StableCoin

[Red]

What you write is very similar to what I wrote in the GEM thread.
It is too late at night for me to verify your math, but I think you are on the right track.

Your analysis seems pretty sound. Before I start agreeing with you, let me propose one change (at the risk of going out to la-la-land):

This might mitigate the effectively-zero subsidy issue.

The first section you wrote is very similar to what I wrote here:
https://bitcointalk.org/index.php?topic=47628.msg569053#msg569053

I used a constant tax as part of the ration you propose.
I rebated the tax when monetary policy should be neutral.
I destroyed the tax when monetary policy needs to reduce the availability of circulating coins.

There are some differences. I let anyone who thought they could profitably generate coins do so at anytime.
As such I didn't have a variable difficulty over time. I had a monotonically increasing difficulty.
I compared the number of coins generated to the tax to calculate which of the above monetary policies was appropriate.

Now, switching back to agreeing with you; here's another problem with the design: when a new mining technology appears that allows people to generate a lot more coins with a lot less electricity (e.g. ASICs), there will be a sudden major inflation. If the ratio of coins generated to electricity consumed increases by a factor of 100, I see no reason the money supply wouldn't also increase a hundredfold. At this point, the miners have most of the coins (they just generated 99% of all coins generated to date), and transaction fees are now 100 times higher. Whoops. We were s'posed to be getting a StableCoin, but instead we got a ZimbabweCoin.

This is very similar to what I wrote here:
https://bitcointalk.org/index.php?topic=47628.msg566657#msg566657

We are on pretty much the same page.

In another thread I also worked to try to minimize the amount of energy needed to keep the whole system in balance. I'd have to dig that up to remember what I did. It just seems silly to burn so much CPU time when it isn't actually necessary at all.

I wrote about the lack of necessity a couple of days ago. It took a lot of posts to explain it to people. Here are the best:
https://bitcointalk.org/index.php?topic=175676.msg1832843#msg1832843
https://bitcointalk.org/index.php?topic=175676.msg1833134#msg1833134
https://bitcointalk.org/index.php?topic=175676.msg1833246#msg1833246
It kind of freaks people out when you tell them that mining is no longer necessary.

Your point about the correlation between mining difficulty and exchange price is an interesting one to think about again. I'll revisit your actual equations again tomorrow when I'm fresh.

[1715027023]

1715027023

[1715027023]

1715027023

[Red]

In addition to the ones already listed in the thread, I also proposed this:

New musings for a stable currency

That's a nice concise write up Etlase2. I like it. But when you said:

Flat tax: If over a significant number of blocks, the block creation rate averages over 20 minutes (vs. the 10 minute goal), a 1% flat tax will be applied to all existing coins and difficulty will be reduced by 25%.

Do you mean a demurrage style tax where you remove hoarded coins as well? Or is this just on the current transactions?

I'm pretty confident demurrage doesn't do anything to affect exchange rates. The clearest example of this I can give is to point out the case of coins in lost wallets. These coins theoretically exist but they never circulate. So when it comes to pricing lost coins can affect neither supply nor demand. Therefore any demurrage that destroys X% of lost coins does nothing to effect the coin supply EVER. Likewise, any demurrage of hoarded/not currently circulating coins does nothing to effect immediate supply.

[Impaler]

Red I've concluded that you need to have a market to predict inflation and deflation.  Obviously some external market is a single-point-of-failure and would not work, so markets need to be created INSIDE the block-chain itself, after all markets are just sets of user transactions handled by defined rules and we know how to do that in Bitcoin type networks.

The trick is that the market can only manipulate a few things, it can't trade coins vs commodities because it has no control over external commodities and no way to convey them or even compel people to deliver them.  Thus this market needs to utilize only things that are completely under the control of the block-chain rules, and that's not much.  The protocol can be control all balances, and it knows the passage of time (roughly), but that should be enough because all we care about is the value of coins today relative to the same coins later, we don't really care about it relative to other 'stuff' only too itself.

So the solution is to have this market make some kind of special 'instrument' like a Bond in a sense that creates or exposes this time-value change in value.  Then people put bets on each side putting their money on the line that their will be inflation or deflation.  The 'clearing price' tells us what is the consensus prediction and the instrument makes the correct prediction pay off.  Now we just adjust money supply counter to the prediction, if their inflation predicted monetary base contracts, if deflation is predicted it expands.

