A Better Coin

[JohnDoe]

To the extent that computing improvements are predictable, they are not a problem. In my formulation, the system generates more coins than it destroys (via txn fees) if difficulty grows more rapidly than 50% per year, and destroys more coins than it generates if difficulty grows less rapidly than 50% per year.

Well computing improvements are unpredictable, so it is a problem. There would be too much uncertainty with your method so nobody would adopt it. Also you can't destroy transaction fees, they have to go to miners or else there's no incentive for them to include those transactions in their blocks ahead of non-paying ones.

What you suggest is a mild exponential inflation. Besides the "incentive" issue (which I think we are disregarding for the sake of argument), it still doesn't prevent early adopter, ahem, "problem". Early miners would still have lots and lots of coins, because, if you consider the evolution of Bitcoin, the difficulty could (and would) increase 1000-fold after it goes mainstream. Assuming difficulty stays parallel with technological advancement after that, with %5 annual increase in inflation (105% inflation total), it would take 141 years for it to match early adopters' advantage.

I'm not sure you get the early adopter problem. In Bitcoin if you hold 210,000 coins then you own 1% of the whole economy, forever. If you held the same amount in this new chain your power becomes ever smaller. Sure at first it might be 50% of the economy but eventually it will become 0.000001%. The effect is compounded in that early adopters will be mining say 1000 coins per block while late adopters will be mining 1 million/block.

[1713507585]

1713507585

[1713507585]

1713507585

[1713507585]

1713507585

[1713507585]

1713507585

[cunicula]

Well computing improvements are unpredictable, so it is a problem. There would be too much uncertainty with your method so nobody would adopt it. Also you can't destroy transaction fees, they have to go to miners or else there's no incentive for them to include those transactions in their blocks ahead of non-paying ones.

Computing improvements have been pretty predictable for the last 20 years.
http://en.wikipedia.org/wiki/Moore's_law

Moore's law could change (which would change the inflation/deflation rate), but price changes would remain predictable once people became aware of the new 'law'.

You can destory txn fees of course. The miners don't need to get all of them. 

[JohnDoe]

Computing improvements have been pretty predictable for the last 20 years.
http://en.wikipedia.org/wiki/Moore's_law

Moore's law could change (which would change the inflation/deflation rate), but price changes would remain predictable once people became aware of the new 'law'.

Even if you could predict the strength of the network in any point in the future it's still not good enough. What if quantum computers make the difficulty suddenly skyrocket and create hyperinflation for decades to come? Everyone would dump the currency. You need a money supply growth rate that is always constant and close to the average GDP growth rate to try to maintain prices as stable as possible so that we can have a healthy economy.

You can destory txn fees of course. The miners don't need to get all of them. 

Why would anyone pay transaction fees if miners don't care which transactions are included in their blocks?

[memvola]

I'm not sure you get the early adopter problem. In Bitcoin if you hold 210,000 coins then you own 1% of the whole economy, forever. If you held the same amount in this new chain your power becomes ever smaller. Sure at first it might be 50% of the economy but eventually it will become 0.000001%. The effect is compounded in that early adopters will be mining say 1000 coins per block while late adopters will be mining 1 million/block.

Sure. What I was saying is that "eventually" might take a long time depending on the amount of inflation you introduce. And "forever" might not be so long in practical human scales. Bitcoin is somewhat mimicking gold production, which seems to have worked to some degree for human society.

Let's say that there are two different things some people take issue with in Bitcoin's coin generation scheme. First is, inflation decreases with time, which puts people who mine the first 4 years to an advantage. Second is, difficulty was very low before lay people found out about it and raised its value, and a handful of early adopters had, hypothetically, hoarded huge amounts of coins before that.

If you only wanted to solve the first problem, you could use a linear inflation scheme by changing the inflation multiplier from 0.5 to 1. God forbid, you can even fork the chain at the moment of the switch next year and people who prefer linear inflation can carry their bitcoins to the new branch. This would decrease the basic adoption incentive, but on the other hand, if it is perceived as better grounds for a currency, it would create a different incentive.

