Bitcoin doesn't have a fixed supply, but it could. Should it?

[seekererebus]

I've often heard it referenced how BTC has a fixed supply, and that being one of it's greatest advantages. It doesn't however, what it actually has is a decreasing supply with an unverifiable loss rate.

As is pointed out in this article: https://medium.com/@alcio/how-to-destroy-bitcoins-255bb6f2142e There are three ways bitcoins can be lost or destroyed. They can be sent to bitcoin addresses that have no known private key, the private key can be lost, or they can be destroyed by being unredeemed by a block miner. A few thousand bitcoin are verifiably lost to burn addresses and unredeemed block rewards. The amount attached to lost private keys is unverifiable.

There's nothing that can be done about unredeemed block rewards. However, the amount can be perfectly accounted for, and bitcoin core could be updated to guarantee 100% of the block reward is claimed in future; assuming it hasn't already. Burn addresses and lost keys are another matter. This is where the "Should it" part comes in.

We could, if we really wanted to, implement an expiration timer on UTXO's of 21 Million blocks (~400 years, block count chosen for symbolic value). Expired UTXO's would be added to the block reward on expiration, and would apply to all UTXO's all the way back to the Genesis block. Such a timer is long enough that even family trusts planning to hold for tens of thousands of years would experience negligible loss, and virtually zero bitcoin with private keys recoverable by the inheritor of the original holder would ever expire. It would bring with it a massive advantage to the network as a whole: a truly fixed, fully verifiable supply. Not of 21M bitcoin because of already unredeemed block rewards, but very close.

I'm inclined to think the trade-off is worth it. Quite possible I'm alone on this, and I'm certainly unwilling to break from consensus over the matter (not that I'll live long enough to do so). I still think it's worth considering. It's not like there's a time crunch on it; we have to hard fork bitcoin long before this would even matter to fix the UNIX time bug after all.

[Welsh]

Until there's a need, I can't see anything being implemented about restoring lost coins back into the network after a certain amount of time has passed. Besides, 21 million isn't the important figure, the important part of it is it can never increase. It doesn't matter if we only have 18, 17 or 14 million in circulation. The important part is that number never increases beyond 21 million.

Everyone gets a little too involved about owning one whole Bitcoin, but Bitcoin doesn't have to be used like that, hence why we have denominations which could potentially be increased further to allow even more room for smaller amounts.

[franky1]

I've often heard it referenced how BTC has a fixed supply, and that being one of it's greatest advantages. It doesn't however, what it actually has is a decreasing supply with an unverifiable loss rate.

the end supply limit is a known limit. the current PRODUCTION rate is a known amount

the production rate is known to half every 210,000 blocks (~4 years) and doing the math can calculate everything you need to know.

as for then what people do within these limits. well thats within the limits. its not like are creating extra

..
as for wanting to put into rule a way to take funds off keys (even though u suggest after 400 years) goes against the security of requiring key signing in the first place
that would break bitcoin, allowing such a back door even with a massive delay

once you implement a rule that allows spending via not using a key signing a tx message process.. whats next. changing the suggested 400 years into 40 months. ?.. now that would break bitcoin

[mk4]

Until there's a need, I can't see anything being implemented about restoring lost coins back into the network after a certain amount of time has passed.

I've thought of this as well. It's definitely controversial because some people simply prefer holding for long long periods of time, but what if we set the timer to something like 20 years? Is it justified to redistribute the coins from a wallet if there wasn't an outbound transaction in 20 years or more?

[franky1]

Until there's a need, I can't see anything being implemented about restoring lost coins back into the network after a certain amount of time has passed.

I've thought of this as well. It's definitely controversial because some people simply prefer holding for long long periods of time, but what if we set the timer to something like 20 years? Is it justified to redistribute the coins from a wallet if there wasn't an outbound transaction in 20 years or more?

so you* want to steal satoshis stash in just 6 years time... um nah

not your key not your coin. so dont be greedy trying to get some other persons coin just because they have more patience than you*
(*not u personally but "you" as in the content of the proposal)

bitcoins security is the absolute requirement of a private key to sign a tx movement of coin.. any change to that which does not require key signing WILL BREAK BITCOINs security model

..
its as bad as saying
"in 2029 the first 20,000 people to post a picture of themselves with a 'shoe on head' can get 50coins unlocked from a 2009 stash"

what kind of security/value system is that

[Upgrade00]

I've often heard it referenced how BTC has a fixed supply, and that being one of it's greatest advantages. It doesn't however, what it actually has is a decreasing supply with an unverifiable loss rate.

