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 https://github.com/ethereum/EIPs/pull/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.

[rfikki]

Not sure if this is a serious proposal, but I will bite....

This would be a major change to the platform, the OP does not seem to provide any specific reasons for wanting this change to the platform, other then a general statement about economic sustainability.

What are the primary effects on the economic sustainability of the platform vis-a-vis a max supply of MAX_SUPPLY = 120,204,432 vs. MAX_SUPPLY = 144,052,828, would it make sense to agree on any alternate number(s) for any reason? If no change is adopted, what are the drawbacks as to economic sustainability? What is total supply and how is it determined currently?

[prestonvanloon]

@DanielRX I don't know the correlation of uncles to supply, but I imagine they are related. It would always be a non-constant amount since transaction fees would differ each time.

[ctapang]

I am glad you published this on April 1st, @vbuterin. I hope you are just kidding because this is going to be a very crucial decision for Ethereum. I do not agree that this will be good for Ethereum, unless you want ethers to have the same fate as bitcoins. Bitcoins can never become a medium of exchange because its exchange rate with respect to commodities will always be trending up, with occasional corrections. The same thing will happen to ethers if the quantity is fixed. The question for you is this: do you want the currently most popular cryptocurrency to be elitist, forever an object of speculation, or do you want it to become a common medium of exchange, something we all can use in our daily lives, to buy lunch, to buy groceries, to pay for rent, etc?

Related article in Medium: https://medium.com/@carlostapang_15490/cryptocurrency-as-media-of-exchange-761dce398517

[scottscheper]

In order to ensure the economic sustainability of the platform under the widest possible variety of circumstances

I would like some examples so I can better understand how a hard cap would help achieve this

[naga2727]

april fools still not done yet??

[ethereumdegen]

Why not just keep ether capless and use tokens as a medium of exchange with scarcity? For example use scarce/capped PoW mined decentralized tokens or even PoS scarce/capped tokens. If you need an example, a few already exist. That way ether can be the network security mechanism and won't need to also fill the role as a monetary exchange mechanism / scarce speculation asset. I think trying to make Ether more like Bitcoin may detract from it's ability to secure the network. It's hard to engineer something to fit many roles. It's easy to engineer something to fit a very specific role. This topic requires more research however.

[pkrasam]

I wonder what will happen if the MAX_SUPPLY = 1.000,000,000 1 RU = (1 - CURRENT_SUPPLY / MAX_SUPPLY) ETH

EIP	CURRENT_SUPPLY	MAX_SUPPLY	RU	ETH
EIP 0.1	98,500,000	120,204,432	0.1806	ETH
EIP 0.2	98,500,000	120,000,000	0.1792	ETH
EIP 0.3	98,500,000	1,000,000,000	0.9015	ETH

[johnEth]

I am in support of a hard cap of 120 million because it will increase the profile of projects on the ethereum network by introducing value and scarcity. Anything that has limited supply by virtue of there being a finite number will mean only quality projects will be supported and rise to the top. Frivolous projects will be squeezed out.

[johnEth]

In short having a hard cap will have the effect of decreasing scammy ICOs and ponzi schemes and only the best projects will be funded. This solves multiple current issues with one stone.

[morok9]

Excuse me,can you tell me the EIP is a new coin?

[ctapang]

@johnEth , I don't see how "Anything that has limited supply by virtue of there being a finite number will mean only quality projects will be supported and rise to the top." Are you saying that there are no expensive scams, that people who invest large amounts of money are not subject to fraud?

[johnEth]

Another perhaps unintended consequence of limiting the supply is to decrease the network traffic and load. By limiting the supply people will want to make wiser decisions before sending their ether over the network. Charging people to use the network will only make them choose another chain but limiting the supply will positively have the same effect without scaring people off.

[johnEth]

@ctapang, no. This is simple economics 101. When you cap the amount of fuel (ether) over time, people are more careful where they spend their fuel. Thus they will due better due diligence, think before sending their limited supply of fuel. So over time quality projects will rise to the top. In effect, this is brilliant. Only the best and worth while projects will be funded and used. Ethereum will be the best quality network and all everyone will want to use it of course but only the best will be able to afford to be on the network.

