Tokenomics Proposal [Community Consideration]

After reading the entire thread I have to say thank you to all of the people putting some serious time into thinking through their ideas and presenting them here. Also thank you to the Dfinity Team for demonstrating that they want to move in the direction of the community driving the future of the IC.

Given the current discussion above, this is a summary of my current stance on tokenomics:

  1. Like @wpb and some others have said I agree that we should not make any changes to the current programmatic mechanics of the tokenomics. I think more time should be given to capture some metrics to determine how the tokenomics are actually playing out. Without some metrics on the key variables, making changes is somewhat uninformed (if not completely random). This is dangerous because the “levers” and the strength of those levers aren’t fully understood.

  2. I agree that more metrics should be made transparent. The primary ones being:
    a. Percent of ICP locked up over varying time periods (this would answer some of the behavioral
    questions that @lastmjs has asked above),
    b. A measurement of neuron liquid voting power so that a “take over” of the IC can be defended
    c. Some analysis should be completed to confirm whether it is possible to reach ICP price appreciation
    (or even stable) state through the mechanics of the deflationary/inflationary mechanics. @Trevor
    asks some good questions on this above.

  3. I’m really on the fence regarding unclaimed ICP. I can’t decide if it should just be left alone, burned, or re-distributed. This is a really tough one that probably has some legal issues that vary by country/jurisdiction.

  4. Due to the security reasons expressed, I think we should do whatever we can to avoid the creation of a Neuron trading market.

These are some really tough concepts that will take a lot of smart people debating in order to come to the best possible outcome.


I fully agree with this. I think the tokenomics model is working just fine


How do people still think it’s good token economics. I was just reading for days but could not resist anymore ICP has terrible token economics.
First, let’s look at this;
Dscvr thread: here you go fools .
and, now the current situation is for a 4 years seeds (the ones bought each 0.03$) going to dump on current buyers. It’s like a leaking boat and if there is not any investment coming people going to lose 70% of their money at the end of 4 years. Besides any gain, it’s really risky, and when more people learn this they are going to avoid investment on icp, without the help of the community, icp is going to leak more like a snowball effect. Even after 4 years foundation etc is going to mint 300M coins? is it like a joke? nothing is transparent about token economics since day1 its understandable with that much-broken level. I believe that ICP not going to survive well with that broken token economics people are not blind as creators of this thinks.
Honestly, I have no idea how to fix this but maybe stopping the aging of seed neurons and increasing the maturity rate of neurons created after launch as 2-3x maybe can lead to a bit equal distribution.
There should be 75-80% of coins on exchanges, not 10%. It was a broken launch, who launches with 0.005% coin on the first day and make their coin on the top 3 as there was 40$B investment on it at the first day (probably the real market cap was around 200M$) and with showing circulating supply 127M and dump other 10-20M coins weeks later. It’s has terrible token economics and with this level of tech, it could be in the top 3 for real atm. Really sad these happened to ICP.


@icpmax, I’m having a tough time following this, but I think your primary critique is about the token distribution–not the tokenomics. We can’t do anything about the distribution of ICP at this point. And that’s not what this thread is about. This thread is about tweaking the set of incentives around the ICP token to meet some desired purpose (whether it is to increase the security of the network, to increase demand for token, etc.).


For most, I disagree with @icpmax. Those seed investors (which I am not and have bought my few first ICP at 320$) had invested 10k, 30k or whatever (at .03 cents each ICP) back years ago when Internet Computer was only on paper, not much more then a dream. These investors had the most faith that the dream would come true. Without these investors, IC would just not exist. Now, if these investors believed enough to pour thousand of dollars in a dream, why would they sell now that IC is alive and work amazingly? These people believe much more in Crypto then fiat money. Why would they sell? To buy other cryptos? or get back to fiat? Where will they invest their fiat money obtained from that dump? What is better then ICP as investment for the next 20 years? If you have a clue, let me know, I may be interested.
I just believe (but may be wrong) these seeds will invest most of their new obtained ICP in the NNS and have great returns, while the price would go up.
I don’t see many other ways (other then Dfinity have done) to obtain funds for a project on paper only.
Our risk to invest in ICP now, a great working layer 1 network, is millions times less then for these seeds investors at the time.
What I found to be bad at launch wad the lack on clear information, which is a general statement and complain as I can see. Now, we have to live with it and go forward, if we believe in the technology of course, which I do from the very beginning I learn about IC. (When ICP was selling at 320$ back in May)
This is my thinking but respect yours.