As you propose a geometric progression, looks like you also want to solve the second problem. As you said, difficulty is not predictable, but that is a double edged sword. By guessing a safe multiplier, you are still trying to predict the future, including the mining difficulty. If the ratio is too high, this could still result in hyperinflation. If it is too low, second problem will still be there for a meaningful interval of time. I'm not saying that ratios lower or equal to 1 are better predictions, they are just better understood and easy to work with.

[matsh]

Psychological barrier to entry is simply this: Few people are going to be willing to buy a single bitcoin for $1,000.  It's a psychological thing.  Having $1,000 in your bank account just "feels" a lot better than having 1 bitcoin, even if they're both worth the same amount and both have the same purchasing power.  To a lot of people, buying a handful of bitcoins for a lot more than a handful of dollars feels like they are getting ripped off.  Unless the virtual currency can have rough parity with the dollar, it will not ever be fully accepted by the general public.

One Google share is currently at 557 dollars. That high price haven't stopped people from buying them.

[SgtSpike]

Psychological barrier to entry is simply this: Few people are going to be willing to buy a single bitcoin for $1,000.  It's a psychological thing.  Having $1,000 in your bank account just "feels" a lot better than having 1 bitcoin, even if they're both worth the same amount and both have the same purchasing power.  To a lot of people, buying a handful of bitcoins for a lot more than a handful of dollars feels like they are getting ripped off.  Unless the virtual currency can have rough parity with the dollar, it will not ever be fully accepted by the general public.

One Google share is currently at 557 dollars. That high price haven't stopped people from buying them.

You don't have a clue how many people haven't bought Google stocks because of price.  And really, without some expensive polling, it would be impossible for you to know anyway.

[FlipPro]

I support this idea.
 
Bitcoin is an excellent commodity, but we can't run away from the fact that it stinks as money(because of it's instability).

I had an idea called Infinity Coin/Credits/Dollars/Digits where the coins print forever, and the distribution rate is solely regulated by the difficulty of the network. This makes it so that the more demand there is the less of the money is available and vice/versa giving miners/traders/business owners all reasonable incentives to adopt the currency. The currency would be more evenly distributed in the beginning, and will be inflating forever (at a very slow and predictable rate). It would be like creating a self regulated global federal reserve that's fair and transparent. This blockchain could directly compete with the deflating chain (Bitcoin) and inspire real cash adoption.

It would be nice to see someone take this Idea and I would surely participate.

EDIT:This would be the ideal system for most Mainstream economist. It would be really good if you got some of them on board with a decentralized/inflationary self-regulated system that operates off pure supply/demand, and projects a stable long-term financial reality for all businesses. SGTspike I would highly recommend you to talk to people who are professionals in this field and you will see that their opinions aren't to far from ours, try to get them involved. The problem isn't longterm inflation, the problem is greedy politicians manipulating the longterm odds to benefit their short term gains...

[SgtSpike]

Thank you for the support Matthew!  I don't expect everyone to agree with my conclusions regarding bitcoin and the issues it has, but that is ok.  They'll have a chance to attempt to prove themselves right soon enough.

I suppose the only thing stopping me from actually going through with it is not having the knowledge to change and compile the C code myself.  If I can find a C programmer who believes in the idea in the same way that I do, then we could start it up.

The specific variables would need to be nailed down as well.  More research needs to be done on what the actual loss rate would be, to determine the appropriate block reward to reach eventual parity with the dollar.  My estimates are based solely on lost coins from bitcoin, but it is difficult to say how well that loss rate could be extrapolated across the general population.  On the one hand, many of those lost coins were lost when people were more careless because the amounts that they had were not worth much.  On the other hand, the people here could be considered above-average with regards to computer knowledge and use, so would be expected to lose fewer coins than the general population.