[ctapang]

Limiting quantity makes sense only in the beginning, when a cryptocurrency is attracting attention and subsequent holders. As soon as a cryptocurrency has achieved popularity and has gained so must trust (as Ethereum has), then its quantity should be determined by its velocity, or how much it is used in daily commerce. I believe the timing for going into Proof of Stake is perfect: Ethereum has gained the most number of adherents, and now its quantity must be determined by velocity alone. Proof of Stake (PoS) should allow the network to pay "interest" on ethers that are parked in PoS "bonded validators" that confirm transactions. These parked ethers represent economic activity: when a merchant accepts ethers, she deposits extra ethers earned in such accounts, when we get paid in ethers in our jobs, we also deposit extra ethers we earn in such PoS accounts, etc. The more ethers we earn, the more ethers can be parked in bonded validators, and more NEW ethers are introduced into the Ethereum economy.

[ctapang]

Putting a hard limit on quantity (like in Bitcoin) makes a cryptocurrency volatile and unfit to be a medium of exchange. Nobody wants to spend (nor accept) a currency that changes its value drastically on a daily basis. Stability is a requirement for any medium of exchange.

[mohsenghajar]

Network security is ensured by mining and mining rewards are paid in eth. I guess the idea is to cap max supply so that the price goes up, and/or to attract more miners and gain increased security. Not sure what good does e.g. a capped supply ERC20 token do.

On Sun, Apr 1, 2018 at 10:33 PM, Admazzola notifications@github.com wrote:

Why not just keep ether capless and use tokens as a medium of exchange with scarcity? For example use scarce/capped PoW mined decentralized tokens or even PoS scarce/capped tokens. If you need an example, a few already exist. That way ether can be the network security mechanism and won't need to also fill the role as a monetary exchange mechanism / scarce speculation asset. I think trying to make Ether more like Bitcoin may detract from it's ability to secure the network. It's hard to engineer something to fit many roles. It's easy to engineer something to fit a very specific role. This topic requires more research however.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/ethereum/EIPs/issues/960#issuecomment-377839994, or mute the thread https://github.com/notifications/unsubscribe-auth/ANH0wh4Qhh4IvU95K1VefDoMapNijfCIks5tkY4QgaJpZM4TCuGY .

[johnEth]

A capped supply confers value in the medium of exchange, eth. Supply is a non issue even with a cap because each eth can be broken down to a million pieces so there will be plenty of eth to do what ctapang wants to do. I believe the Proof of Stake and a cap of 120 million eth combination will knock down multiple issues with one stone. This actually is an incredible brilliant idea from Mr. Buterin.

[ctapang]

@johnEth , I've heard your reasoning before, from Bitcoin enthusiasts. Yes, it is true that an ether is fungible into smaller units, wei being the smallest unit. However, that is a solution to something that is not a problem. It is possible to move the decimal point either way, and that is as expected. The problem is volatility versus stability. All currencies are also subject to the law of supply and demand: if you limit the supply in an environment with an ever-increasing demand, naturally the exchange rate will always trend upwards, with occasional, violent corrections along the way.

[johnEth]

Volatility is not due to limited supply. Fiat, gold, silver they all have a limited finite supply. The American dollar has a certain number of units circulating around the globe so volatility is not the result of limited supply. Volatility is due to a rapidly growing space.

[ctapang]

In the end, it's up to the Ethereum leadership how they envision the ether cryptocurrency to be: do they want it to be a medium of exchange, like most fiat money is now, or do they want it to be a store of value, like gold or bitcoins? I would like to use ethers to pay for my lunch, groceries, rent, etc. I would like to receive my pay in ethers also. But if it becomes like Bitcoin, then it cannot be a medium of exchange.

Note that many companies have tried to use Bitcoin as medium of exchange, and it has not become one, after almost nine years.