Proposal for Adjusting Polkadot's Inflation System: Reducing Issuance and Complexity

Jonas, have you seen my proposal?

I would like to know your opinion

@jonas @filippoweb3 @alice_und_bob @lolmcshizz

Between you 4 - the best (imo) parameters have been put forward.

If there is consensus amongst you then there will most likely be consensus in OpenGov.

Thank you all for tackling this issue - a bright future awaits.

1 Like

Please donā€™t insult others for having a different opinion.
Keep it civil.

2 Likes

Iā€™ve asked some validators about the topic and it seems likely they would have to increase their commission on nominators for operating safely if the proposal would pass.

2 Likes

I am supportive of the parameters proposed by @jonas in the OP:

There appears to be some confusion as to what these numbers mean so to clarify - wiith these parameters inflation would be reduced from 10% to 8%, of which 85% will go to staking rewards. The Treasury would see fixed inflows of 15% of inflation - or 1.2% of existing supply annually.

Currently stakers are earning 13.54% APY against an inflation rate of 10% - outperforming inflation by 3.54%. The proposed parameters would result in stakers earning 11.62% APY against an inflation rate of 8% - outperforming inflation by 3.62%. This is a win-win scenario.

This balances:

  • the need to reduce inflation (this is one of the hottest topics for DOT for an eternity now)
  • the need for stable Treasury inflow - there are many discussions around budgeting Treasury spends, this is a pointless task if we cannot regulate the inflow as well as the outflow
  • maintains a double-digit APY for stakers - gud number

I think that inflation to some fixed % is easier to explain than a fixed token amount of inflation (the linear model) - itā€™s also more easily comparable to other POS networks where % is the common measurement.

I also believe that weā€™re still very much in the early ā€œgrowth at all costsā€ part of Polkadotā€™s life and therefore donā€™t need to be too aggressive with the reduction at this stage - and should be looking to experiment where possible with this now, before it becomes enshrined and untouchable forever with JAM :slight_smile: Iā€™m still holding out hope that Gav will propose some ā€œhalvingā€ style reduction in inflation as part of JAM implementation as this is far more exciting than a simple linear model.

As a final note I am surprised to see people in the same breath talk about constantly selling their staking rewards for income (proposed 85% of the inflation spend) but also how Treasury spending is having such a negative impact (proposed 15% of the inflation spend) :thinking:

6 Likes

I think it cannot be changed anymore as advanced as this thread is, but I am observing that we are mixing two different discussions here.

  1. What mechanism should we set up for tokenomics and what parameters do we give OpenGov to regulate it?
  2. What should the actual parameters look like?

Of course we need discussion around parameters, because they influence mechanism design.

But our primary goal at the current phase should be to determine the right mechanism. We need a thread that is clearly concentrated on that. And because naturally people want to have a discussion around the actual parameters, this should be a separate thread.

The part about ā€œoutperforming inflationā€ is interesting and I actually wasnā€™t aware of it. Will this relationship continue if inflation is reduced further later?

Regarding your final note:
In most parts of the world, staking rewards count as taxable income (whether sold or not) and must therefore be (partially) sold in order to avoid generating net losses. I also think that the approach of considering them as dividends is legitimate.
The treasuryā€™s expenditure can be interpreted as a relative loss of value for token holders, since it had no deposits of itā€™s own beforehand. Of course, this is only true if they are exorbitant and do not offer any corresponding added value through their sale. If this is not the case, however, treasury expenditure is certainly justified.

2 Likes

Of course, this is impossible. You cannot reduce inflation and get the same staking rewards, especially given that treasury inflow is increased compared to right now.

The current rewards are approximately:
9.5% (inflation going to stakers) / 0.59 (current staking rate) āˆ’ 10% inflation =
6.1 % rewards

After implementing the suggested change, the total inflation is 8%; 1.2% goes to the treasury and 6.8% goes to stakers. This results in:
6.8% (inflation going to stakers) / 0.59 (current staking rate) āˆ’ 8% (inflation) =
3.5 % rewards

So yes, the suggested change will reduce staking rewards considerably.
(Please note that the calculations above use approximate numbers)

1 Like

Can inflation be linked to staking rate?
Making inflation an elastic state?
I think 6% to 2% inflation is a reasonable range.
Polkadot should focus on user experience and developer convenience.
Instead of generating competitiveness with a 28-day unlock time and a high APR.
When value can be effectively enhanced, APR is a supplementary tool rather than a primary tool.

