[Signal Request] Should we add new PSMs?

A PSM is a bank deposit owned by Maker. Currently, we have put all our liquidity at Circle.

Now if Circle is the Lehman from the next crisis, this is bad. This is why bank usually diversifies their counterparty exposure. So from a risk perspective, I think having many PSM is good.

It is also good from a business perspective as favoring competition between fiat-backed stablecoin can only help to stay relevant. And we should be unbiased. This is good for business development as well.

On-chain liquidity is not that important I would say as you can always mint as much as needed. PAX has low liquidity while BUSD is better but you can swap those at Paxos for no cost and in an instant. The backing is the same anyway.

In my view, the aim should be to decrease our liquidity reserves to ~10-20% and diversify the counterparty risk between at least 5 counterparties. i.e. a 2-4% exposure at most.

I would also favor more auditing. I would be fine with TUSD and USDT if we could have confidence in them. If we provide a 200M PSM to Paxos, I’m quite sure we could get more information.


I think what makes the best asset for the PSM is which one has the best peg to the usd, whatever happens.

The main reason for having the psm is to be as close as possible to the usd. If usdc is under at 0.98 that means we will buy usdc at 0.98 for the price of 1 dollar.

It seems for me one asset should be good enough, as far as it full fills the initial purpose.

In my view it is bad to have a higher percentage of stable coin, because it just doesn’t fit the initial purpose. I would prefer to have a better rate to attract more users and more collaterals.

A few questions for “No” voters.

  • Why are you not voting for “Sweep Only” where applicable?
  • How is BUSD different than PAX?
  • Can you give more colour on your “No” to USDT/TUSD/BUSD? I assume it has to do with trusting the issuers.


Sweep only is a good option for taking existing stablecoin vaults and moving them into a PSM.

  • TUSD was great at first, and then they sold out to a questionable actor, therefore sweep only is my choice there.
  • PAX unfortunately has such low liquidity and adoption but maybe it will pick up. A small PSM DC is better than sweeping imo.

Binance has not treated MKR or DAI very well, from a BD standpoint, and I think that’s why people lean against BUSD. To me it’s irrelevant, and I agree with Sebs points about this being good for BD. Their regulatory legitimacy is nice so thats why I’m voting yes on a BUSD PSM.

USDT/TUSD - I voted no.

  • For USDT it’s a PR thing. From what I understand they released their backing info and they have some portfolio of decent assets backing it(75% cash equivalents), but i’m not totally aware of the risks and feel general discomfort around USDT(maybe unjustly now.) Also I think the space should do what it can to reduce USDT’s market share–it’s enormous.
  • As for TUSD, fuck Justin Sun is my reason.

Just to clarify some confusion about “sweep” vs. “no”.

  • This proposal would only “sweep” Vaults that are already under-collateralized into their respective PSMs. Anything else is confiscation. So “sweep” is like saying: “liquidate this Vault type into its PSM”.

  • There is no other elegant liquidation mechanism for these until we patch the vow to have a collateral specific bad debt queue. We can’t just put this out for liquidation with a duration of 1 year, otherwise it will become bad debt within a few days. We could limit the ilk.box for these so they only do 100k DAI at a time and liquidate over a year that way, but that will be costly for incentives to bark. It’s much cleaner to liquidate into the PSM and let anyone come along and take the collateral for the DAI outstanding whenever they want. If you’re voting no on TUSD, you presumably are assuming LIQ-2.0 fixes the liquidation problem here, and perhaps it does if we limit the amount out for liquidations and let it constantly liquidate for a year, but I think “sweep” is a much more elegant solution.


USDC is currently under some suspicions.

The interesting part is the following where the accountant no longer say how much US Dollar in the bank account. The only reason I can think of is that they are investing in something were valuation is not so easy and they are not sure if it should be on a cost basis or on a mark to market. They are climbing the risk ladder (because cash in a bank account is easy to value).

Not saying that there is any problem here and now. Not so sure in 5 years or if the short term rates in the US dive below 0.


With regulations looming it seems that we need a solution to the circular nature of holding other stable coins as a means of protecting the peg. Even holding gold in a vault would be better than being exposed to stable coins that may or may not be holding cash. The other day I mentioned treasuries or Japanese Notes would be a great alternative, but perhaps Gold would be best at least until there is a way to hold treasuries. I really don’t think this is a far-fetched view and monitoring the regulation chatter makes this all the more critical.


Maybe that’s why USDC has been able to scale so fast compared to GUSD… Seems like everyone is going fractional to gain a competitive advantage.

