Is this idea to counter lost bitcoins possible?

[bitfreak!]

I know this topic has been discussed in depth before, but I can't be bothered rehashing any old thread or read through some long threads to find an answer to my question. First of all, many of you will think that lost bitcoins don't matter, and I partially agree; the divisibility of bitcoin makes it nearly impossible to have a number of bitcoins in circulation which is too small. But on the other hand it does pose a real threat to the stability and feasibility of bitcoin in the very long term. I mean we will eventually reach a point where there's just a ridiculously small amount of BTC in circulation and I think it's better to avoid that scenario if possible, if only to help alleviate the concerns of people who think this is a real problem.

The solution is quite simple, if an address holding bitcoins remains inactive for maybe 50 or 100 years or maybe even more, then the bitcoin becomes minable again, or some how put back into circulation. If you don't want to lose your bitcoins then simply transfer them before this time period elapses, that seems completely fair to me. I'm aware that this idea has been suggested before, but I never read enough to be sure about how feasible this idea is and what problems may prevent it from happening. I don't see why anyone would have a problem with this system. It will ensure that all bitcoins eventually go back into circulation and that seems very healthy to me, rather than a constant decline in the number of bitcoins.

Keep in mind I know very little about the complex inner workings and technical details of bitcoin so keep your answers relatively simple.

[xxjs]

It could be done, but there is no point to it. The principle of a fixed, decreasing inflation rate is paramount, from the viewpoint of existing coin holders.

[bitfreak!]

It could be done, but there is no point to it. The principle of a fixed, decreasing inflation rate is paramount, from the viewpoint of existing coin holders.

Inflation (of the money supply) will decrease regardless, because new coins are created slower and slower until we reach the limit of 21 million. However, even if this idea was implemented, deflation (in terms of change in BTC value) will still happen because presumably the economic activity within the bitcoin economy will increase on average. Even now the value of BTC is going up because the growth of the economy out paces the creation of new BTC. That is really why bitcoin is deflationary, not because the money supply will get smaller and smaller. Either way I see no reason why it is "paramount" for the money supply to continuously decrease, that is absurd reasoning which could only come from someone who has a lot of BTC stashed away.

edit: made a few specifications to above text to make it clearer.

[DeathAndTaxes]

yes your exact proposal has been proposed at least a couple hundred times in the past.  Nothing novel or unique about it.   It will never happen.  There is no reason for it to happen and plenty of reason for it to not happen.

I'm aware that this idea has been suggested before, but I never read enough to be sure about how feasible this idea is and what problems may prevent it from happening. I don't see why anyone would have a problem with this system.

There is nothing which would prevent this from occuring even tomorrow other than the need for near unanimous consensus among bitcoin users.  You won't get that, it will never happen.  Most users don't support it but even if they did to avoid a hard fork and split of the block chain you would need nearly unamanimous support (i.e. 90%+ of users, 90%+ of merchants, 90% of service/wallet providers, 90%+ of developers, and 90%+ of miners).  The risk of a hard fork and split of the network at this point doesn't warrant any dubious gain from having coins put back into circulation.

TL/DR you are trying to solve a problem which doesn't exist and thus it will never be "solved".

[bitfreak!]

There is no reason for it to happen and plenty of reason for it to not happen.

There is clearly a reason for it to happen. It seems obvious to me that it's much healthier to have a money supply which remains stable, then to have a money supply which decreases into infinity; that's like the opposite absurdity of infinite inflation, both extremes are moronic in the long term. From a mathematical, practical, and economic perspective, the most desirable option is obvious. Now please enlighten me with the reasons why it shouldn't happen, other than you want your BTC to go up in value due to a decrease in the money supply.

[xxjs]

This has already been discussed all the way. Basically, I would go for it if the lost coins were allocated proportionally to existing coin holders. But wait! - that is exactly the same as not doing it, existing coins would gail in exchange value. Maybe you would give the lost coins to somebody else, the world government? You could as well rename Bitcoin to Bernank.

[bitfreak!]

This has already been discussed all the way. Basically, I would go for it if the lost coins were allocated proportionally to existing coin holders. But wait! - that is exactly the same as not doing it, existing coins would gail in exchange value.

If the coins were re-mined they would be going to people who use their computing power to mine the coins, exactly as they are distributed right now. And it is very different from not re-mining them, for the reasons I have stated thus far. I fail to see any reason why some of you would be so strongly against this idea, and your lowbrow sensationalist remark concerning Bernanke only indicates to me that you have a hidden agenda here.

