Meta: cap total ether supply at ~120 million

[vbuterin]

Author: Vitalik Buterin
Category: Meta
Published: 2018 Apr 1

In order to ensure the economic sustainability of the platform under the widest possible variety of circumstances, and in light of the fact that issuing new coins to proof of work miners is no longer an effective way of promoting an egalitarian coin distribution or any other significant policy goal, I propose that we agree on a hard cap for the total quantity of ETH.

During the next hard fork that alters reward distributions (likely phase 1 Casper), this proposal requires redenominating all in-protocol rewards, including mining rewards, staking interest, staking rewards in the sharding system and any other future rewards that may be devised, in "reward units", where the definition of reward units is:

1 RU = (1 - CURRENT_SUPPLY / MAX_SUPPLY) ETH

I recommend setting MAX_SUPPLY = 120,204,432, or exactly 2x the amount of ETH sold in the original ether sale.

Assuming MAX_SUPPLY = 120 million, and given the current supply of 98.5 million, that means that 1 RU is now equal to 1 - 98.5m/120m ~= 0.1792 ETH, so if a hard fork were to be implemented today, the 3 ETH block reward would become 16.74 RU. In one month, the ETH supply will grow to ~99.1 million, so 1 RU will reduce to 0.1742 ETH, and so the block reward in ETH would be 16.74 * 0.1742 = 2.91555.

In the longer term, the supply would exponentially approach the max cap and the rewards would exponentially approach zero, so if hypothetically Ethereum stays with proof of work forever, this would halve rewards every 744 days. In reality, however, rewards will decrease greatly with the switch to proof of stake, and fees such as rent (as well as slashed validators) will decrease the ETH supply, so the actual ETH supply will reach some equilibrium below MAX_SUPPLY where rewards and penalties/fees cancel out, and so rewards will always remain at some positive level above zero.

If for some reason this EIP is adopted at a point where it is too late to set a max cap at 120 million, it is also possible to set a higher max cap. I would recommend 144,052,828 ETH, or exactly 2x the total amount released in the genesis block including both the sale and premines.

[aunyks]

This is a huge sign of the project's maturity! However, I'm not sure that I see a substantial advantage to fixing the supply. Is it simply to introduce monetary scarcity, or am I missing a key component?

[OperationNine]

I think a round number such as 120,000,000 would be easier for people to calculate, similar to Bitcoins 21,000,000. Not sure if exactly 2x is entirely necessary. But I think it's great to have a proposal discussing economic suggestions of this sort!

140M might be seen as too high in some cases. For example BTC circulating supply is currently 16,951,300 with a max of 21,000,000. That is 23.88% more in total inflation from todays numbers.

ETH circulating supply is 98,545,046 with a theoretical max supply of 120,000,000 would mean a total of 21.77% more in total inflation.

140M would give us 42.06% more inflation (close to double the rate of BTC)

I read an article quoting you as saying: “Introducing some kind of sinks into ethereum is definitely something we’re looking at. By sinks, I mean fees that lead to the token actually being destroyed.”

Is this something that could still potentially become apart of Casper? Wouldn't this then lead to a decreasing supply over time determined by how much the network is being used?

I like this idea a lot personally because it could be said that this would give ETH the properties of increased security based on the network usage increase. The more Ethereum is being used you would have some baked in security increase going hand and hand with this.

As the incentive to 51% attack might increase from more value being placed on the network through more transactions, the more difficult it would become on its own as ETH becomes scarcer.

[sammy007]

Why not obtain a banking license and stop messing with chains?

[Buttaa]

yes. what I am curious about is, what motivated you to request a cap now and not at the beginning?
what changed your mind? @vbuterin

[bokkypoobah]

It's only because ETH price is tanking

[th4s]

April's fool?

[smartcontracts]

Please excuse my ignorance on the topic: what are the arguments for either side?

[WowSyler]

ik

[Souptacular]

Needs to be Standards Track EIP rather than Meta per EIP 1.