If the penalty were so severe that it ate into the principal, and if the totality of said penalty were burned or distributed to, say, only locked neurons, then MAYBE I’d agree. But even then it’s a maybe.

And regarding neuron transfers… one thought is for neurons to be transferrable to other II’s that are subordinate to my master II (i.e., created by me via my master II). And perhaps there needs to be a rule that a neuron can only be transferred to a subordinate II that has been in existence for a year (or some other arbitrary time)… or perhaps the % of the neuron that can be transferred is dependent upon the age of the subordinate II.

I’m not a fan of transferring neurons, except for legacy/estate needs.


Anyone who believes neurons should not be transferable is making the same mistake as Bitcoin’s decaying block reward: it is a bad security assumption for the future.

If an external blockchain implements chain key crypto and private smart contracts, one will be able to create neurons on addresses controlled by the external blockchain. Then later one could transfer ownership of this neuron on the other blockchain, while on the IC it appears nothing happened. Only the buyer of the neuron would know the transfer happened, preventing NNS voters from blacklisting transferred neurons.

If the security of the IC/NNS depends on neurons not being transferrable, it faces a ticking time bomb of another blockchain implementing CKC and private smart contracts. If just 1% of locked ICP wants to be sold annually, that is still a huge market to build a product for. Then an attacker could still perform whatever governance attack they want, while only losing the discount on locked ICP. The discount will still be non-trivial, because neuron sellers have to incentivize buyers over just buying their own ICP and staking. Plus, the sale won’t happen instantly, so the attacker still risks the price of ICP declining after a bad proposal is passed.

The counter to the governance attack is for neuron buyers to be aware of the voting status of the neurons they wish to buy. Buyers will not want to buy neurons that are actively voting because of the risk of becoming a bagholder after a governance attack. I expect neuron markets to prevent listing of neurons for sale that are actively voting to save buyers the trouble of manually looking that up.

If the IC directly allows neuron transfers, it could implement some light rules/restrictions to enhance security, that are not enough to incentive the creation of an external neuron market. For instance, adding small delays to voting/following/transferring actions could prevent attackers from instantly voting then liquidating a neuron. Or create a special neuron state that allows it to be transferred but not vote. It would be better for the IC to have some control over neuron markets than for them to have free reign on external blockchains.


I personally think that other projects like Polkadot raise some sincere questions regarding the tokenomics and governance:

Here is the link on Polkadot’s governance mechanism.

To vote, a voter generally must lock their tokens up for at least the enactment delay period beyond the end of the referendum. This is in order to ensure that some minimal economic buy-in to the result is needed and to dissuade vote selling.
It is possible to vote without locking at all, but your vote is worth a small fraction of a normal vote, given your stake. At the same time, holding only a small amount of tokens does not mean that the holder cannot influence the referendum result, thanks to time-locking. You can read more about this at Voluntary Locking.


Peter: Votes No with 10 DOT for a 128 week lock period => 10 * 6 = 60 Votes

Logan: Votes Yes with 20 DOT for a 4 week lock period => 20 * 1 = 20 Votes

Kevin: Votes Yes with 15 DOT for a 8 week lock period => 15 * 2 = 30 Votes

So that raises for me the first question: Why are we afraid of shorter lock-up periods? Polkadot shows exactly how this could be solved.

If one doesn’t lock up his DOT for a vote, he’ll only get voting power with a factor of 0.1. = > 128weeks against zero delay equals a factor of 60(!) in voting power weight difference.
That means someone who locks up 1Dot for 128weeks has got the same voting power as someone with 60DOT.
In my opinion that should be enough to actually counter the argument: “Only long-term holders should be able to vote/benefit from staking etc.”
The huge advantage of a system like Polkadot is the that it actually let’s everyone benefit. ICP feels more like a punishment with the 6month lock-up atm. I understand the arguments, but in comparison with Polkadot we should ask ourselves: Why?

I don’t have a good answer for this. Polkadot differs between staking and governance voting. I know that it’s not the case here. However even staking is possible with a lock-up delay of 28days on Polkadot.
And I actually don’t see any argument why that shouldn’t be brought to ICP.

If we don’t find proper solutions to these questions, we’ll probably end up like cowards. The only argument I can find is that a minimum lock of 6month, brings stability to the price, since ICP are longer out of circulation.
That’s also the complete opposite of what we should oppose to the fiat system: If we make it hard to liquidiate, because we’re afraid that the price could go down, what does that say about the system and ourselves? Are we that afraid that the free market could punish us for some flaws?
It’s like a country embracing capital controls in order to prevent capital flight and help the exchange rate…
That’s normally a completely desperate act. Look at some countries with high inflation and ask yourselves how that turned out for them.

