On bitcoin's very long term future without miner rewards

[omarino]

I've been thinking about the future of bitcoin in the decades to come. Like many of you I think that bitcoin will become the world's reserve currency, if it weren't for one little problem: miner reward. I know this problem has been discussed in the past, but I have not seen much discussion in the present. There is a great paper on the topic https://www.cs.princeton.edu/~arvindn/publications/mining_CCS.pdf, which shows how unstable the ecosystem becomes once block rewards are taken out of the equation. Fees will not replace rewards in the sense that incentives change. How do you think it will affect future development, and when do you think this will have the first concrete manifestations both in network dynamics and price?
Do you think it will fork in other versions where there is always a tail reward like in other cryptocurrencies? If that were to happen it would destroys bitcoin's case as having a finite supply.
These issues are far ahead and I'll probably be dead before they will actually become tangible issues, but I think it's important to discuss anyways.
I'd love to hear opinions from someone who's more knowledgeable than me, so thank you in advance for any reply.

[Charles-Tim]

Bitcoin is a deflationary currerency, which means the marketcap and price will be increasing. Although, the price over a short time can decrease, but Bitcoin nature is inevitable which will make the price to have a net increase over long time period. This alone will make the price to be increasing in a way the price as at then when all bitcoin are already mined to be increasing because of limited supply. Because of the price increase, transaction fee should be enough to make miners to continue mining bitcoin.

Do you think it will fork in other versions where there is always a tail reward like in other cryptocurrencies? If that were to happen it would destroys bitcoin's case as having a finite supply.

Anything that will result to finite supply of bitcoin to be altered can not occur, it can only lead to hard fork while bitcoin will remain Bitcoin with a total supply of 21 million. And I think bitcoin miners (community) will not agree with such because increasing the total supply of bitcoin will only lead to price devaluation and depreciation when there is no need, it has postitive effect for bitcoin price to appreciate.

[omarino]

Bitcoin is a deflationary currerency, which means the marketcap and price will be increasing. Although, the price over a short time can decrease, but Bitcoin nature is inevitable which will make the price to have a net increase over long time period. This alone will make the price to be increasing in a way the price as at then when all bitcoin are already mined to be increasing because of limited supply. Because of the price increase, transaction fee should be enough to make miners to continue mining bitcoin.

As is proved in the paper, an environment where fees are the only source of revenue for miners makes selfish behavior not align with the correct functioning of the network, unlike what happens in the present.

Anything that will result to finite supply of bitcoin to be altered can not occur, it can only lead to hard fork while bitcoin will remain Bitcoin with a total supply of 21 million. And I think bitcoin miners (community) will not agree with such because increasing the total supply of bitcoin will only lead to price devaluation and depreciation when there is no need, it has postitive effect for bitcoin price to appreciate.

Yes, I meant a hard fork which could happen if enough miners want solve the detrimental problems that I've been talking about.

[gmaxwell]

The paper mistaken, because it analyses an abstraction which is too different from the real system. In particular it disregards that blocksize is limited, which is a factor known to address this instability since at least 2011. (There are other countermeasures as well, such as locktime based anti-fee sniping)

https://medium.com/@bergealex4/bitcoin-is-unstable-without-the-block-size-size-limit-70db07070a54

Do you think it will fork in other versions where there is always a tail reward like in other cryptocurrencies?

Absolutely not.

a hard fork which could happen if enough miners want

When it comes to hard forks what miners want is completely and utterly irrelevant. The consensus rules the users use are what decide who is or who isn't a miner.  If someone violates the rules individually and autonomously by the users then they just aren't a miner anymore, they might as well be sending viagra spam instead of blocks for all the users nodes would care about them.

I'll probably be dead before they will actually become tangible issues,

I wouldn't be so quick to assume that. The subsidy decline is geometric and will become small long before its zero.  About 10% of miner income comes from fees today.  And an income stabilizing backlog is no longer a theory but a practical reality.