Infinite deflation of the money supply is no more or less insane than infinite inflation of the money supply. If bitcoins answer and opposal to the current inflationary money system is to take on a form exactly the opposite to that system, instead of reaching a nice balance between the two extremes, then you guys clearly do not want to create a long term fair and stable currency to oppose the mainstream, but something which benefits a few of you in the opposite way that our current money system benefits a few.

[DeathAndTaxes]

There is no reason for it to happen and plenty of reason for it to not happen.

There is clearly a reason for it to happen. It seems obvious to me that it's much healthier to have a money supply which remains stable, then to have a money supply which decreases into infinity; that's like the opposite absurdity of infinite inflation, both extremes are moronic in the long term. From a mathematical, practical, and economic perspective, the most desirable option is obvious. Now please enlighten me with the reasons why it shouldn't happen, other than you want your BTC to go up in value due to a decrease in the money supply.

One I don't accept that the money supply needs to be fixed exactly.  The loss rate over time will slow to a trickle and the money supply decreasing by 0.5% a year or (or even increasing by 0.5% a year) isn't a material problem.   For all practical purposes the supply is fixed and price is based on demand (which over the next 100 years or so changes in demand will dwarf any valuation changes due to slowly shrinking supply).  However lets pretend the money supply "should be fixed".


There are three major issues.

First.  Your solution won't make the money supply stable.  Valuation is based on the available money supply.  If anything your proposal will cause rapid fluctations in supply, and mining rates (to potentially cash in on a once in a lifetime windfall in mining profits).   Lets (pulling numbers out of my ass) that in the first 4 years roughly 2 million BTC were lost.   We would expect over time this rate of loss to decline as the obvious value of BTC becomes apparent as well as improved technology and best practices (deterministic wallets, auto backups, hardware wallets, etc).  So the "lost coins" are to a certain extent already price in.     Currently the price is based on effective supply  Nobody needs to know how many are lost the market effect of supply and demand will set the price based on effective money supply.   As the annualized loss rates continues to decline (in nominal BTC terms) the effect on overall valuations will be small.  The difference between an asset whose supply is truly and perfectly fixed vs shrinking at say 1% is immaterial.  The change is price will be driven mostly by changes in demand.   As an example say 10 years from now loss rate is <1% per year and demand is growing at 35% per year.   35% vs 36% does it really matter?  The price is for all intents and purposes determined by increased demand.


However whatever "expiration date" is used means that in roughly x years there will be a massive INCREASE in the effective supply when that window hits.   Say 2M coins were lost in the first 4 years.  That means in 50-54 years the coin supply (which has had slowing inflation for decades now) will supply see a minting explosion of 2M BTC (which could be worth billions of USD or more).   The fear and uncertainty on the true supply would have the exact opposite effect as intended.   As we approached the expiration date a lot of volatility would occur as people try to estimate how much new supply is going to crash into the market.   Another issue is the risk of overmining.   Mining will reach an equilibrium with transaction fees and in 50 or so years should be relatively stable however the expiration of these early 2M coins (which could be worth a small fortune) would cause a massive mining surge.  The network will built out massively as the ROI when including this once in a lifetime bonanza of confiscated coins would warrant expeditures than "normal" mining wouldn't.  What happens if the hashrate jumps 200x and then 99.5% shutoff after that free money window is gone?  Oops.  Network needs 400 weeks to reset and until that happens block times are ~1 day each?

Second.  It is unfair.  There is a social contract in Bitcoin.  Bitcoin is voluntary.  Nobody is forced to use it and the rules are well known.   However what you propose is an "Ex post facto" change.  It is material unfair.  People entered in Bitcoin because on certain proclaimed truths one of which being that (for better or worse) transactions are irreversible.  This is a form of reversibility.  You are proposing to change the social contract after the fact and that is ALWAYS unfair.  A system built from day one with this change (and other hard fork changes) would be different but to impose it upon people after the fact isn't just unfair it is immoral.

Third.  This is all academic.   It simply is NEVER EVER going to happen.  Just like the hundred or so last threads nothing is different.  Making this kind of hard fork is simply "defacto" no possible (although technically possible) due to the consensus rules built into Bitcoin.   Unless a change requiring a hard fork has overwhelming support the other "normal bitcoin" fork will live on.  So either one fork will die off or they would compete directly with each other causing chaos and uncertainty which devalues both forks.   "Can the real Bitcoin please stand up?"  

