults to basically accordion DAI liquidity ourselves. (Another reason to have a treasury funded with not 50M but more like 500M of assets).

I see that I am probably the odd one out of the entire community on this. Ironically my comments in the past inspired the creation of the PSM I just wanted it to be banded so we could get proper signals from the markets about DAI demand.

I honestly want to see more discussion about how we are going to grow system deposits so we can grow DAI liquidity and whether there needs to be some centralization of DAI liquidity management beyond the PSM model.

4 Likes

I was campaigning to keep the tout at 0.01% because in my idea it would be more efficient to push the dai to 0.99, increase the supply and having the PSM as back stop to be sure we don’t slip too low.

But to archive this, we would need a huge effort on rate to be under the market rate.

However I doubt the community and MKR holders are in line with it. The strategic line that seems to be willing by MKR voters is more keeping high marging, an high PSM and don’t play the Dai supply increase. In that case a tout at 0 is the correct strategy as we start unwinding at around 1.0004 before curve get flat and we can get another round of fees as soon as the dai goes back to 1.0006.

It is also why I think those parameters need to be managed by the MOMC as they need to be turn up as soon as we start to be more aggressive with the rates.

@ultraschuppi I have errors occuring when trying to access the https://maker.blockanalitica.com/ links you provided.

Maybe they don’t exist yet?

I support increasing the USDC-PSM tin due to the fast growth. We’ll still be providing an important backstop to control the DAI peg. This will also encourage growth of these new PSMs. I would’ve been more conservative and just suggested a 0.15% tin, but 0.2% may be more effective.

One other reason for a small tout for these new alternative PSMs is that it’ll slightly encourage market participants to shed USDC from the USDC-PSMs before these new PSMs.

Yeah, I agree but the PSM is also what we have right now and I think diversifying our PSM collateral is a strong priority.

Also some people consider that we have a dual mandate of both trading fees and peg management. I think we should focus on effectively using the PSM for peg management. We shouldn’t expect high turnover daily. The PSM is not as nimble as other participants.

It’s incentives. Stablecoin supply is dynamic and if there’s strong demand for DAI like in the past, these PSMs will get used. It’s not about % of USDC supply but $ amount of excess DAI demand.

I also think we should use the USDC as another source of revenue. I think it only makes sense.

1 Like

It is great to see our continued expansion on PSM, let us work hard to make this cake bigger. great!

thanks for pointing out - is fixed now!

1 Like

I believe that PSMs are important, and I think that we must follow the path of not remaining in a strategy and complement it with other better ways to grow Maker.

This is my personal view of the PSM situation and DC limits;

We wish to diversify USDC counterparty risk with other similar products. This risk can be divided between willingness to serve obligations and ability to serve obligations.

In terms of willingness to serve, I doubt CENTRE/Circle (USDC) would “attack” MakerDAO with restricting our use of USDC reserves freely; this would mean that USDC is not suited for any kind of permissionless protocol and would lose its current dominance of CEX stablecoin in DeFi. Other, smaller stablecoins we are considering here in my opinion increase our exposure towards this specific risk (willingness to serve), as there is a higher probability that a non dominant product tries to do something scathy, as they have less to lose in the first place.

Ability to serve obligations can be further divided to financial restrictions; insolvent or illiquid organisation - and legal restrictions; change in the legal system these companies operate in leads to inability to continue serving their obligations.

Comparison of companies involved with relevant stablecoins and regulatory status;

  • USDC; Governed by CENTRE (Founded by Coinbase & Circle (Several different investors, from large crypto focused VCs, DCG to Goldman Sachs and Bitman)), incorporated in the US. Currently the only issuer of USDC is Circle, but they plan to have more potential issuers in the future. Circle is regulated by FinCEN as a Money Services Business in the US. Beyond the US, Circle is regulated as an Electronic Money Institute by the FCA (UK). Circle accounts are held with FDIC insured depository institutions.
  • GUSD; Operated by Gemini Trust Company, LLC (Owned by Winklevoss twins via their holding), incorporated in the US. Gemini is regulated by New York Department of Financial Services (NYSDF) and their reserves are eligible for Federal Deposit Insurance Corporation (FDIC) insurance up to $250k per user.
  • PAX & BUSD; Operated by Paxos Technology Solutions LLC, assets are custodied by Paxos Trust Company LLC (Several investors, including DCG and few individuals), incorporated in the US. Paxos is also regulated by NYDFS and customers are eligible for up to $250k FDIC insurance for funds held at Paxos Wallet account.
    • PAX & BUSD are issued and custodied by Paxos, they are interchangeable between each other within Paxos platform and the US dollar (USD:BUSD:PAX).