[bumerang007]

a great idea, it will give an opportunity to correctly distribute the rewards of all participants in the network

[tbrannt]

@ButtaTRiBot

I have seen this suggestion several times now. I think it makes sense as it adds an easy social contract. A hard number that's not really subjective.

[apertamono]

@OperationNine: 123,456,789 would be easier to remember.

[lichnosam]

Виталька , переходи к делу.

[zangheri]

Once estabished a non controversial MAX_SUPPLY, we could further reduce it by a % (from 0 to 100) of empty blocks mined. Must define what is an empty block (% of full capacity).

[negamax]

It's great that ETH community is discussing about a hard cap. This will ensure better economic incentives and propagation. I propose halving built into this at 91st day of the year. So somewhere around 1st April

[3esmit]

I don't see this have any to do with "price".
Also, this is not decided until now because PoS economics are not yet defined.
This is important for the economics, specially for the case of "burned ether for paying storage" that can be "reissued" by validators, and together with the rent fee is a brilliant way to solving, seems like going to work!

[Arachnid]

There's not currently any in-protocol mechanism for determining current supply. How would this work?

[ktechmidas]

April fools I think?

[nyancodex]

Stop the troll plz =.=

[MaxFriedman17]

First, I would like the confirmation that this is not due to April's fool. I do not think it is but anyways.

Second, I think it is a must for any cryptoasset to decide the Total cap (or if there will not be a Total cap, then decide the number of tokens to be created yearly as soon as possible). It is not "serious" (maybe not the correct word) to have an asset with intrinsic value without knowking how or at which pace it will be created. That is actually what current FIAT does.

Regards.

[blocxsjm]

How about bringing out Casper first Vitalik? That should solve your issues, not a hard cap.

[narmirzaei]

It’s April 1

[mohsenghajar]

This, the ether sinks, and casper, all three are needed imo.

[ethereum-pos]

PoS will may calibrate another parametrs, but fix 120m supply more clearly for innoncent-crypto-minds. In another case I would suggest building Bitcoin emission on erc20 token format. eBitcoin but with PoS emitation of Bitcoin PoW, nice joke for 1apr)

[prestonvanloon]

@Arachnid nodes evaluate the state and reject chains that violate the max supply condition.
It must be possible if etherscan and others can determine total supply.

[Arachnid]

@prestonvanloon There is no consensus figure for total supply at present. Etherscan et al compute it for themselves based on historical block rewards (not practical for a fast-synced node).

[prestonvanloon]

@Arachnid I agree. Maybe we could add supply property to new blocks.
The full/fast sync nodes could agree that block N+1 increased the supply by X and block N had a supply of M and M+X does not violate the total supply. However, it would be difficult for non-archival nodes to reach consensus about the established supply.

[Arachnid]

Regarding the idea as a whole (and treating it as serious for the moment): I think introducing a hard cap needs better justification than this.

The way I see it, network costs (eg, security) can be paid for out of inflation or fees, or a combination of both. Paying costs out of inflation encourages use and discourages HODLing, while paying costs with fees has the opposite effect, disincentivising transacting. All else aside, I'd prefer to fund using inflation for that reason.

[DanielRX]

@prestonvanloon Is the increase in supply not linked to the number of uncles the block has? So it would be a non-constant amount, meaning a fast node needs to verify each block and the uncles?

[realcodywburns]

I'm assuming this is a total joke. You would need a new opcode like #700 and nearly completely revise how block reward is calculated at the same time. Non-trivial tasks. Also doesn't account for stuck or burned ether.

Edit: For clarity, having a hard capped supply is an incredibly bad idea and is only good for pumping the price of a coin because of 'muh scarcity' . Unless a method can be articulated of the knowing exact consumption statistics of ether in the year 2053 on March 2nd(or any arbitrarily selected day) it is foolish/ponzi-ish to suggest that they will only need n tokens especially with >50% of the tokens already having been premined.

[AliAshrafD]

Disclaimer:
@ctpang my latest paragraph, asking for putting an end to protocol manipulation was not meant to hurt you, not even about you. Obviously you are not the one who is about to committing such a manipulation and as I understand you are against the idea of putting a cap.