I just haven't figured out how to structure this instrument, It probably needs to involve waiting to receive coins at some later date or having some coins 'frozen' for some period of time.  Possibly the value being bid with is time rather then coins, their are lots of possibilities but it needs to expose that time differential for both inflation and deflation.

[Red]

Red I've concluded that you need to have a market to predict inflation and deflation.

This is a really insightful way to phrase it. I was kicking around a similar idea that involved "gambling" some coin you have now against generating new coins.

I agree with you that measurable human behavior needs to be the sentinel to determine when new coins need to be generated. I don't think an algorithm using a derivative value (like block difficulty) will be reactive enough to avoid oscillations. And if a system can oscillate there will always be speculators to attenuate the oscillation. I think using speculators to dampen oscillation is a better use of the resource.

Then here you come and point out the obvious with your insights about time. It's a "futures market." We speculators who trade in bitcoin futures. Genius!

I just haven't figured out how to structure this instrument, It probably needs to involve waiting to receive coins at some later date or having some coins 'frozen' for some period of time.  Possibly the value being bid with is time rather then coins, their are lots of possibilities but it needs to expose that time differential for both inflation and deflation.

I was having the same structuring problem with my "gambling" idea. How do you gamble in coins and pay off in coins, when the variable you need to make the decision is external to the system (fiat/external value per coin).

I'm not sure I know how to do it with a futures market either, but your insight about the ability to gamble time as well gives us one more variable! Woot!

So how do we do it? :-)

[hobbes]

Instead of the number of transactions you could also use the transaction fees. They probably are somewhat bound to the value of the coin.

Inflationary coins will never have a chance against noninflating coins. Who in his right mind would put his money in there if he could simply put it into a more stable coin? Thus Bitcoin and the like always win.

[herzmeister]

Binding to some however defined external (and complex) property will increase complexity and induce/increase the possibility of failure regarded to design flaws.

What is "stable in value"? While btc crashed against the dollar, I saw no significant changes in the relation to ltc/ppc.

To get it "stable in value" against the dollar you need a market capitalization in dollar much bigger in relation to the availible speculation capital in dollar. And even then there could be hedging.

I don't mean stable against the dollar, I mean stable in itself, as if StableCoin was the national currency of StableCountry. This can be achieved if its money supply can be made to grow (or shrink) by its actual adoption and usage. It will be traded freely at exchanges for other currencies (or baskets of goods, BigMacs or whatever), but with the aim to be stable in exchange value (if the other currency is stable too). The problem of course is we have no economic metrics other than its blockchain (or whatever method is used as a distributed accounting database).

There's little connection between the transaction rate and the monetary velocity. I can move money between wallets or I can pay using a PayPal like service, thereby introducing false or hiding real transactions from the chain. This is not simply statistical noise because high worth speculators will abuse your mechanism on purpose to move the exchange rate. Nevermind that there's very little connection between the exchange rate and the monetary velocity.

To control the exchange rate you need to watch it directly, because it's influenced by external factors: fundamental supply and demand, economic cycles, trade deficits relative to the fiat world, speculators, etc.

The goal is to eliminate external dependencies and statistical noise and to come to faithful predictions about actual adoption and use by internal metrics alone, like amount and volume of transactions. Maybe more metrics can be found. Of course people move wallets around, but this should also even out in the long term. Transaction fees may be able to discourage all too much manipulation. They will have to be fixed to a reasonable level. The exchange rate is intended to be boring and not attract speculators.

Miners can control the money supply using a voting mechanism.
They simply write how much reward there should be for finding a block. Newcomers will most likely enter a high value, old miners most likely enter zero.
So the inflation automatically adjusts to the hashrate the newcomers are able to provide which should be correlated to economic growth,

This can be combined with a demurrage scheme where coins decay into nothing if the currency shrinks. (If the hashrate goes down for example)

Sounds interesting; I originally did not intend a voting mechanism, but if metrics cannot be discovered automatically, why not let money supply growth be voted upon somewhat democratically. I don't understand why old miners would vote for no new coins for themselves though.

i have a theory, and i know that im going to be attacked for this, that bitcoin is so volatile in part because its deflationary. Basically the mentality goes something like this. I should buy bitcoin because its deflationary so if it becomes widely adopted it should increase in value forever. Of course if this was the case everyone should just buy bitcoins and never work again. but then bitcoin wouldn't ever buy anything if that happened since nothing would be produced to be bought. So it both makes sense for the individual to buy bitcoins and horde them forever but it doesnt make sense for society as a group to do this. The net result of this is a series of speculative booms as everyone realizes they need to get on the train to infinity, but crashes when they becomes obviously over valued.