Next thing: 6month minimum lock up in combination with 1ICP treshold for spawning a neuron
Who benefits more? Someone with 1000 ICP or someone with 10ICP?
Someone with 1k ICP and an annual inflation rate of 10% would generate ~8.3 ICP a month or 2 ICP a week. With 7 days delay for unlocking the new neuron, he’d actually have much liquid ICP available. He could restake them again to earn interest on his interest (making him benefit even more), or he could also sell them, which completely counters many of the arguments. I’ve already described it here.

The staking guide explains it:

There is a “Spawn Neuron” button that is grayed out until I have at least 1 ICP in rewards. You can compute how many newly minted ICP you have by taking your maturity % and multiplying by the number of ICP you have staked in the neuron. Once your ICP rewards are greater than 1 ICP it will let you spawn a new neuron with the newly minted ICP staked in it. Once you spawn a new neuron it will have a 7-day dissolve delay after which you can either send it to a different wallet or add it to your existing staked wallet.

Someone with only 10ICP would have to wait much longer to be able to earn interest on his interest. Futhermore he would also have less liquidity than someone with a high stake, earning the 1ICP really fast.
What did the person with “only” 10 ICP do wrong? Right, he didn’t have as much money or wasn’t as lucky as someone else to get such a high amount of ICP that early. Now he is punished, and the whale benefits exponentially. I don’t have to explain that even the current banking system is more equal since we all get the same low interest rate.
I personally see it this way: A while could dump on the market with his freshly minted ICP (staking reward) while I still can’t spawn a new Neuron. Here you are clearly prefered if you have a high stake or to turn it around:
=> One actually gets punished for only having a small stake.
IMO this is an oligarchy on steroids.

Neuron age is an interesting concept, however if we bring it back to the inequality discussion it’s not clear if that isn’t exactly the opposite of the wanted effect.

Let’s take Dfinitys stake (Idk how high atm): Probably 20% of the Genesis supply
If that is staked => 10% a year.
20% on ~400Mio ICP => 80 Mio ICP
10% a year => 8Mio ICP a year (8Mio ICP @60$ are 480 Mio$ a year from staking only(!))
Now additional benefits due to Neuron age (one of the first stakers and 1ICP treshold liquidity, earning interest on interest) => Dfinity can easily pay staff from staking only. How should that thing ever become decentralized with that whale tokenomics?

I have to disagree with your statement:

we should not make significant changes the tokenomics.

You’ve written, that you wanted actual data first. Fair way to go. However I don’t read that there is sensitivity on this matter. I read many comments which say the following: Wait, let’s watch it. etc.
And that’s what really concerns me.

Having bad tokenomics is one thing, but not seeing the flaws is a whole other story. I’ve actually never seen a project like that: Without doubt the greatest tech. Without doubt horrible tokenomics, and I clearly stand to my opinion (my provided facts should speak for themselves). @Trevor wrote another great comment in here about some more flaws.

If ICP wants to succeed it will need great tech, great tokenomics and community support. Putting it on the long bench and thinking that this will somehow fix itself, will let the problems appear squared in a few years. By embracing an oligarchy, ICP loses points to tokenomics and this thing will come back like a boomerang after years. Benefiting the whales will let us lose community support (think of our fiat system). Maybe not today, but it will get worse with each year. Look at the wealth distribution in the US. ICP can do this within 5 years. If you don’t believe me, take some numbers and calculate them in Matlab via interest of interest on high stakes.

tldr: IMO ICP tokenomics is horrible. Without major governance/staking upgrades, the system could set itself up for failure.


  • Polkadot governance should be studied and partly copied to ICP
  • 1 ICP treshold for spawning a new Neuron should be lowered, or completely abolished (better).
  • 6 month minimum lock-up period should be significantly lowered
  • Neuron Following should be decentralized to more parties
  • More transparancy regarding staking stats is required (already in discussion in here?)
  • Discussion on how to avoid oligarchic structures is required
  • Discussions about the treasury should be established (e.g. treasury votes with ICP?)
1 Like

Thank you for sharing your thoughts on these topics. I respect your opinion. I just wanted to say that I agree with one of your proposals, which is that the 1 ICP threshold to spawn a neuron does disproportionally benefit larger token holders who are compounding their voting rewards and are willing to manually spawn, dissolve, and top up as quickly as possible. For anyone that is not compounding, it makes no difference as the reward rate of return is equal (assuming equivalent dissolve delay and age) no matter how many ICP are locked. In practice, I suspect a lot of folks are compounding and want the compounding to happen automatically. It is my hope that the automatic compounding feature that Dfinity foundation is working on will be implemented on a daily basis so this issue goes away. If it doesn’t address this issue, I think a strong case should be made for enabling compounding to equally benefit everyone no matter how many ICP are staked. There are probably multiple ways of accomplishing that goal and I would advocate for Dfinity foundation to pursue that type of proposal regarding the tokenomics.