FlipPro, could you explain the relationship between difficulty and distribution that you are talking about with your proposal? 

[JohnDoe]

After much soul searching I'm now utterly convinced that investment and deflation will always be incompatible so I'm fully on board with this project.

As you propose a geometric progression, looks like you also want to solve the second problem. As you said, difficulty is not predictable, but that is a double edged sword. By guessing a safe multiplier, you are still trying to predict the future, including the mining difficulty. If the ratio is too high, this could still result in hyperinflation. If it is too low, second problem will still be there for a meaningful interval of time. I'm not saying that ratios lower or equal to 1 are better predictions, they are just better understood and easy to work with.

The scheme I propose doesn't care about mining difficulty. It does try to predict the future but only that block generation will stay roughly at the intended interval (6 per hour, around 52560 blocks per year) so that the money supply growth rate stays on target. There will certainly be some deviation but hyperinflation would be next to impossible with this scheme because of difficulty adjustment. Some of the deviation could be mitigated by making retargets closer together, like every 1008 blocks instead of 2016.

[JohnDoe]

I suppose the only thing stopping me from actually going through with it is not having the knowledge to change and compile the C code myself.  If I can find a C programmer who believes in the idea in the same way that I do, then we could start it up.

The specific variables would need to be nailed down as well.  More research needs to be done on what the actual loss rate would be, to determine the appropriate block reward to reach eventual parity with the dollar.  My estimates are based solely on lost coins from bitcoin, but it is difficult to say how well that loss rate could be extrapolated across the general population.  On the one hand, many of those lost coins were lost when people were more careless because the amounts that they had were not worth much.  On the other hand, the people here could be considered above-average with regards to computer knowledge and use, so would be expected to lose fewer coins than the general population.

I believe your proposal of capping the reward at 500k coins will eventually produce deflation. If you try to match generation rate to loss rate then the money supply becomes finite just like Bitcoin.

and thus anything that doesn't sound libertarian enough will set off their alarm bells.

This chain wouldn't violate anything about Austrian school theory though.

[wareen]

While I agree that Bitcoin is not perfect and could certainly be improved, I really don't see how a new digital currency which has built in inflation could possibly compete against Bitcoin. You would effectively remove its usefulness as a store of value, no?

If I invest money, it is either because I need/want something directly or because I think the investment will gain in value thus enabling me to buy more of the things I need/want in the future. If I invest in BetterCoin (directly or via the electricity for mining) I neither get something I immediately need/want (no intrinsic value and nothing that I can't buy otherwise) nor does it help me to buy more of the things I need/want in the future because it will lose value over time.
Counting on other people adopting it on a large scale in the future (making it gain in value) seems like a very risky bet to me (due to the same argument) - certainly one not many people would be willing to take.

I honestly don't see a way around this - if you start two blockchains, ceteris paribus, the less-inflationary will attract more adopters. Also, I don't think it is a very reasonable assumption that Bitcoin will be the only accepted currency at some point in the future - therefore I also don't think that the deflationary nature of Bitcoin will ever really hinder innovation or growth of the economy.

I'm now utterly convinced that investment and deflation will always be incompatible

I'm no economist so forgive my ignorance but why do people keep buying electronic gadgets although they get cheaper and more powerful all the time? Deflation doesn't seem to be incompatible with investments there

[SgtSpike]

I suppose the only thing stopping me from actually going through with it is not having the knowledge to change and compile the C code myself.  If I can find a C programmer who believes in the idea in the same way that I do, then we could start it up.

The specific variables would need to be nailed down as well.  More research needs to be done on what the actual loss rate would be, to determine the appropriate block reward to reach eventual parity with the dollar.  My estimates are based solely on lost coins from bitcoin, but it is difficult to say how well that loss rate could be extrapolated across the general population.  On the one hand, many of those lost coins were lost when people were more careless because the amounts that they had were not worth much.  On the other hand, the people here could be considered above-average with regards to computer knowledge and use, so would be expected to lose fewer coins than the general population.