1 Like

Sir, they have expanded to others—such as Solana USDC, Algorand USDC, and many others that I cannot keep up with them. I never worked at an auditing firm but I’m pretty sure it is a stressful occupation that usually takes more than 4-business weeks to complete.

I disagree that Binance has treated MKR or Dai badly. They been a big supporter of Maker ecosystem and although it took a while, they listed both tokens for free (as this is a public information). Also BUSD team themselves went through the process to apply for collateral onboarding. So if some believe that way, I want to clarify that that’s not the case.


Thanks for that! I only speak from the memory of poor sentiment back before they did the listings. They did get a lot better I see.

1 Like

Did you read the whole article?

The piece itself is under-researched (the comments clear some things up) and is just trying to poke holes at random findings.

I wouldn’t trust this author - every 2nd post of his is about how this whole “crypto bubble” is about to explode, apparently built on the promise that most crypto volume is solely because these stablecoin issuers are printing free money and buying up tokens with them.

Let’s not just link random articles and create the false impression that things are not stable elsewhere. I’m no USDC proponent, but it’d help this community to keep a realistic pulse on the industry.

Yes, and it is entertaining. I don’t know why you need to trust the author at all. I highlighted what I thought was important and it is a fact, not an opinion.

If I had any serious issue with USDC right now, I would yell a bit louder. This is not the case.

Can folks who are voting “No” on TUSD instead of “Sweep” tell me why? We need an efficient way to get rid of this collateral and moving to a PSM would be the most efficient.

@PaperImperium @g_dip @alexis @jernejml @Lozadaluis12 @Pan @Andy_McCall

Yes sure,
We don’t need a PSM to liquidate TUSD, there is 22M only.
And anyway USDC PSM is already at tout 0 so we won’t be able to sell it. And not even under 1.0004.

Seems a complicated solution/time consuming for peanuts. Either we liquidate because there is a risk or we keep it until we get under 1, then we can start the liquidation.

Put the TUSD inside the PSM is even more a risk as we won’t be able to sell it in case we have to.

And regarding the others PSM, why I voted No, it is just because there are illiquid you can’t sell 1M Pax without moving the price above 1.001 which is our margin.

For me, at this point only usdt can have value to add as PSM. But here we are … usdt …

1 Like

Similar to @alexis i think the protocol should be trying to get TUSD off the books, and maker doesn’t need anything complicated to do that.

Sure you might end up losing a bit of money in the liquidations, but IMO that is just one of the occupational hazards here. Plus, given the exposure I don’t really see that as much of an issue. Assuming maker is able to liquidate for ~90% - 95% of FMV you’d only be talking about losses in the $1-2MM range.

Not really all that big of a deal IMO.

So I worry about having multiple PSMs, after I have thought about it more and discussed it with several others who pointed out some possible dangers I had not originally thought of.

Right now our big risk is exposure to USDC – mainly in the form of them involuntarily being forced not to honor USDC sitting in the PSM, as it seems unlikely they would want to destroy their own brand, reputation, and peg by voluntarily monkeying around with us.

Note that that risk – regulators based in the US where Circle is deciding to do something that negatively impacts us – is not lessened by including other US-based centralized stablecoin issuers. It does reduce the risk of Circle deciding to take a swing at us, but we likely hold enough of their float that we present an existential threat to them as well, since we could force a mutually assured destruction by asking to redeem our $2 billion in USDC on a whim.

We do, however, add additional risk by creating a multiple-legged PSM. Suddenly there is not just one peg to maintain, but multiple. That means that, one stablecoin issuer could mint a bunch of their coin unbacked, move it into our PSM for DAI, use the DAI to take the redeemable stablecoin of another issuer from our PSM, move it onto their books. Net result? They grow their circulation and balance sheet for free, and at our expense.

Just as we (and Circle!) have benefitted by the current PSM, some other stablecoin issuer could do the same trick but entirely through our PSM and not through the secondary market.

We cannot/should not operate multiple PSMs at the same time without a solution to the above scenario. Otherwise, we risk the DAI supply shrinking by billions instantly, and all the havoc that would wreak both on our business and on the markets as a whole.


Why ‘at our expense’?

1 Like

Because it can move $2 billion USDC from backing $2 billion DAI to backing $2 billion GUSD/whatever. Not good for DAI stability. More counterparties, more complexity, not more DAI.

I don’t see the problem here. How is using the PSM route different than swapping from one stablecoin to another using Curve? To take your example, suppose Gemini swaps GUSD to USDC using Curve? So what? Suppose Gemini swaps GUSD to DAI via a PSM and then uses the DAI to draw USDC out of the current PSM. So what? It seems like the only difference is that by using through the PSM, Maker collects some fees.