[omarino]

The paper mistaken, because it analyses an abstraction which is too different from the real system. In particular it disregards that blocksize is limited, which is a factor known to address this instability since at least 2011. (There are other countermeasures as well, such as locktime based anti-fee sniping)
https://medium.com/@bergealex4/bitcoin-is-unstable-without-the-block-size-size-limit-70db07070a54

I honestly couldn't understand what different it makes on the matter whether block size is limited or not. Regarding the countermeasure, a miner could just use a modified code for the node to get rid of it. It's a countermeasure only on the surface.

When it comes to hard forks what miners want is completely and utterly irrelevant. The consensus rules the users use are what decide who is or who isn't a miner.  If someone violates the rules individually and autonomously by the users then they just aren't a miner anymore, they might as well be sending viagra spam instead of blocks for all the users nodes would care about them.

What I ultimately meant is that if the fork is for the good of bitcoin in the long run, it will inevitably be used. I'm talking about forks with the purpose of saving bitcoin (if it were to become necessary), not forks based on different views like bch.

I'll probably be dead before they will actually become tangible issues,

I wouldn't be so quick to assume that. The subsidy decline is geometric and will become small long before its zero.  About 10% of miner income comes from fees today.  And an income stabilizing backlog is no longer a theory but a practical reality.

Yeah I guess it's going to be gradual rather than sudden.

[gmaxwell]

I honestly couldn't understand what different it makes on the matter whether block size is limited or not.

Because when you have mined a full block and there is a backlog, the difference in income between mining the next block (all full of pending transactions) and going backwards and remining the prior block to take its fees is small (and comes at a cost of decreased chance of eventually being in the longest chain).  This is particularly true because new transactions that are arriving cannot be included in the prior block if it is remined.

Regarding the countermeasure, a miner could just use a modified code for the node to get rid of it. It's a countermeasure only on the surface.

No, if a miner violates anti-fee-sniping their block is invalid and will be painlessly ignored by the network.

[omarino]

I honestly couldn't understand what different it makes on the matter whether block size is limited or not.

Because when you have mined a full block and there is a backlog, the difference in income between mining the next block (all full of pending transactions) and going backwards and remining the prior block to take its fees is small (and comes at a cost of decreased risk of eventually being in the longest chain).  This is particularly true because nee transactions that are arriving cannot be included in the prior block if it is remined.

Ok yeah this might mitigate the problem.

Regarding the countermeasure, a miner could just use a modified code for the node to get rid of it. It's a countermeasure only on the surface.

No, if a miner violates anti-fee-sniping their block is invalid and will be painlessly ignored by the network.

Remember that in that environment the selfish behavior would be for every miner to modify its code to take advantage of this strategy. One would have to hope that most miners wouldn't do that for the good of the network overall but that's not selfish behavior.

[HeRetiK]

Do you think it will fork in other versions where there is always a tail reward like in other cryptocurrencies? If that were to happen it would destroys bitcoin's case as having a finite supply.

Any change which violate Principles of Bitcoin almost certainty will be rejected.

Regardless of the Principles of Bitcoin such a hard fork could happen. It's just that it most likely won't be accepted by the market since no one would want to buy such a coin over BTC.

Remember that in that environment the selfish behavior would be for every miner to modify its code to take advantage of this strategy. One would have to hope that most miners wouldn't do that for the good of the network overall but that's not selfish behavior.

If it's enforced on the protocol level, there's no way around it other than forking of into an alt coin.

[omarino]

I doubt every miner will collude to increase block size stealthily

I wasn't referring to the block size but to the countermeasure mentioned in the medium article, that is

Wallets supporting this will create transactions that can only be processed on top of the current block, severely undermining the chances of successful “fee-sniping”.

[gmaxwell]

Wallets supporting this will create transactions that can only be processed on top of the current block, severely undermining the chances of successful “fee-sniping”.

Miners can't violate that, it's physically impossible.  If they break the locktime rule their blocks are no longer blocks and get ignored. Even if 100% of previously active miners do so, that just means is that they stopped mining.  It's no different from a technical perspective than if they all decided to award themselves an additional 100 BTC per block.

The anti-snipe protection isn't especially powerful, but it's just another tool in the set... it amplifies the incentive protections that exist from the risk of orphaning vs the small marginal gain that exists in the presence of backlog.

[Apollo Avanis]

21 million coins won't be the final limit. Majority will eventually agree to keep btc block rewards at a sustainable level.