I believe your proposal of capping the reward at 500k coins will eventually produce deflation. If you try to match generation rate to loss rate then the money supply becomes finite just like Bitcoin.

and thus anything that doesn't sound libertarian enough will set off their alarm bells.

This chain wouldn't violate anything about Austrian school theory though.

It would produce deflation if the number of coins lost happened to be greater than the number of coins created over a given period of time, but that should have matching variance on the inflationary side as well.  The goal is to have a currency with the same number of usable coins at any given time.  Bitcoin is deflationary because people lose coins, so the goal would be to circumvent it.  I think that constant inflation results in overinvestment, bad investments, and too much debt, so I don't believe inflation is a good route to go in general.

While I agree that Bitcoin is not perfect and could certainly be improved, I really don't see how a new digital currency which has built in inflation could possibly compete against Bitcoin. You would effectively remove its usefulness as a store of value, no?

If I invest money, it is either because I need/want something directly or because I think the investment will gain in value thus enabling me to buy more of the things I need/want in the future. If I invest in BetterCoin (directly or via the electricity for mining) I neither get something I immediately need/want (no intrinsic value and nothing that I can't buy otherwise) nor does it help me to buy more of the things I need/want in the future because it will lose value over time.
Counting on other people adopting it on a large scale in the future (making it gain in value) seems like a very risky bet to me (due to the same argument) - certainly one not many people would be willing to take.

I honestly don't see a way around this - if you start two blockchains, ceteris paribus, the less-inflationary will attract more adopters. Also, I don't think it is a very reasonable assumption that Bitcoin will be the only accepted currency at some point in the future - therefore I also don't think that the deflationary nature of Bitcoin will ever really hinder innovation or growth of the economy.

I'm now utterly convinced that investment and deflation will always be incompatible

I'm no economist so forgive my ignorance but why do people keep buying electronic gadgets although they get cheaper and more powerful all the time? Deflation doesn't seem to be incompatible with investments there

That's because people don't buy electronic gadgets to invest - they buy them to use.  It's like saying a banana is worth less when you eat it.

[wareen]

I'm now utterly convinced that investment and deflation will always be incompatible

I'm no economist so forgive my ignorance but why do people keep buying electronic gadgets although they get cheaper and more powerful all the time? Deflation doesn't seem to be incompatible with investments there

That's because people don't buy electronic gadgets to invest - they buy them to use.  It's like saying a banana is worth less when you eat it.

True, but the point is, that people are more than willing to spend their money on things even if they get cheaper over time measured in their usual currency (eg. USD). So why do you think would people start to hoard money big time in a deflationary currency? If they want something, they'll buy it - even if it would be a few percent cheaper in a year (and Bitcoin's deflation won't ever be more than a few percent per year).

[JohnDoe]

I'm no economist so forgive my ignorance but why do people keep buying electronic gadgets although they get cheaper and more powerful all the time? Deflation doesn't seem to be incompatible with investments there

Not sure how that relates with what I'm talking about. I'm not denying time preference theory. Note that spending is not the same as investing.

When the purchasing power of a currency increases investment is discouraged because the risk of not getting a return on your investment or outright default becomes higher. It is way better to just hoard your money. You can read the following threads for more detailed discussion on this subject:

https://bitcointalk.org/index.php?topic=8329.0
https://bitcointalk.org/index.php?topic=28276.0

[JohnDoe]

It would produce deflation if the number of coins lost happened to be greater than the number of coins created over a given period of time, but that should have matching variance on the inflationary side as well.  The goal is to have a currency with the same number of usable coins at any given time.  Bitcoin is deflationary because people lose coins, so the goal would be to circumvent it.

Oh so you were talking about deflation in the money supply sense rather than price deflation. If you try to maintain a constant money supply then the purchasing power of the currency will keep increasing and raising capital will become just as hard as it would be with Bitcoin.