It was an official request issued on behalf of the majority of Ethereum community, or at least an important fraction of it, to Ethereum Foundation and VB.

[amchercashin]

I think we can't avoid protocol manipulation because monetary inflation is not resolved problem.

At the start of the ethereum the model with constant positive monetary inflation was chosen, like 18m of coins annually. I can understand reasons behind this: the lack of knowledge of system behavior and lack of time to make decision.

Maybe that's not bad for the start. But this is the delay of problem. Constant positive monetary inflation effectively leads to zero monetary inflation in the future. So it's not that different from this proposal to make it zero much faster.

I will say again, I think everybody would agree that monetary inflation policy is critical question for all the possible risks to the Ethereum: from blockchain security to people and companys incentives to actually use Ethereum.

Why the policy should be chosen arbitrary, like let's change from constant to decreasing? Why the parameters chosen arbitrary like let's decrease it to zero this exact pace? This question should be studied at least with the same effort like secure of PoS or Sharding implementration.

Ethereum has advantage to be not the first concurrency. We can observe how earlier crypto like bitcoin behave (from mine narrow point of view it's 99% speculation, 0.9% illegal goods (and there is better crypto for that already) and 0.1% - pizza buying accident : ) rtally I doubt the amount actually used to buy goods. but i happily change my mind if there are different statistics)

Isn't it better to brainstorm all available options like:

• constant, with different k

• decreasing

• linear dependency on available variables, for example why not to tie it to transaction fees: like miner gets fee from user and the same (or not same) amount from issuance so everybody pays: like half who made transaction and half all other except this miner (actually the intention for this to make that second part balance the increase in "production", so nobody "pays"). So more transaction means more issuance, zero means zero. Or maybe better not to fees but to transacted value? of both?

• or some non linear dependency?

Write out pros and cons. And maybe then discuss what's better?

[AliAshrafD]

A wise man has said:

A system is called persistent if it can survive in spite of the infinite absence of its creator(s).

I'm not sure who is that wise man and I appreciate any help to find him. Wanna make a temple in his name ;)

Ethereum, a persistent system or an under development experimental project , subject to deconstructionism? This is the question!

Just one comment to be taken in consideration: Those who reject the first in favor of the second option, defining Ethereum as a project, have the least right to use this EIP as an evidence for their commitment to help the price to go to the moon. Other than members of a closed circle, the believers, who else would buy such a coin with no promise and commitment to a protocole? The currency of an un-persistent crypto system.

[bumerang007]

All look https://www.etherchain.org/coinvote/poll/35

[dsyeag]

I agree with @Arachnid 's point that funding network costs through issuance is desirable to avoid ETH hoarding. In the interest of the long-term development of the protocol, I think adding an issuance-based reward for the development team makes sense as well. ETH hodlers should be happy to be diluted by some small % a year for development efforts that add significant value to the network.

The tough part is clearly developing governance around distribution of the developer reward, but I think it's a problem that could eventually be solved (through a DAO mechanism or maybe a competing node implementation model where different dev teams reap reward proportionate to blocks mined using their implementation). Anyway, I think once a hard cap is introduced, it will be very difficult to remove from a political perspective, so long-term development sustainability is worth considering here.

[unknown1235]

I strongly believe having a 120 million hard cap will be a great benefit for the Ethereum community and developers in the long run.
It will keep inflation down and reduce the high volatility that we are currently experiencing. Overall this will strengthen Ethereum Community and developers as a whole.

[kameir]

The question will always be whether a value is being created (or represented) through new ETH. I get that functioning as currency was not the initial purpose of ETH but it is the de-facto currency for virtual assets today. From a practical standpoint, ETH might, therefore, benefit from relating more to the current implementations of money. This way it seems more likely to find widespread adoption. The most obvious shortcoming for the later is ETH's divisibility into 18 decimal places. It would be impractical at this point to create price-labels for everyday items - which would be great to see.

[rp3599]

The 120 million hard-cap will be the best thing that can happen for community as a whole. And ethereum will also serve as a investment commodity and will attract more investors and more innovations and more apps to make use of valuable dapp currency.