1 Like

I think @jonasā€™s proposal has made a lot of significant improvements over the current system, which is clearly outdated. I am in favor of this proposal. I very much agree with the general ideas that (1) the current inflation rate is high (2) the current design is over-complicated without a clear benefit.

I am rather new to the forum, but I tried to read as many comments as possible in this and other related threads. Something I do think is missing from the discussion in this thread is that higher inflation does not necessarily benefit the stakers. Inflation is a transfer of wealth from holders to stakers, and it does not directly generate any value to the ecosystem by itself. While inflation gives stakers more DOT, the unit value of each DOT is lowered by it. The ā€œoutperforming inflationā€ as @lolmcshizz has pointed out is a better metric to look at real return. Even today, the inflation adjusted of the stakers is only 3.54%. This is lower than the return from money market funds, which is essentially risk-free.

In addition, from an economistā€™s standpoint, it is not even clear that inflation benefits stakers at all. We do need small users for the ecosystem to thrive. But since inflation works against holders, an inflation that is too high might drive small users (and potentially a large number of them) out of the system. This will eventually come back to hurt the stakers.

If you look at crypto data (letā€™s ignore nation states for now), the correlation between inflation and project success is rather low. In fact, many of the very successful projects have very low inflation even from the very beginning.

So overall, I would actually propose an even lower inflation rate (between 0-4%), at least in the long run. I understand the current consideration that some inflation is used to fund the Treasury.

This statement is incorrect. Currently, the inflation-adjusted staking rewards exceed 6%.

1 Like

Even if the inflation-adjusted reward exceed 6%, it is still very low. You can get a higher rate by investing much safer asset classes (e.g. mortgage or corporate bonds) with much much lower volatility in the tradfi world.

My point is not about how accurate this calculation is, but rather that a high staking APR does not reward the stakers as much as it appears.

If we started to utilize coretime buy and sales to pay stakers/validators lets say 20%ā€¦ would we be able to reduce inflation more?

Distributing the proceeds from Coretime sales instead of burning them would effectively increase inflation.

But coretime sales and buys are coming from dot already in circulation, correct? So if 20% went to stakers/ validators and 80% got burnedā€¦ we should be able to lower inflation more than just 8%?

correct me if Iā€™m wrong but inflation is also used to reward validators and stakers correct? so, if we found a way to lighten that dependence by using a bit of coretime buy and sales wouldnā€™t that allow us to lower inflation more?

When I say buy and sales Iā€™m also including coretime purchases on the secondary marketplace. That dot is also supposed to be burned?

DOT holder/staker here.

IMHO I am troubled by having a forum discussion on one of the foundational parts of the Polkadot ecosystem. Its good for discourse but it does not give me the confidence that it will result in proper decisions on this matter.

With all respect to Jonas, thank you for the elaborate work, I find it troubling that just a handful of people are discussing the matter and even an lower amount of verified and qualified experts involved.

There has to be more scientific rigor in this process.

3 Likes

Weā€™ve just finished a live stream. Make sure to check it out.

DOT needs to drastically reduce inflation and set max suply.
Inflation needs to be tied to some mechanism and remain elastic.

I completely agree with your words. The problem is not inflation, but the lack of good projects and activities. I have seen a message many times that polka dot is techies, and only Manta has launched all good projects, and it is going to leave or not develop its Atlantic at all, because, in their opinion, the ecosystem is deadā€¦ Okay, sheā€™s not dead, but it feels like sheā€™s showing the last signs of life. I repeat, the problem is not inflation, but the lack of good projects and activities.
Inflation can be solved later. Why solve it now if thatā€™s not the MAIN problemā€¦

3 Likes

Donā€™t you realize that the high staking rewards influenced by inflation make it difficult for other applications to offer comparable returns, hindering project development? We should lower some staking rewards to encourage participation in other applications, rather than only engaging in staking for higher returns.