[tromp]

21 million coins won't be the final limit. Majority will eventually agree to keep btc block rewards at a sustainable level.

No; we'll just see 2 camps of people. The supply-cappers and the tail-emitters.
Both agree that mining rewards should always be very significant, in order to achieve long term security and stability.
These two conflicting visions also see different outcomes of inevitable coin loss,
which I believe is very unlikely to drop below 0.1% and may be closer to 1% on a yearly basis.

Capped Supply

A supply cap requires a block reward dwindling to insignificance (long before it reaches 0).
So the significant mining reward must come from large fees. Mining stability requires a steady backlog of fee paying transactions, which in turn requires a highly constrained block size. Coin loss will erode the circulating supply over time
and will eventually necessitate subdividing the smallest unit.

Tail Emission

With an eventual (or even from launch) constant block reward, fees are only needed to deter spam, and block size is less constrained. It could be more dynamic to deal with surges in transaction demand, or slowly grow over time. Coin loss will put a softcap on circulating supply.

People who stop believing in the 21M bitcoin cap could of course try to fork away, but such a fork would very much go against Bitcoin's main principle of immutability. and would likely fare no better than altcoins that embraced the tail emission from the start, which is probably what those people would instead gravitate to.

Bitcoin cash is interesting for being in the Capped Supply camp, while being unwilling to highly constrain the block size. Which looks like a recipe for long term insecurity and/or instability.

[KIZILAGA]

  i think Satoshi made a time-adjusted self destruction design here.İntentionally He put a limit to cap supply becuz he knew the mentality of money flows in this 0-negative rates area,printed trillio dollars,euros.. A small part of this money will be located a scarce supplied financial instrument, higher demand than supply..So it will gain popularity and adoption with this rapid Price increase until no one can denies the existence.A 600 billion mcap monster.

 Admit it,none of us,mostly wont be get involved in the idea or tech if Bitcoin didnt accelerate price with limited cap. Miner and asic firms wont invest if it was stayed low priced.We are all price and profit oriented,we want to beat inflation and censorship. Yet,VC,Miners,Asic firms,and you..already discussing,thinking. 21 million +%3 inflation or %5 inflation ,blocksize etc.


  Miners and everybody needs rewards. 1.5 million left coins. it is time detonated. Satoshi knew it and trapped the whole financial system.Bitcoin is an only spearhead.

  Real hard money is Gold. Still living at the every dark corner of the world.

  An unlimited supply,miner rewards for thousand centuries. Still inflated but less.İmmense liquidity.Perfect design tested for 1000 years.İmmortal.

  Bitcoin is monetary policy  immutable,which makes it Bitcoin. There is nothing you can do. When you change it, you are like ecb,fed,boj.And if stays 21 million this way again it dies slowly.



  
 

[HeRetiK]

 i think Satoshi made a time-adjusted self destruction design here.İntentionally He put a limit to cap supply becuz he knew the mentality of money flows in this 0-negative rates area,printed trillio dollars,euros.. A small part of this money will be located a scarce supplied financial instrument, higher demand than supply..So it will gain popularity and adoption with this rapid Price increase until no one can denies the existence.A 600 billion mcap monster.

 Admit it,none of us,mostly wont be get involved in the idea or tech if Bitcoin didnt accelerate price with limited cap. Miner and asic firms wont invest if it was stayed low priced.We are all price and profit oriented,we want to beat inflation and censorship. Yet,VC,Miners,Asic firms,and you..already discussing,thinking. 21 million +%3 inflation or %5 inflation ,blocksize etc.

I think it's undisputed that its scarcity is part of Bitcoin's success. Personally I feel that trailing emission is definitely worth discussing as an alternative monetary policy (in the context of alts). But without Bitcoin gaining traction due to its limited supply, causing a Cambrian explosion of alts, we'd never reach the point where we could discuss this issue in a practical sense in the first place.

Actually I'd even go as far as to say that the abrupt supply shocks caused by the halvings are part of the plan. An asset affected by sudden surges is more likely to gain attention than one that's constantly growing. And the four years between halvings seem to be just enough time that people can recuperate after a crash without Bitcoin being completely lost out of public view, making it the perfect FOMO trigger.