[mskwik]

I think another major issue people have with it is how do you know that the "lost" coins are really lost.  I could make a paper wallet and give it to my grandfather (who knows nothing about computer security) and if he locked the paper away and just sent coins to that address in 50 years they would still be there and usable with that paper.  If they had to be moved every couple years then I wouldn't be expecting any bitcoins in my inheritance.  Something like the physical Casascius coins have a similar use case, they could be traded around as much as you want and (assuming the tamperproofing holds up) 50 years later the coins backing them will still be there in the blockchain unmoved (despite however many hands the physical coin may have gone through).

[Seal]

I raised this suggestion a while back and still believe it to be a good idea however didn't get much consensus:

Interesting point, I see where you're coming from... that its a shame that so many bitcoins will never be realized as they're just forgotten about.

Just a thought... if an expiry system was introduced, similar to land leases, 99 years would be a good benchmark for expiry... and the miners could once again mine the coin. Not sure if this would be technically possible though.

That way the reintroduction will be very slow and not affect the economy a huge amount, however it would ensure 100% circulation of all 21million bitcoins each 100 years.

If in the future you have old coins that are > 90 years old, I'm sure it wouldn't be too much trouble to send them to a new address or re-create that new offline wallet.

Why not 80? 40?

If you could make changes like that it would break trust, can't so won't so we're fine.

Historically I'm using an age old concept as a suggestion of how the policy could work:

http://en.wikipedia.org/wiki/99-year_lease

Under the traditional American common law doctrine, the 99-year term was not literal, but merely an arbitrary time span beyond the life expectancy of any possible lessee or lessor.

Who's trust would it break given that the owner of the coins would have died meaning the coins will be out of circulation anyway?

Redistribution of expired coins can be done once again through mining such as how coins are currently distributed.

Whole thread here:

https://bitcointalk.org/index.php?topic=112525.msg1218706#msg1218706

[bitfreak!]

The loss rate over time will slow to a trickle and the money supply decreasing by 0.5% a year or (or even increasing by 0.5% a year) isn't a material problem.

I doubt it will slow to a trickle, it will slow at a rate proportional to the amount still left in circulation. People are always going to make mistakes no matter how many safe guards are put in place.

However whatever "expiration date" is used means that in roughly x years there will be a massive INCREASE in the effective supply.

The expiration time for any address would start when that address was last active, meaning the last time it had coins sent to or from it. Unless a huge amount of coins were lost at the exact same time (or rather, last active at the same time) I don't see how this concern is realistic.

You are proposing to change the social contract after the fact and that is ALWAYS unfair.

It would not be unfair if the majority of us agree to change the rules, that's what makes bitcoin so great, we can change the protocol according to consensus if we think it needs to be changed. And I think my arguments here are more than valid. The only people who will consider such a change "unfair" are people who are hoping to gain from it.

So either one fork will die off or they would compete directly with each other causing chaos and uncertainty which devalues both forks.   "Can the real Bitcoin please stand up?

This is the most realistic concern but I still hold that we need a stable money supply.

[bitfreak!]

I could make a paper wallet and give it to my grandfather (who knows nothing about computer security) and if he locked the paper away and just sent coins to that address in 50 years they would still be there and usable with that paper.

Then make the expiration time 100 years, or 150 years... what ever, as long as it can ensure a stable money supply. All one need do to re-validate the address and make it last another 100 or 150 years is send a small amount of coins to the address or remove some.

[bitfreak!]

However whatever "expiration date" is used means that in roughly x years there will be a massive INCREASE in the effective supply when that window hits.   Say 2M coins were lost in the first 4 years.  That means in 50-54 years the coin supply (which has had slowing inflation for decades now) will supply see a minting explosion of 2M BTC (which could be worth billions of USD or more).   The fear and uncertainty on the true supply would have the exact opposite effect as intended. As we approached the expiration date a lot of volatility would occur as people try to estimate how much new supply is going to crash into the market.   Another issue is the risk of overmining.   Mining will reach an equilibrium with transaction fees and in 50 or so years should be relatively stable however the expiration of these early 2M coins (which could be worth a small fortune) would cause a massive mining surge.  The network will built out massively as the ROI when including this once in a lifetime bonanza of confiscated coins would warrant expeditures than "normal" mining wouldn't.  What happens if the hashrate jumps 200x and then 99.5% shutoff after that free money window is gone?  Oops.  Network needs 400 weeks to reset and until that happens block times are ~1 day each?

As the first inactive bitcoins start to be re-mined it may cause a little bit of shock and uncertainty, however that will quickly settle down and the system will become accustom to handling the reintroduction of inactive coins and begin to flow nicely as a stable currency system with a stable money supply. There will be no point when the "free money window" suddenly disappears, because once it opens it will stay open... since people will continue to lose coins up until the window opens, and beyond. Once it is open it will stay open indefinitely and miners will always be able to re-mine coins lost 50 or 100 years ago, or with what ever time limit we go with.

