PSM Governance Timeline

I think a tight fee band is a good idea, but will 50M DAI really bring the peg all the way down to $1, or even $1.0004? I’m doubtful. In addition to a tight, symmetrical fee band (0.04% fee in and fee out), I would also advocate for some wider, asymmetrical fee bands, eg:

2% fee in, 0.2% fee out, 50M DC
5% fee in, 0.1% fee out, 100M DC

I bet independent market makers will be much more comfortable selling/shorting DAI if they are confident the DAI price is capped in the 1.02-1.05 range.

I’m optimistic 50m Dai sell order will restore the peg. Almost all the Dai generated over the last few weeks has been deposited in Compound where it does nothing to restore the peg. People are afraid to sell Dai right now because of difficulty buying it back and lack of confidence peg will be restored.

So this range would effectively provide the same function as USDC-B vault?

Isn’t PSM designed to be a liquidity provider of last resort?

  • Fee In : 1.5%, then 1%, then 0.5% (stop if the peg is restored)
  • Fee Out : 0% (to reduce USDC custody risk at the minimum)
  • Debt Ceiling : unlimited (more on that below)
  • ERC20 : USDC, then the Paxos stablecoins

Those setting would enable the ability to buy DAI at 1.015 USDC then 1.005. We can discuss if a cap at 1.005 is close enough to 1 for DAI users (I would say close enough).

When peg return close to 1 (starting at 1.003 for Uniswap users), the PSM USDC will be drained by arbitrage (less expensive than any alternative).

On the unlimited cap, it means the willingness to increase the ceiling each week if needed. It will avoid any significant arbitrage risk if USDC fails. Still, this “whatever it takes” approach will send the right message to the public, anyone buying DAI at 1.015 or above will lose 1%. The ceiling will have to be reduced if not used.

The main risk is if the community think DAI deserve a premium to USD (or any other USD alternative). Let’s hope collateral onboarding will provide enough supply while keeping counterparty risks low (diversified). In any case, the peg will be hit, the more the USDC PSM is filled, the less valuable DAI is (at the limit USDC risk + MakerDAO risk can’t be more valuable as only USDC risk).

If we want to provide liquidity of first resort, we can provide liquidity to trading pairs on Uniswap and balancer. That would be great to bring value to DAI usage. Much more an ecosystem building strategy. I like it. But this isn’t the same problem.

Due to the recent communication from the Foundation about the PSM, I don’t believe that it makes sense to continue polling past the current point, and I will not be putting up more PSM polls at the current time.

While we could vote on parameters, I suspect that by the time the community has organised around producing the PSM themselves, the macro environment will have changed significantly enough that they would no longer necessarily be valid.

If members of the community do want to continue pursuing the PSM concept, I would suggest that someone creates a Declaration of Intent (MIP13) and passes that through the MIPs process.


The sandbagging on the PSM poll is extremely disturbing. Until MKR has delegation + automatic vote deadline extension on lead changes, that kind of last-minute vote dump to switch the outcome has terrible optics.

The community obviously wants PSM, no other topic has had this level of engagement for a long time.

1 Like

The on chain poll results say otherwise:

1 Like

@lix Lets be honest, it’s a razor thin line.

@IslandHunting With regards to the sandbagging, the only thing I can suggest in the meantime is to make sure you vote with your MKR if you have it. There is something like 280k in DSChief, only 40k voted in the PSM poll.

1 Like