I think that constant inflation results in overinvestment, bad investments, and too much debt, so I don't believe inflation is a good route to go in general.

I agree with the Austrian business cycle theory in that inflation in the real world produces malinvestments but that's because of some key elements that this proposed chain doesn't have. The first is that in any nation with a central bank all money is created as debt, first by the banks taking loans from the central bank and then consumers taking loans from those banks. In here new money is mined instead, and it can be hoarded, spent or invested instead of lent. The other difference is that central banks can conjure money out of thin air with no cost, so they are able to set interest rates artificially low which leads to easy credit and thus artificial booms and malinvestments. In this chain there is a real cost to creating money so interest rates will be at market value, there will never be easy credit.

[SgtSpike]

It would produce deflation if the number of coins lost happened to be greater than the number of coins created over a given period of time, but that should have matching variance on the inflationary side as well.  The goal is to have a currency with the same number of usable coins at any given time.  Bitcoin is deflationary because people lose coins, so the goal would be to circumvent it.

Oh so you were talking about deflation in the money supply sense rather than price deflation. If you try to maintain a constant money supply then the purchasing power of the currency will keep increasing and raising capital will become just as hard as it would be with Bitcoin.

I think that constant inflation results in overinvestment, bad investments, and too much debt, so I don't believe inflation is a good route to go in general.

I agree with the Austrian business cycle theory in that inflation in the real world produces malinvestments but that's because of some key elements that this proposed chain doesn't have. The first is that in any nation with a central bank all money is created as debt, first by the banks taking loans from the central bank and then consumers taking loans from those banks. In here new money is mined instead, and it can be hoarded, spent or invested instead of lent. The other difference is that central banks can conjure money out of thin air with no cost, so they are able to set interest rates artificially low which leads to easy credit and thus artificial booms and malinvestments. In this chain there is a real cost to creating money so interest rates will be at market value, there will never be easy credit.

Why would maintaining a constant money supply alter the purchasing power of the currency?  The goal of the currency would be to maintain purchasing power, hence the constant supply, so if that wouldn't be true, I am interested in hearing why...  The only thing I can think of is world population increase, but I can't think of how one would reliably tie a coin reward to world population count.

[d'aniel]

JohnDoe,

This is how I view the deflationary property of Bitcoin, and how it pertains to lending.  It's contrary to yours and many others' views of it, but it seems so clear to me, so I'm interested in hearing what you and others think about it here.

First of all, I'll consider the case if/when a steady user base has formed and the adoption rate has levelled off, as this is the only case where your perpetually fixed 10% annual bitcoin supply increase might make sense.  Otherwise, you're probably going to need human feedback in your system.

Because the bitcoin supply will at this time be roughly fixed/very slowly increasing, any deflation will be caused by overall growth in the bitcoin economy.  Thus, the deflation rate and overall growth rate are equal, modulo lost coins and short-term fluctuations in velocity.

So if the deflation rate is steady and predictable enough (true by the above assumption), then people will only borrow to invest if they are sure their investment can at least keep up with the overall growth rate of the bitcoin economy, plus the usual premium paid for risk/time value of money/lender profit.

And if the loan is thus not made, and the money hoarded instead, then the deferred consumption that the hoarded money represents results in concomitant price decreases in the bitcoin economy that would not have occurred if the loan had been made.  So the overall bitcoin economy is the beneficiary of the deferred consumption, rather than the borrower whose investment couldn't keep up with its growth rate.  And this should be the case, should it not?

When viewed this way, the deflation rate is seen as a valuable measure of something like the average ROI in the bitcoin economy, and loans that are made uneconomical due to it are revealed clearly by it to be just that.  And there are clearly still beneficiaries of deferred consumption in the event that bitcoins are hoarded instead of lent - they are instead the bitcoin users who would have otherwise been competing to buy the goods that the borrower would have bought.  They're Bastiat's "unseen", whereas the borrower would have been the "seen".