All of these companies are incorporated in the US and are thus subject to the same risk of change in the legal system which could have a negative impact on their ability to serve obligations. In my opinion, if a situation arises where a regulator or some other authority decides to combat or restrict these products, it will be done in a systematic manner which means there is very little diversification benefit (legal reasons). Note that FDIC insurance probably does not help in our case, as the DAO is not a legal entity.

Comparison of reserves management and financial institutions involved;

  • USDC; I couldn’t find information about which exact financial institutions they are using (reserves are in the form of US dollars), while their monthly reserves audits (April example) are published by an independent accounting firm Grant Thornton LLP.
  • GUSD; Reserves are custodied by State Street Bank and Trust (reserves are in the form of US dollars), while their monthly reserves audits (April example) are published by an independent accounting firm BMP & CO. LLP.
  • PAX & BUSD; I couldn’t find information about which exact financial institutions they are using (reserves are in the form of US dollars and US treasuries), while their monthly reserves audits (April example) are published by independent accounting firm WithumSmith+Brown, PC (“Withum”), which confirms that on the specific date and time, assets held in “Reserve Accounts” exceeded or were equal to amount of issued tokens.

It seems that issuers of relevant stablecoins use different financial institutions as counterparties which do offer some risk diversification benefits (financial reasons).

The analysis of counterparties here in regards to their legal structures, financial institutions they operate with and asset reserves is based on information on their websites and other public databases such as Crunchbase, etc. More in depth analysis would require much more time investment and is in my opinion out of the scope of current PSM reserve asset concerns.

In regards to the DC for relevant stablecoins;

We have to understand that all other USD nominated coins are in some relation to USDC and also USDT due to how these assets are paired across different AMM markets in regards to their prices. USDC is currently the largest CEX stablecoin paired with DAI and has much higher supply than other considered stablecoins and as supply of USDC increases in general, so does allocation of it across AMM markets which means we will be forced to further increase USDC PSM reserves DC inevitably of other stablecoin PSMs if we want to maintain the USD:DAI peg.

Current supply of USDC is 24.276b, GUSD - 203m, PAX - 789m, BUSD - 9.412b. Due to how unproportionally higher the supply of USDC is compared to other coins, the only considered stablecoin which can make a difference at current time is BUSD. Because PAX and BUSD are basically almost the same asset (managed by same entity and interchangeable between each other inside Paxos platform) and additionally the actual usage of PAX in DeFi is much smaller than BUSD, I think PAX PSM is not really needed or will be effective for arbitraging dai premium. Furthermore, GUSD is currently small, but based on some information, they are planning to significantly increase their effort to increase usage inside DeFi so it is worth adding.

We are currently exposed to more than 10.8% of total USDC supply across PSM-USDC and different vaults; hopefully with new PSMs we will be able to lower this exposure. In my opinion, there isn’t a perfect initial or final DC for each stablecoin we should aim for, there is the effectiveness factor for arbitraging and realistically we will need to increase these DCs as supply of foreign stablecoins (USDC, GUSD, BUSD, etc) increases on Ethereum.

Further on, we will increase the DC based on usage and effectiveness of these coins to arbitrage the dai peg.

Stablecoins which are directly paired with DAI in highly liquid markets are the most effective. This only leaves us with one serious additional option available other than USDC, which is the notorious USDT. This is not a recommendation to include USDT, but merely an explanation, that all other stablecoins are much less effective for arbitraging the dai premium.

6 Likes

There doesn’t seem to be too much consensus over Debt Ceiling amounts here. I’d like to encourage more people to vote before the signal closes tomorrow, and suggest that those that have already voted consider compromise options.

Thanks all.

1 Like