Edit: Another way to give large and small token holder equality from compounding would be to change the definition of when the Spawn Neuron is allowed. The fact that it is pegged to 1 ICP is what creates the inequality. If it were pegged to Maturity (the ratio between ICP reward and ICP staked) then that inequality would go away.


I believe that in order to achieve the maximum number of stacked neurons, it is necessary to offer a non - proprietary value. Let’s think about it? What can be valuable for a true fan of the IC ecosystem? Allocations in startups. We need to somehow encourage true stakeholders by providing guaranteed allocations in projects that are interesting for them. What can be the model? Let’s fantasize. You can add the Noblesse parameter equal to the Maturity parameter. From which it will be possible to spawn a sub-neuron. This sub-neuron will be able to vote for a certain project and get a proportional guaranteed allocation. After voting, the sub-neuron disappears. It is important that these neurons cannot leave their host to avoid speculation and market valuation. I believe that it will also be beneficial for the project owners.
First of all: this will give them the power of public recognition.
Secondly: the owners of strong sub-neurons cannot be weak hands

There is an expression Canadian people are saying: “Don’t try to use the shovel before it snow”. We can assume and try to solve tons of problems that may never arise. Good for discussion though. My stance, for what it is worth, is wait and see how it goes and solve problems as they arise. So far, I am only impressed about IC technology, Dfinity and every one involved (developers, etc…). Neurons rewards were design to rewards more investors with long term view. But short term are quiet well serve as well to me. Whales (I am not at all) are whales for a reason. They believe and invested in a paper only system. Now, I would think they have all incentive to have IC to work at his best, to get their investment at his best as well.


I’d like to point out Diego explained most ICP minted are from neuron rewards not node providers, so part of my concerns are alliviated.

I think personally there are several different topics here being discussed and they should go into relevant threads. I would call mine “Inflationary rate.”

While I’m aware the ecosystem will lead to ICP getting burned, it would be good to get some numbers behind that so the math works out. I think this ties into better understanding of costs and what it would take to have tokens burned = tokens minted with a 5-10% inflation rate plus node provider fees.


I absolutely agree on your view that automatic compounding doesn’t solve the issue. It actually further embraces the problem. Now you’ve to take a look on your neuron once in a while. High stakes might forget to spawn new neurons. There is also human work (still really low amount) in the process, which slows it a bit down.

With automatic compouding they can lean back while the ICP itself helps them to use the oligarchic structure perfectly with zero additional effort. That’s insane.
No wonder that whales want automatic compounding => The rich want to get richer, and that as easily as possible. I’ve written that exactly in my other thread.

Furthermore what I didn’t mention is that the period for reaching the 1 ICP gets smaller and smaller, the higher the staked amount gets. => That even puts more dynamic into the exponential curve for whales.
Imagine an automatic script which does that after a few votes/days.

Could you explain your statement more exactly, pls? It’s interesting:

If it were pegged to a ratio between ICP reward and ICP staked then that inequality would go away.

The problem with this is, that someone with a high stake, could simply split his neuron up a few times and avoid getting into the problem. That might incentivice to get smaller neurons and put less dynamic into the exponential curve. However if you take a look at my example in the other thread, you’ll see that the difference of someone having 2 neurons with amount/2 or one neuron with amount is not that severe.

It will get interesting if someone is forced to split up his neurons to an amount equal of a smaller stakeholder. However this would need serious calculations, and further analysis if there is a way to simply avoid it. If someone with 1k ICP will get the same percentual reward after the same amount of time as someone with 10ICP, the system is fair (excluding Neuron age at that point).

Hey @MichaelAnnh, I just want to clarify what I was trying to say in the previous post…

I am in favor of automatic compounding.

I am in favor of compounding methods that ensure reward rate of returns benefit large and small stake holders equally. Daily automatic compounding achieves this goal, but there are manual methods as well.

If automatic compounding is not going to be implemented, then pegging the manual Spawn Neuron action to a fixed Maturity threshold instead of a fixed ICP reward threshold also achieves this goal of equality. However, pegging to a fixed Maturity creates new problems such as how long it will take to achieve that Maturity.


