Negative DSR / TRFM

Has anyone definitively confirmed that implementing a negative DSR is not feasible and if so why? It has now been proven beyond a doubt that launching a synthetic asset without a proper funding mechanism was silly.

If we can’t introduce that funding mechanism via a change to DSR, then the least socially damaging version of TRFM should be considered as an interim measure. Minting DAI 1:1 with USDC is not a valid peg-fixing solution.

A negative DSR also removes a lot of complexity from the system because MKR holders can choose a DSR SF spread that will be their profit, while the market will set the actual DSR and SF values. This guarantees all risk taken by MKR holders is paid for, unlike the current model where we have 140M of ETH debt not producing any income for us.

With a spread and negative DSR, we could fix the spread to 2% and the market might set the following rates:

  • -3% DSR
  • -1% SF

With these rates all will be resolved and harmony between parties achieved. MKR holders get paid for taking risk, DAI minters get paid for arbing, DAI holders pay for the privilege of owning DAI and whatever value that provides them.


If no one deposits DAI into DSR.

The DSR being mandatory is implied

1 Like

If you do this, DAI holders will gradually shrink, and the negative effect will be great.

It would reduce demand for dai, which is exactly what we want. The discussions we’ve had before about negative dsr imply it would be a major rework of existing contracts, but it does seem to be one of the less disruptive solutions to our problems. I am in favor if we can confirm there is a feasible way to implement a negative DSR.

P.S. If we set the DSR to a large negative value doesn’t it directly negate most of the pressure from comp rewards? Say if we set it to like -10% or -20%? I don’t necessarily think this is what we should do, but it is nice that we would have that much control over external tampering with our supply and demand incentives

If you want to limit the demand for DAI, you can solve this problem by significantly increasing SF (for example, SF reaches 15-20%), but this will shrink the DAI market.

We should print more DAI, this is the only correct choice. Now we cannot limit the demand for DAI or stimulate the demand for DAI.

Therefore, printing more DAI is the right choice. How to print more DAI is a problem we must study.

Let’s clarify some terminology before replying:

DAI Supply: CDP holders mint DAI against collateral.
Stability Fee: Paid/Received by CDP holders
DAI Demand: DAI holders buy DAI to use as they see fit
DAI Savings Rate (DSR): Paid/Received by DAI holders

To increase the supply of DAI, you must decrease the stability fee to make minting new DAI more attractive. To decrease the demand for DAI, you must decrease the DSR, making it less attractive to hold DAI. Currently, the DSR is at zero and we cannot make it go negative, sadly.

If this all makes sense, can you please delete your posts so we don’t derail this discussion?

1 Like

I find it hard to believe any integration partner is going to be comfortable integrating dai into their platforms if users loose money using it. The theory may check out, from a practical stand point I just don’t see the idea feasible.


Right now DAI holders are stealing value that should be going to CDP holders. Market forces are not being allowed to work and the hopeless peg situation is the result. Sure we can ignore the problem but it wont be long before other synthetic platforms come and replace us

I understand your point but I think negative rates are not the solution because they kill adoption. At the end of the day there is no perfect solution, only compromises. What we as a community need to define is which compromises are we willing to give without killing the dai core value proposition. Do we want a semi descentralized pegged stable coin with adoption or a pure descentralized stable coin without it? As long as we keep onboarding assets in a fast manner with a special scope in real world assets I stand by the first option.


I personally tend to print some unsupported DAI. Overdraw the future.

1 Like

I would just focus short term on implementing a conservative PSM, fees should go to the surplus buffer. Once and if the peg is in control increase the base rate and start generating a war chest bigger than black thursday. Perhaps include an accrual into the base rate for such purpose.

1 Like

I agree with PSM. I’m just worried that if we still can’t make PEG reach $1 after adopting PSM, printing some unsupported DAI is also an option. Because USDT does not have 100% asset support, but it is still operating normally.