It is one of its greatest advantages. In a global economy which is plagued by inflation, having an asset that is not prone to supply shocks and with a finite supply which is highly divisible is a positive.
Unverifiable loss does not undermine the value of a limited supply, and bitcoins being divisible mitigates the risk of lack of supply.

A system to reinclude unspent UTXOs after 400 years to circulation isn't a bad idea, but if not implemented properly could lead to a spike in supply at the time the coins are included and influence the price. It could also influence the perception of decentralization, if bitcoins can be automatically moved from one wallet to another.

We could, if we really wanted to, implement an expiration timer on UTXO's of 21 Million blocks (~400 years, block count chosen for symbolic value). Expired UTXO's would be added to the block reward on expiration, and would apply to all UTXO's all the way back to the Genesis block.

There would be no block reward at that time.

Is it justified to redistribute the coins from a wallet if there wasn't an outbound transaction in 20 years or more?

I can think of a couple of logical reasons to not have an outbound transaction in 20 years. Not justifiable in my opinion.

[franky1]

20 years 400 years.. doesnt matter. breaking the signing policy of fund movement breaks bitcoin no matter the delay

as for "20 years" (satoshi stash at risk in just 6 years!!. thus bad idea again

when people start working at 18 they put funds into a pensions pot. and dont touch it until they are 68
that is 50 years they dont touch their retirement money.. just saying

[Welsh]

I've thought of this as well. It's definitely controversial because some people simply prefer holding for long long periods of time, but what if we set the timer to something like 20 years? Is it justified to redistribute the coins from a wallet if there wasn't an outbound transaction in 20 years or more?

I'm personally not comfortable with it. I can see the argument for longer periods of time, but still probably wouldn't support it. Honestly, I don't think the loss of coins will be that big of a problem to justify any action taken. If we were in dire needs, then a solution probably would need to be thought of. However, we are far from that, and I can't see us getting there any time soon. There's probably going to be generations upon generations, and Bitcoin might well be replaced by the time that happens.

I think there's definitely moral issues when it comes to injecting coins back into the network after a certain period of time. Bitcoin is attempting to replace conventional ideas of money, and doing something like this would be a step backwards in my opinion, and would question the integrity.

[mk4]

I can think of a couple of logical reasons to not have an outbound transaction in 20 years. Not justifiable in my opinion.

It doesn't need to be a huge transaction. Probably even just a sub $1 transaction sent out just to show that someone is still in control of the wallet once every x years.

But yea, I'm really not that convinced that it's a good idea, but it's debatable and definitely makes a very interesting discussion that's for sure. Even though it's obviously a very very controversial idea.

[sbrys]

There will also probably never be a need to recuperate them. Conceptually enough Satoshis to handle the global monetary system.

[Upgrade00]

It doesn't need to be a huge transaction. Probably even just a sub $1 transaction sent out just to show that someone is still in control of the wallet once every x years.

This would impact someone's privacy. Sending out transactions from an address reveals the public key of that address and to an extent reduces the privacy of the user.
Some would not want be comfortable with that and to maintain that privacy might need to sweep their address every so often.

But yea, I'm really not that convinced that it's a good idea, but it's debatable and definitely makes a very interesting discussion that's for sure. Even though it's obviously a very very controversial idea.

It does make for an interesting discussion and whichever side you're on there are pros and cons to it.

[BlackHatCoiner]

Or we just don't, to maintain this awesome, non-negotiatable, steady inflation rate.

Bitcoin doesn't have a fixed supply, but it could. Should it?

What do you mean it doesn't have a fixed supply? It very much does. Just enter bitcoin-cli gettxoutsetinfo, and it will return you the total coins in circulation your Bitcoin node considers spendable. Every 10 minutes the supply increases in a fixed manner, and the subsidy is halved every 210,000 blocks in also a fixed manner.

[seekererebus]

Is it justified to redistribute the coins from a wallet if there wasn't an outbound transaction in 20 years or more?