[umurb]

Still disagree. Putting a hard cap would only have a short-term effect and make us all happy for a limited amount of time. If we really want web3.0 to succeed, we should tame our greed and wait for mass adoption. I am certain that one day the world will be complaining about why Ethereum is not generating more Ethers a day.

…

On Fri, Jun 15, 2018 at 4:50 PM, rakeshkumar ***@***.***> wrote: The 120 million hard-cap will be the best thing that can happen for community as a whole. And ethereum will also serve as a investment commodity and will attract more investors and more innovations and more apps to make use of valuable dapp currency. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#960 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AjyYvL42IHqacpc1_xJyrd_JwuWIt3CEks5t9B4mgaJpZM4TCuGY> .

[adamluc]

I have thought quite a bit about the supply cap issue for ETH, and I believe there should be a supply cap. The way I approached this topic was thinking through the primary components of what makes ETH have value and what the impacts would be if a supply cap is implemented. The components I utilized as a framework are:

1. Medium of Exchange -> The ability to use ETH as a currency/asset to exchange for other goods and services. This needs to be high velocity, and eventually accepted as a legal tender.
2. Store of Value -> The ability to store ETH and exchange at some point in the future with ETH retaining purchasing power vs. other currencies/assets.
3. Computation Cost -> the cost of computation, which is denominated in 18 decimals and determined by market ( ETH gas). In my view, this should be kept on par with other systems executing transactions on one blockchain vs. another (for similar types of transactions, eg. Sending value or arbitrary data, or verifying a zk-proof). This will be more important when blockchains start to connect more to one another via technologies such as Cosmos and Polkadot.
4. Security of the Network -> The value of ETH is correlated to the security of the network, e.g. -> how much would it cost to attack the network and either fork it or destabilize it?

There are many other components that can be leveraged, however I thought it is best to find 3-4 primary components to derive a framework to think through the issue.

If a supply cap is not enacted and ETH is maintained as an inflationary asset:

1. Medium of exchange: Using ETH as a medium of exchange would be impacted positively. ETH would have higher velocity, allowing for ETH to move more freely within the ETH ecosystem.
   a. Legal tender here is not impacted in my view by ETH supply remaining uncapped, legal tender would need to be adopted by governments.
2. Store of Value: Supply and demand dictate price, and if there is an increasing amount of ETH inflation without increased demand (demand may be strong over the next few years, however there will likely be a point of diminished demand at some point in the future), price will drop and ETH as a store of value of will be diminished. An ETH today will likely be worth more than an ETH tomorrow, similar to how government backed currencies operate.
3. Computation Cost: In some instances, velocity may impact the computation by increasing cost however, from a macro perspective, I suspect this will not be impacted much in the long-run from continuous inflation.
4. Security of the Network: Since the Ethereum network will be moving to proof of stake, the price of ETH impacts it’s security, eg. If the price drops over time, the cost to attack the network will be lower.

If a supply cap is enacted and ETH is maintained at a finite supply:

1. Medium of exchange: Using ETH as a medium of exchange would likely be impacted negatively. ETH would have lower velocity, effectively reducing the amount of ETH traversing the ETH ecosystem, and thus ETH’s utility.
   a. Legal tender here is not impacted in my view by ETH supply being capped, legal tender would need to be adopted by governments.
2. Store of Value: Supply and demand dictate price, and if there is a finite amount of ETH with increased demand, price will increase and ETH as a store of value will be fortified. An ETH today will likely be worth the same or less than an ETH tomorrow, counter to how government backed currencies operate.
3. Computation Cost: In some instances, velocity may impact the cost of computation by increasing the cost however, from a macro perspective, I suspect this will not be impacted much from slightly decreasing velocity. Having 18 decimals I expect would help here.
4. Security of the Network: Since the Ethereum network will be moving to proof of stake, the price of ETH impacts it’s security, eg. If the price increases over time, the cost to attack the network will be higher. This fortifies ETH’s position as a system that will be secure against sovereign entities.