Wow - a lot of very thoughtful discussion here. In separate posts to come soon, I’d like to offer my own two cents. But first, I thought it would be useful to all involved here to try to summarize ideas raised so far (apologies in advance for not tagging all of who have advocated support for given ideas - the forum here doesn’t seem to permit tagging more than 10 members in a single post):

Some key ideas raised to this point:

  • 1. Re-calibrating ICP inflation and burning parameters - to achieve today something closer to a non-inflationary supply of ICP (@Trevor, @BrianGaller - who advocates for some study to be conducted to see whether this is needed).
  • 2. Providing the ability to transfer neurons to another - to the beneficiary of an estate if need be (@wpb), but also potentially to anyone else at any time, possibly with penalties (@Alixthe, @talkingant).
  • 3. Creating visibility into “liquid voting power” - both to provide general awareness of where control of the NNS lies at any given moment (@wpb), and also providing clarity regarding who owns how much ICP (@Trevor).
  • 4. Finding better ways to achieve the 90% staking goal - both to decentralize faster and in the process achieving certain guarantees against 51% attacks (@wpb).
  • 5. Supporting shorter ICP lockup periods in staking (@MichaelAnnh).

Some other good ideas raised to this point:

  • 6. Fine-tuning staking reward mechanics - namely, automatic compounding of staking rewards, and/or changing the threshold for spawning new neurons, so as not to disadvantage smaller ICP stakers (@wpb, @MichaelAnnh).
  • 7. Providing greater control over one’s own neurons - i.e., combining and/or splitting neurons (@wpb).
  • 8. Ensuring that centralized exchanges reflect only circulating supply of ICP (unlocked supply), rather than all supply (@tlbalog).

Please do let me know if I’ve in any way inadvertently misrepresented anyone’s position here, and I will undertake to adjust. Thoughts on some of these topics to come soon.


On point 1. - recalibrating ICP inflation and burning parameters:

I believe that current parameters are quite nicely calibrated for the steady-state economics that we might anticipate for the IC. I’ve reached out to both @Trevor and @BrianGaller directly on Telegram to share thoughts, and invite anyone who may be interested to connect with me on Telegram (not sharing here, as thoughts have implications for ICP’s price, and this Forum isn’t meant for price-talk).


On point 2. - providing the ability to transfer neurons to another:

I definitely think it’s worthwhile to consider how to transfer to the beneficiary of an estate. This would provide peace of mind, and if designed nicely, could, I would imagine, be made hard to abuse.

On the idea of enabling neuron transfer in general, the security implications are worrisome. But I think @talkingant raises some really interesting points. It hadn’t occurred to me that, if another blockchain platform were to adopt chain key technology, users of that blockchain could create transferable neurons there, with no way to know or mitigate on the IC itself. I’m quite curious to learn more about how this would work. Because if neuron transfer really can’t be prevented, then yes, much better to prepare for the inevitable, sooner rather than later, and find other ways to mitigate the unavoidable risks of neuron transfer.

Still, even if chain key’s adoption by another platform would enable neuron transfer, it will most likely take a long time for other platforms to adopt chain key. So we have time to figure out how to manage associated risks.

So I think we’re still better off at the moment aiming to keep neuron transfer disabled, until we can find workable ways to mitigate risks - risks that really could undo the IC, if not managed properly.

Only once we’ve found appropriate mitigations - and just as soon as we have - then I think it would be great to enable neuron transfer… I think the novel creative opportunities represented by neuron transfer could be significant… just need to make sure we do this in a safe way.


On point 3. - creating visibility into “liquid voting power":

I think this is a great, uncontroversial idea that should be pursued without hesitation.

Same goes for point 4. - Finding better ways to achieve the 90% staking goal.


On point 5. - supporting shorter ICP lockup periods in staking:

I think this is interesting, and worth considering… so long as voting power is sufficiently small for trivial lockup periods so as to prevent interested parties from swooping in to successfully make a quick gain at the expense of those who are invested in the IC’s long-term success. I think this idea deserves more thought.


Finally, I think all of points 6., 7., and 8. are good, fairly uncontroversial ideas:

    1. Fine-tuning staking reward mechanics - namely, automatic compounding of staking rewards, and/or changing the threshold for spawning new neurons so as not to disadvantage smaller ICP stakers.
    1. Providing greater control over one’s own neurons - i.e., combining and/or splitting neurons.
    1. Ensuring that centralized exchanges reflect only circulating supply of ICP (unlocked supply), rather than all supply.