20 years is completely unacceptable. The entire idea of this is that unrecoverable bitcoin is recirculated into the economy; no UTXO that could be reasonably expected to have it's private key still known should be recirculated. The reason ~400 years was proposed is because if a UTXO is that old but is still recoverable, it is being held by a seriously dedicated family or company trust. Such a trust could be expected to occasionally self-transfer those funds.

20 years 400 years.. doesnt matter. breaking the signing policy of fund movement breaks bitcoin no matter the delay

I don't know enough about the code to know whether a retroactively applied expiration would unavoidably create a security vulnerability. If it can't be done without such a vulnerability, then I agree it shouldn't be.

I've often heard it referenced how BTC has a fixed supply, and that being one of it's greatest advantages. It doesn't however, what it actually has is a decreasing supply with an unverifiable loss rate.

It is one of its greatest advantages. In a global economy which is plagued by inflation, having an asset that is not prone to supply shocks and with a finite supply which is highly divisible is a positive.
Unverifiable loss does not undermine the value of a limited supply, and bitcoins being divisible mitigates the risk of lack of supply.

An inflationary money increases demand pressure without true demand being affected. Price signals are thus manipulated and are impure. A deflationary money decreases demand pressure without true demand being affected. Price signals are still manipulated and are impure. A fixed money doesn't affect demand pressure at all. Price signals are pure and only affected by true supply and demand.

A deflationary money will also require it to be made more divisible from time. Technically there are 2.1 quadrillion units of bitcoin, with 100M satoshi per bitcoin. It's somewhat likely that within the next 20 years, 1 satoshi will have the buying power of $0.01 at current value. If bitcoin continues to have minor deflation from lost UTXO's, at some point we will be forced to split further subdivide bitcoin, into 1B satoshi per bitcoin. That's 21 quadrillion units, a 10x increase of supply. Though as all holders of bitcoin will also have their holdings increased by 10x, the ratio of each holder to the total monetary network remains unchanged. That ratio is very important, as it determines just how much energy a holder of the money has a claim on compared to all the energy stored by the entire monetary network.

[Upgrade00]

An inflationary money increases demand pressure without true demand being affected. Price signals are thus manipulated and are impure. A deflationary money decreases demand pressure without true demand being affected. Price signals are still manipulated and are impure. A fixed money doesn't affect demand pressure at all. Price signals are pure and only affected by true supply and demand.

Demand in these cases is not constant but changing in response to users sentiment about the asset class.
Bitcoin is widely considered to be deflationary with a steady increase in purchasing power of holders, but this is not entirely down to a finite supply and as such cannot be termed as a manipulation.
What constitutes a "fixed money" in your context?

... Though as all holders of bitcoin will also have their holdings increased by 10x, the ratio of each holder to the total monetary network remains unchanged. That ratio is very important, as it determines just how much energy a holder of the money has a claim on compared to all the energy stored by the entire monetary network.

Multiple times in the past decade, holders have had the value of their holdings increased by 10× without the ratio of their holdings in relation to the total supply changing, why do you consider the ratio important and what are the downsides?

[seekererebus]

An inflationary money increases demand pressure without true demand being affected. Price signals are thus manipulated and are impure. A deflationary money decreases demand pressure without true demand being affected. Price signals are still manipulated and are impure. A fixed money doesn't affect demand pressure at all. Price signals are pure and only affected by true supply and demand.

Demand in these cases is not constant but changing in response to users sentiment about the asset class.
Bitcoin is widely considered to be deflationary with a steady increase in purchasing power of holders, but this is not entirely down to a finite supply and as such cannot be termed as a manipulation.
What constitutes a "fixed money" in your context?

... Though as all holders of bitcoin will also have their holdings increased by 10x, the ratio of each holder to the total monetary network remains unchanged. That ratio is very important, as it determines just how much energy a holder of the money has a claim on compared to all the energy stored by the entire monetary network.

Multiple times in the past decade, holders have had the value of their holdings increased by 10× without the ratio of their holdings in relation to the total supply changing, why do you consider the ratio important and what are the downsides?

A "fixed" money would be a money with a precisely fixed supply. No inflation, no deflation, ever.