[Rotsor]

What grudge do you hold against inactive coin-holders?? You can achieve the same simply by putting a lower bound on the block reward. No need to use silly methods. Still not going to happen any time soon.

[bitfreak!]

What grudge do you hold against inactive coin-holders??

I could ask you what grudge you have against a stable money supply. I have no problem with people who might like to leave coins inactive for long periods of time, my problem is lost coins which create infinite deflation. I strongly feel that infinite deflation could be just as harmful as infinite inflation over the long term. I mean honestly, what is the problem with having to send or remove some BTC to an address which has been sitting idle for 50 or 100 years? Is it really that hard to show the network that the address is still active? All your arguments are clearly based on the point that you want to profit from lost coins, and I find it sad that you are all placing profit over the long term health and stability of the network.

[bitfreak!]

To further illustrate the point I'm trying to make here, let me draw a comparison to gold, as people often like to compare bitcoin to gold in the way they function.

Gold can be lost, but not in the same way bitcoins are lost. When a person loses some gold, it doesn't just vanish into to thin air with virtually no chance of ever being recovered again, in the way bitcoins are lost. It will be some where waiting for someone to find it again at some point, even if they have to mine it out of the ground again. Would any of you really like a world where gold could vanish into thin air, and the supply of gold would slowly get smaller and smaller? Well of course some of you probably would judging by this thread so far.

But my point is that's not a healthy economic system, anymore than infinite inflation is a healthy economic system. Both are fundamentally flawed at a very basic level. Like gold, we need some system which will allow us to rediscover the lost bitcoins and bring them back into the system. The idea of a limited supply is great and fantastic, but it's not just a limited supply, it's an ever decreasing limited supply. It's easy to see that a stable money supply is by far the most superior currency structure, and bitcoin should be designed in the most superior way.

[bitfreak!]

Here I am responding to some comments made on the other proposal thread concerning lost coins:

• Extremely hard, as it requires consensus from almost everyone in the community, some of whom would consider this proposal as theft (I don't, but only if it had been a clear rule from the start).
• Very dangerous, as it risks a block chain split between old and new nodes.
• Hardly an improvement, as it doesn't change the distribution of coins, and the rate of lost coins is expected to drop as the economy around them grows.

1. It is not theft if after 100 years or more a person fails to indicate that an address is still active. If they want to secure their ownership of the coins then they can take action to prove the coins are not lost.
2. The risk is much worth the long term advantage in my opinion. I don't see why this should be such a controversial decision if we are aiming to make bitcoin better and fairer.
3. That depends on how you see the problem. From my perspective it's a huge improvement to secure the long term viability of bitcoin.

[Rotsor]

I could ask you what grudge you have against a stable money supply. ... my problem is ... infinite deflation.

There is none. I'm not against the slow inflow of coins through mining at all. Here, I've even told you how to achieve your goals without resorting to silly measures:

You can achieve the same simply by putting a lower bound on the block reward.

Yes, just make the minimum block reward equal to 1 BTC and the problem of infinite deflation is solved. My solution is superior to yours in these ways:
* It is much simpler, which means: easier to explain, easier to implement.
* It doesn't require long-term storage to be "refreshed", meaning less things to worry about.
* It creates a stable supply instead of allowing random market shocks. Imagine a million coins lost in 2009 hitting some lucky miner in 2109.
* It can not be subverted by a bot making 1-satoshi payments to all the addresses that are about to expire.

[bitfreak!]

You can achieve the same simply by putting a lower bound on the block reward.

Wouldn't that just lead to infinite inflation in the case where we get less coins lost then coins created?

* It can not be subverted by a bot making 1-satoshi payments to all the addresses that are about to expire.

That's a good point. Then the system should be designed so that only withdrawals will prove an address is still active.

[Rotsor]

You can achieve the same simply by putting a lower bound on the block reward.

Wouldn't that just lead to infinite inflation in the case where we get less coins lost then coins created?

No. As the currency supply increases, the effective inflation rate decreases (because it's equal to M/S where M is the constant minting rate and S is the growing currency supply). The deflation rate will likely stay constant because people having more are going to lose more. This way the equilibrium will be achieved, when the deflation rate equals the inflation rate and they cancel each other out.

* It can not be subverted by a bot making 1-satoshi payments to all the addresses that are about to expire.

That's a good point. Then the system should be designed so that only withdrawals will prove an address is still active.

... which puts a huge burden onto the offline, or worse, m-of-n wallet keepers.