[cunicula]

D'aniel, Wow, have you taken a course on monetary economics? Your intuition is very good. I'm impressed.

Basic long-run results that you allude to:
a) the deflation rate cannot exceed the growth rate of the coin's economy  [otherwise the price is a bubble]
b) the deflation rate cannot exceed the risk-adjusted expected rate of return on other assets [otherwise people seeking a better return will either sell assets to buy coin or sell coin to buy assets]

Unfortunately, the short-run is always anyone's guess.

[d'aniel]

Early adopters.  Well, if Bitcoins were to replace all currencies worldwide, and only 21 million coins were ever produced, then Satoshi, with his supposed 1.5 million coins, would hold approximately 7% of the world's currency.  And you don't see that as a problem?  Even if we only consider US money supply, which is around $14T, he would be holding $1T.  I know I am not the only person who has a problem with that, and I am sure enough people would have a problem with it to prevent mass adoption of Bitcoins.  It's not that I would be upset that he had that much money, it's that I don't want to see anyone with that much power over the currency.

Just a couple of quick points on this: the daily Forex turnover is $4T, so $1T does not seem like such a big deal when viewed this way.  Many markets have been much more cornered than this in the past, and every single fiat currency in existence now are good examples.  Plus, unlike the central banks that control these, once Satoshi's money is spent, it's gone.  No more can be conjured up.  Plus, it's highly doubtful that all or most of it would be held on to during an extremely unlikely rise to $14T market cap, unless Satoshi has some serious Scrooge McDuck syndrome, or is dead.

Deflationary spiral was shown as a problem during the great depression, when the dollar was based on a gold standard.  While it may not have been deflationary, it was certainly less inflationary than the currency we have today.  The country entered into a depression, along with many other countries, and each country only recovered when it started printing more money.  Also, as you have pointed out, raising capital for new innovations or companies will be extremely difficult while using a deflationary currency because the opportunity costs are so high.

While there was a correlation between deflation and depression during the great depression, based on the broader data this is not generally true: http://www.google.ca/url?sa=t&source=web&cd=1&ved=0CBcQFjAA&url=http%3A%2F%2Fciteseerx.ist.psu.edu%2Fviewdoc%2Fdownload%3Fdoi%3D10.1.1.147.6290%26rep%3Drep1%26type%3Dpdf&ei=9-lJTq_JH4HeiALqqMiTBw&usg=AFQjCNHyeqoF52JQhW83o9O-FDXdSQHJxA

Furthermore, is it really clear that deflation was the cause, and not the result of the depression?  It makes more sense to me that you need a wave of bankruptcies first in order to destroy the debt based money in the fractional reserve system before you can get falling prices.

I'm no economist though, just pointing out what seems logical to me.

[d'aniel]

Lastly I want to add that if my above characterization of lending deflationary currencies is wrong, and it is a problem even once Bitcoin's user base stabilizes, then a fractionally bitcoin-backed Bitcoin-like currency could always be launched that has whatever desired inflation rates you like.  This can be done using a low trust "distributed central bank" (the trust can be distributed over multiple organizations in multiple jurisdiction using the CHECKMULTISIG OP code) that issues all new currency, and offers redemption back to bitcoins.

This also allows for human feedback, so it's not so limited in its design as Bitcoin had to be.

It solves the initial distribution/adoption problems.

It also avoids to some extent competing for Bitcoin's existing and potential user base - something that should be seen as very undesirable in these early stages - since this isn't currently necessary due to the fact that the user base hasn't even come close to stabilizing, and because some proportion of bitcoins must back the new currency.

Because the specie that backs the currency (bitcoins) is likely impossible to confiscate in this scenario, I doubt this added centralization would be a problem.

Also, demurrage is much easier to implement with Bitcoin-like currencies than with traditional ones, so it can be used instead of inflation, with the same result, in order to avoid any risk of a bank run on the "distributed central bank".