My thinking on this is heavily influenced by the philisophical work Robert Breedlove is engaged in. A monetary network is a kind of social energy storage. Money is a claim on energy. The total energy in a monetary network is thus the overall energy capacity, and the ratio of one's holdings is how much of that energy one owns. When people trade dollars for bitcoin, they move their energy from the dollar network to the bitcoin network. Dollars and fiat generally is a scam because the ratio of all holders is decreased as the currency is created. Energy is taken from existing money and moved into the new money, thus energy is stolen. There are thus two factors that determine how much value your money has: the ratio of your money to the total money supply, and how much energy is claimed by that monetary network. Unfairly change the ratio, and supply demand mechanics must be affected. Moving energy between networks is much more complicated, but hyperbitcoinization promises to leave us with only 1 network.

[Upgrade00]

A "fixed" money would be a money with a precisely fixed supply. No inflation, no deflation, ever.

Bitcoin qualifies as a fixed money in that context. The supply is fixed at a certain amount and is introduced at a constant rate over a pre determined period of time.
The only slight difference is that it's not introduced all at the same time, but rather gradually eased into circulation at a controlled rate.

[serjent05]

I've often heard it referenced how BTC has a fixed supply, and that being one of it's greatest advantages. It doesn't however, what it actually has is a decreasing supply with an unverifiable loss rate.

BTC has a fixed amount of total coins, and though the block reward is deflationary, we cannot deny the fact that there will be only 21m Bitcoin created.  I agree that Bitcoin instead of saying has fixed supply of coins should be considered a deflationary currency because of the fact given by @OP.  

Until there's a need, I can't see anything being implemented about restoring lost coins back into the network after a certain amount of time has passed. Besides, 21 million isn't the important figure, the important part of it is it can never increase. It doesn't matter if we only have 18, 17 or 14 million in circulation. The important part is that number never increases beyond 21 million.

True but doesn't it make Bitcoin more valuable if we tend to realize that Bitcoin supply is deflationary?  It is actually supported that the total coins in circulation deflate every time one holder forgot his private key and is unable to access his Bitcoin.  Isn't this Bitcoin is lost making the remaining Bitcoin in circulation less than what it suppose to be?

[Welsh]

A "fixed" money would be a money with a precisely fixed supply. No inflation, no deflation, ever.

I think you might be thinking fixed as a broader term, but by fixed what a lot of users mean is that the amount of Bitcoin will never increase or decrease, at least via code. Of course, Bitcoin will be lost, and therefore the amount in circulation will be lost, but technically there's still 21 million Bitcoin in existence. I think that's the main difference of opinion here, as for fixing inflation, and deflation I think that's another discussion entirely.

True but doesn't it make Bitcoin more valuable if we tend to realize that Bitcoin supply is deflationary?  It is actually supported that the total coins in circulation deflate every time one holder forgot his private key and is unable to access his Bitcoin.  Isn't this Bitcoin is lost making the remaining Bitcoin in circulation less than what it suppose to be?

Technically, for many users this would make it more valuable. However, I don't think that's the problem at all. I think it's the moral implications of altering something that has been deemed to not change, and the fact that no timescale will be able to determine whether or not someone should lose their coins. Since, that private key or seed may be passed down generations in theory.

[pooya87]

In order to avoid confusion the best term to use is that bitcoin has supply cap, as opposed to fiat that has none! Meaning they can print as much fiat they want while you can't print more bitcoin than 21 million cap.
Whatever happens in the middle (production rate, loss rate, etc.) are variables.

We could, if we really wanted to, implement an expiration timer on UTXO's of 21 Million blocks (~400 years, block count chosen for symbolic value).

We should only make decisions that concern us not 4 generations ahead of us

[franky1]

In order to avoid confusion the best term to use is that bitcoin has supply cap, as opposed to fiat that has none! Meaning they can print as much fiat they want while you can't print more bitcoin than 21 million cap.
Whatever happens in the middle (production rate, loss rate, etc.) are variables.

We could, if we really wanted to, implement an expiration timer on UTXO's of 21 Million blocks (~400 years, block count chosen for symbolic value).

We should only make decisions that concern us not 4 generations ahead of us

or easier to term:
bitcoin supply CAP(ceiling)
fiat debt canopy(it grows)

bitcoin production rate (halves per 4 year until it cant half anymore, at which point is the supply CAP/ceiling)
fiat money(debt) print(continues and varies but never stops)