[Signal Request] Lower PSM-USDC-A tout Parameter

After observing the USDC PSM in action for over a month now, it is becoming clear that the fees in the DAI → USDC direction are uncompetitive. So far the PSM has only produced ~$100 in trades going in the DAI → USDC direction. The Rates Working Group would like to propose lowering the tout parameter on PSM-USDC-A to encourage more trade volume.

There are a few considerations to make in what the new rate should be. I’ll do an overview of each:

Competition Implications

Currently Curve provides the best rates when DAI = USDC for any trades around the ~10M or less mark. This is because their fee is 0.04% whereas ours is 0.1% and they have very little slippage around that level. As you increase the trade size, the Curve slippage increases so the PSM becomes more competitive, but there aren’t that many trades at that size. By lowering the fee we will begin to eat into the market share of Curve and increase trade volume / revenue.

Around the Curve fee there is an asymptotic effect that we need to take into as well. Since the PSM does not have slippage, we will be strictly dominant over Curve if we set tout to 0.04% (assuming 1 DAI = 1 USDC). This is potentially undesirable as it may cause liquidity on Curve to dry up although it’s worth noting that most of the APY is coming from CRV farming so it may not matter that much. Curve may also decide to lower their fees in response.

There is likely a “sweet spot” where we can improve volume for the PSM without affecting Curve liquidity.

USDC Implications

The closer we get to a 0% tout, the quicker the USDC will drain from the Maker collateral pool. This may be more or less desirable for some.

Peg Implications

By tightening the tout parameter we will end up draining the PSM faster than if we didn’t. This will provide less of a buffer in the event of a sudden DAI demand drop-off. It’s unclear if this is a problem, but it’s a consideration none-the-less.

Poll

Should we lower the PSM-USDC-A tout parameter?
  • No, we should leave it at 0.1%
  • Yes, we should lower it to 0.075%
  • Yes, we should lower it to 0.05%
  • Yes, we should lower it to 0.04% (Curve Fee)
  • Yes, we should lower it to 0.025%
  • Yes, we should lower it to 0%
  • Abstain

0 voters

This poll will run until Thursday, February 18th to be included in an on-chain vote the following Monday, February 22nd if any options gain >= 50% of the vote (ignoring Abstains). A ranked choice vote will be used if multiple options are >= 50%.

6 Likes

Aren’t we planning to farm with some of the PSM USDC stablecoin holdings? What about increasing the tout instead so that we end up having more reserves for a longer period.

There are also benefits to having the USDC PSM have bids further out to act as stability for the peg. By pushing the bids up, you’re also going to encourage a steady state where DAI trades at a higher average price.

Is our goal trading fees or is it long term stability of the peg? Not lowering tout means you deprioritized trading fees but with the goal of healthier DAI price support in case of tail events as a results of DAI price downside.

And the USDC isn’t entirely unproductive if we’re not earning trading fees, there’s also the opportunity to earn APY from farming crops.

2 Likes

I don’t really see anything settled about the farming, and even if we did want to do that it’s not trivial to get that code on chain. We would need to get another round of audits, and I do not want to be the sole developer overseeing that like with the PSM. At minimum a farming USDC is months away, and I don’t think it’s clear MKR holders want to do that.

Personally I would be okay with sticking some portion into Aave to earn interest or something instead of farming, but we need code, audits, consensus, etc first.

2 Likes

Sounds like that just adds more overhead/risk to MakerDAO. Probably smarter to focus on using PSM as a peg stability module and less on another savings account.

2 Likes

@Aaron_Bartsch Agreed, but there’s also benefits to for peg stability with lower bids. Peg stability is not about maximizing transaction fees, there’s also benefits for utilizing the USDC reserves for peg stability from peg downside risks. It could be prudent to not have USDC reserves be utilized too quickly. USDC reserves have some strategic importance compared to DAI reserves as we can’t mint USDC.

3 Likes

For some background on DAI on Curve. There’s currently $1.05B DAI on Curve (Mainly M1 DAI, roughly ~138M M2 DAI). This is a majority of outstanding DAI. Not everything needs to be a competition.

Also the PSM will get run-over if there’s ever an unwind.

I’m confused. I thought that the PSM was a stop-gap to help the peg until we could create enough DAI to reinstate a positive Dai Savings Rate (DSR). I oppose adding any extra utility to the PSM that would delay or disincentivize a positive DSR.

2 Likes

As long as DAI is close to peg does it really matter how much reserve we have? A healthy peg is a healthy protocol right?

1 Like

@Joshua_Pritikin DSR sounds fun. Though the 84M USDC in the PSM isn’t really preventing us from increasing the DSR. If we’re trading at 0.995 we could decide to increase the DSR even with USDC reserves.

So I wanted to elaborate a little on the Curve thing. While I have no idea if it’s a good idea to price trades below them in the long-term, in the short term I don’t think it’s likely to cause a major loss of liquidity from the Curve pools. Mainly because:

The APY for 3pool (which sees the most volume of the stablecoin pools) is ~19% in total (without boost.) Only 2% of that APY comes from stablecoin trades.

Another point is that the PSM is bounded by how much USDC is present. If the tout is lower than tin and the PSM empties, it can’t take any more volume until more USDC is traded in, which may not happen immediately given the usage we’ve seen so far.

4 Likes

To reinstate the Dai Savings Rate (DSR), we need to generate DAI from interest bearing assets. The PSM is unlikely to be a sustainable source of income. The fees are structured to generate income from trading, not from generating DAI. If we raise the DSR and then some miscreant generates 500M DAI with the PSM just to collect the DSR, it’s an abuse of the system.

1 Like

Thats why we have debt ceilings, stability fees and tin/tout fees right? I’m not saying put the fees to 0% and I agree we must stay cognizant of the amount of USDC exposure we have and should try to minimize that exposure whenever possible.

I’m not sure what we’re arguing/talking about anymore lol.

Yes, the PSM will only be more attractive for converting DAI->USDC for a short period, and then we won’t have any more USDC reserves.

(Though some people may consider that a good thing)

1 Like

If we raise the DSR, I’d consider slightly raising tin on the PSM to make it slightly more costly to buy DAI from the PSM. If we end up having an actual prolonged peg problem we can lower the DSR and lower tin on the PSM back to 0.1%.

Isn’t this the kind of thing that woud make us want to move from the PSM to Curve-joined vaults?

The APY? Possibly. That would be a whole other proposal though. It would also further expose us to USDT and/or Compound/Aave smart contract risk (in addition to Curve contract risk.)

I would not worry about farming for now. If it comes we can change the parameter.

How I see it, it is more like :
short range around ±0.02% dai/stablecoin without conversion. depending of the volume but ~ 100M to absorb volatility.
than long range, long term assert cDai/aDai/cUsdc/aUsdc … if we have the tools in place, probably ±0.1% more as cover.

Sorry yes I meant the APY. Is Curve riskier than Uniswap? Also yes it does expose us indirectly. A stablecoin pool without USDT would be better. Still eventually I think any USDC we’re ready to have will be better in a joined vault with APY giving us a cut than in the PSM.

1 Like

I think it is more appropriate to gradually reduce the transaction fee, it is best to reduce it to 0.075% first, and then observe.

We will not get any traffic unless we skeeze it even more. The Dai didn’t go down.
I personally think that even with 0.04% we won’t get traffic at all.