1 Like

I agree. And not only integration partners, but all the current permissionless DeFi apps don’t work correctly with ERC-20 tokens whose balance decreases over time. This was the cause of the Balancer pool drain.

The alternative of having Dai not be worth $1 loses all the Dai PMF, in my view (not to mention, no one wants a token whose USD value shrinks).

1 Like

Good points, the core issues I see with either the negative DSR or TRFM are the following:

Negative DSR:

  • Existing contracts could be broken by the unexpected behaviour.
  • Unpredictable gas pricing for transfers due to the negative DSR needing to be applied.


  • Breaks 1DAI = $1USD

I don’t have good answers for the negative DSR issue but on the topic of TRFM, we can ensure that 1DAI <= 1USD and that long-term it should likely be 1DAI == 1USD. We simply lower the target rate below 1 USD when DAI is above the peg and once the peg is restored, we can raise the target rate back towards 1 USD.

This means that if generally we do not have an excess demand for DAI, we will see DAI trade at 1USD with a positive DSR and SF. However, in extreme events, we are given a powerful tool to restore the peg and then recover our USP (unique selling point, 1DAI == 1USD) once the peg has been restored.

I’d also like to point out that DAI trading at $1.02 is also breaking the social contract around DAI and hurting our USP, so inaction is not an option.

What are the main differences between a negative DSR and the TRFM in this situation where dai is still pegged to $1? Aren’t they kind of equivalent?

The current DAI price of $1.02 shows that users are fine with:

  • the negative (mandatory) DSR of -2% and/or
  • 1 DAI backed by $0.98

They are buying an asset worth $1.00 for $1.02.


Those are not the target users, DAI is meant to be used as a medium of exchange and not as an investment vehicle.

When it comes to monetary policy I think we should look towards Central Banks at their ever growing and serious list of tools.

One of the biggest things that can drive people away from cash is ‘inflation’. While the FED has an inflation ‘goal’ or ‘target’ they have a lot of tools to manage markets (and they don’t even care what the price of a dollar is except as compared to the basket of items people buy - in the form of - inflation). How they manage this is with monetary policy. All driven by injecting or removing USD liquidity from various markets. The FED DOES NOT manage inflation by trying to control the value of the USD by directly interviening in markets as per something like the PSM. How they do control inflation is by directly interviening in the short and long term bond markets (by buying or selling bonds) and thereby driving $$ return rates. (While Maker for a time dictated rates this is becoming less so as secondary markets are growing larger and more mature) The other way is by directly injecting liquidity into various markets. For quite some time up until 2008 crisis what markets the FED would directly manipulate were limited, after 2008 and now covid they literally do and will manipulate a whole bunch of markets now. This is a tool that Maker simply has refused to consider. Creating a secondary reserve fund to be used in direct market operations to buy or sell assets to injection and remove DAI liquidity directly.

As mario rightly points out:

The real question is what does the larger community think DAI’s value proposition is? I honestly don’t have any sense of what that is because we have no mechanics to solicit opinion from the greater community. I have wanted for a long time a means for vault, DAI holders + system operators, developers etc. to be able to give feedback to Maker governance. I don’t know if I posted this publicly but the idea that we would pay people who fit certain conditions to answer surveys I think would honestly be useful for the protocol and as a means to guide the answers to the above question of what compromises Maker governance needs to make to satisfy the community as right now we are just guessing what is important based on a few vocal players. We need to hear the voices of thousands if not 10’s of thousands on these issues. Not 100.

  1. decentralized

  2. 1 DAI = 1.00 USD (on major exchanges)

Everything else in secondary and should be changed if #1 or #2 does not work well.

I don’t think anyone will mind occasional DAI price being ±0.5% off the peg (exchange fees are in that range), but being constantly under the peg or constantly above the peg is a problem. Some users will tolerate 1% off the peg, some maybe even 4% but it should not be considered a normal situation.

Mentioning all advantages/use cases of DAI vs USD (cash) would take too much space.

1 Like