[Paouky]

The block reward concerns all feel high up in the air for most people. That is, until the we reach a specific period where the rewards are so abysmally low compared to the value being moved on chain, that transaction finality actually becomes a material concern - this will be unfamiliar and frightening for many, undoubtedly causing disunity.

I sincerely believe we can anticipate a chain-split revolving around this issue come early/mid 2030's; One side would stick to the supply cap and the other side would eliminate all remaining halvings, in effect creating a tail emission of 0.78 or 0.39 per block. It is practically going to be a battle of whether intermediary tools should provide transaction finality, or the chain it self.

[odolvlobo]

21 million coins won't be the final limit. Majority will eventually agree to keep btc block rewards at a sustainable level.

What do you mean by "sustainable"? How do you determine what is sustainable, and how do you know that fees won't provide it?

I sincerely believe we can anticipate a chain-split revolving around this issue come early/mid 2030's; One side would stick to the supply cap and the other side would eliminate all remaining halvings

Why would there be a split? I don't think there is enough information to justify either side at this point, and it might become clear which way is better as more information is gained. For now, your prediction is based on nothing.

I think that there are two major questions left to answer:

1. What is the optimal maximum block size?
2. Will fees alone be enough to secure Bitcoin?

I don't think that anyone knows the answers to those questions yet.

Regarding the optimal block size, I think that there needs to be a way to incentivize miners to reach an equilibrium. Technocrats dictating a policy is the absolute worst way to approach it. An unlimited block size might be the answer, but I suspect that it would not as it would turn into tragedy-of-the-commons problem.

Regarding security through fees alone, there is no way to know if fees are enough without knowing what amount of fees are necessary to secure the block chain, how much people will be willing to pay, and how big the blocks will be.

[pooya87]

I think that there are two major questions left to answer:

Actually the more important question is not about any of that, it is all about the price of bitcoin. Saying block reward won't be enough to secure the network is actually saying that bitcoin price is not going to be high enough (or would drop hard) to ensure hashrate remains high enough to make attacks impossible due to extremely high costs.

So far we have no reason to believe that. With 50BTC per block reward miners were making less than a dollar. Last cycle when they were making 12.5BTC per block that was between $8k to $75k and today it is $230k for only 6.25BTC per block.
You see bitcoin doesn't even need the fees to keep the incentive so much higher than needed for security, not only that but also this won't change any time soon.

This means even if we cut the reward by half today to 3.125BTC (instead of in a couple of years when price could be a lot higher than now) the miners would still be making $115k which is a lot higher than what they were making when the reward was 12.5BTC.

[odolvlobo]

I think that there are two major questions left to answer:

Actually the more important question is not about any of that, it is all about the price of bitcoin. ...

Price only matters as long as there is a subsidy. I assume that when we talk about long-term, we are talking about post-subsidy.

[pooya87]

I think that there are two major questions left to answer:

Actually the more important question is not about any of that, it is all about the price of bitcoin. ...

Price only matters as long as there is a subsidy. I assume that when we talk about long-term, we are talking about post-subsidy.

The quote was talking about 2030 so it was more of the short-long-term. And the post-subsidy is not for about 100 years, I find it kind of moot to discuss it now since we simply can't know how anything would be in 2100 for instance!

[figmentofmyass]

Actually the more important question is not about any of that, it is all about the price of bitcoin. Saying block reward won't be enough to secure the network is actually saying that bitcoin price is not going to be high enough (or would drop hard) to ensure hashrate remains high enough to make attacks impossible due to extremely high costs.

it doesn't seem optimal to depend wholly on price, as is the case with the large block size/low fee revenue model. as block rewards drop and drop, miners will become increasingly sensitive to price changes, potentially leading to large swings in hash rate.

to prevent that, fee revenue must rise to dampen the effect of the falling block subsidy. that means restricting block size, probably rather conservatively.

You see bitcoin doesn't even need the fees to keep the incentive so much higher than needed for security, not only that but also this won't change any time soon.

this is the early adopter phase when the block subsidy is still relatively high. how about in......12 years or 16 years? then the issue of block size will be extremely important because fees will be the primary source of mining income.