Overall, I see little downside in placing a cap. A cap would increase security of the network, as well as provide a store of value independent of government backed currencies. If the question comes down to medium of exchange, perhaps derivatives can be introduced to pay for computation in the network or be utilized for payment of goods or services, similar to Maker’s CDPs. Additionally, perhaps the 18 decimal limit could be increased to accommodate higher prices.

[jamesray1]

Well a supply cap is included in the shasper spec: https://notes.ethereum.org/SCIg8AH5SA-O4C1G1LYZHQ.

[tbrannt]

@jamesray1 good. In the end it's the only thing that makes sense. If we want that token that we pay our security with to have value we should not punish people who hold it.

[ethereumdegen]

ITT: People trying to determine how to make as much money off ETH as possible instead of trying to improve the Ethereum Network. Just let ETH be the gas of the network. It does not need to be a speculative goldmine. It cannot be everything at once, a jack of all trades is a master of none. ERC20 tokens will be better currency anyways according to Vlad.

…

On Thu, Jul 19, 2018 at 6:20 AM, tbrannt ***@***.***> wrote: @jamesray1 <https://github.com/jamesray1> good. In the end it's the only thing that makes sense. If we an arbitrary token that we pay our security with to have value we should not punish people who hold it. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#960 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AF9bL5jKnxX00kUPbYpQKuHeFxgG_2RMks5uIF1egaJpZM4TCuGY> .

[vjune]

Anyone who knows how this proposal goes now? Regardless the value of ETH, I feel bad because of the challengers such as EOS and TRON whose structure, system & idea are a copy of ETHEREUM.

[jamesray1]

First of all I should emphasise that the total supply cap is not as important as the underlying usefulness and therefore value of the blockchain. EOS and TRON are centralized, undermining security due to a smaller attack surface / fewer nodes, which comes with a host of issues such as censorship, cartelization, rent-seeking, etc.

But to answer your question, AIUI there will be a total supply cap in eth 2.0. MAX_CASPER_VOTES is 2^10=1024 validators per shard, the MAX_DEPOSIT is 2^5=32 ETH/validator, and the SHARD_COUNT is also 2^10=1024. So by this the total supply cap is 2^15=32,768 ETH per shard and 2^25 in total = 33,554,432 ETH. However, I must be missing something, as the total circulating supply at present is 103,895,066.69 ETH, so >70 M ETH would have to be burnt to get that number. However, dimensional analysis holds that 2^10 validators/shard * 2^5 ETH/validator * 2^10 shards = 2^25 ETH = 33554432. Note that a token sale with a new Eth 2 blockchain seems out of the question, due to sentiment about token sales (particularly from Vitalik, Hsiao-Wei, etc., as touched on in talks). Perhaps the SHARD_COUNT is initially 2^10 at launch, then increased to 2^11 and 2^12 (IIRC it has been discussed going up to 4096 before, as well as doubling the SHARD_COUNT as the network usage grows).

CTRL+F for these parameters in https://github.com/ethereum/eth2.0-specs/blob/master/specs/core/0_beacon-chain.md.

[b1-88er]

What is the status of this EIP?

[vkli]

Q: Who will mine the blocks after reaching 0?

A: Holochain (p2p/scalable vs. PoW/PoS/blockchain distributed computing system, like https://holochain.org)

[truongnmt]

Wonder is this April Fool or not! lol

[jamesray1]

While I agree that Holochain looks more promising than Ethereum, this thread and GitHub generally isn't the place for promotions. @vkli

[tvanepps]

Wonder is this April Fool or not! lol

Yes, this was a poorly considered April Fools joke at the time.

[leonardge]

Any update on this EIP?

[tvanepps]

Any update on this EIP?

this was an April fools joke

[github-actions]

There has been no activity on this issue for two months. It will be closed in a week if no further activity occurs. If you would like to move this EIP forward, please respond to any outstanding feedback or add a comment indicating that you have addressed all required feedback and are ready for a review.

[github-actions]

This issue was closed due to inactivity. If you are still pursuing it, feel free to reopen it and respond to any feedback or request a review in a comment.