If this theory is accurate what i would do to address it is build in slow, steady, predictable inflation. This would have the added advantage of forcing inflationcoin users to use inflationcoin for what it is intended to be used for, trade, not a get rich quick scheme and not a savings account. After all we have gold coins that work wonderfully as savings vehicle but terribly as a media of exchange, there is no need to reinvent the wheel. Build a coin that compliments gold and doesn't attempt to replace it.

Yes, Bitcoin is an experiment, and the economics are part of that. I don't think deflation is bad per se, it doesn't create depressions. The problem is rather that it may continue to result in high volatility, even if Bitcoin has satisfied and occupied its maximum possible market share and adoption. Bubble and burst cycles if you will. Maybe people will get used to it and learn to make better predictions in the future, which will even things out. But we don't know. That's why it's an experiment.

StableCoin would be StableCoin and not "Inflatacoin" or "Deflatacoin" (Bitcoin), so it may gain acceptance with both merchants and buyers, and maybe also savers who just want to store the information of the value of their labor without speculation and risk.

[ElectricMucus]

My own bank, insurance company, to decide offers, even the electric companies are starting using computational systems to distribute energy.

I just google... cause i think most of the big companies uses it.

http://books.google.es/books?id=pxAKhMKQ_dAC&pg=PA185&lpg=PA185&dq=genetic+algorithm+decision+making+business&source=bl&ots=4yPcFYb6Lz&sig=oSBlw3Xgkbkfqg5KgW70dNOWs1k&hl=en&sa=X&ei=OthtUZjEH-Lb7Ab2kYAY&ved=0CDcQ6AEwAA

And data mining:

http://en.wikipedia.org/wiki/Decision_support_system

The former is an academic paper on the theory behind the issue, not a concrete application, the latter a simple tool to help with optimization tasks.

I get the burning suspicion that you watched the Zeitgeist films a little too often and not much else.
I think we better leave it like that. If you still think your "AI" governed cryptocurrency: Talk is cheap show me the code.

[ElectricMucus]

Miners can control the money supply using a voting mechanism.
They simply write how much reward there should be for finding a block. Newcomers will most likely enter a high value, old miners most likely enter zero.
So the inflation automatically adjusts to the hashrate the newcomers are able to provide which should be correlated to economic growth,

This can be combined with a demurrage scheme where coins decay into nothing if the currency shrinks. (If the hashrate goes down for example)

Sounds interesting; I originally did not intend a voting mechanism, but if metrics cannot be discovered automatically, why not let money supply growth be voted upon somewhat democratically. I don't understand why old miners would vote for no new coins for themselves though.

Automatic schemes might work too, the big question is though how much they will still work later on. The room for experimentation is huge.

The reason I think old miners will vote for less reward is that the voting is intended to be valid for everybody, in a similar way difficulty is calculated. The thought is that old miners will most likely already have a decent stash of coins and will more likely be interested in preserving or increasing the value of the stash than accumulating more.

This concept is part of a more elaborate system which should mimic the way physical gold works in an economy. I don't know if it can be simplified in such a way.
Previous concepts were that coin holders have the possibility to hire "guards" for their stash to keep it from depreciating, while on the other side if they don't their coins get redistributed to newcomers which are the "bandits", "stealing" the coins. The problem I have with that is that is is modelled too much after a mediaeval society and modern economics work different.

The most important part though is that the system provides an engine for new economic growth, in a way like Bitcoin already did with Silk road.
What I always was thinking about is 3D printing and other grassroots manufacturing technologies as a niche.

[Etlase2]

Do you mean a demurrage style tax where you remove hoarded coins as well? Or is this just on the current transactions?

I mean a haircut across all the network balances. Transactions are unaffected. Demurrage, at least to me, implies a predictable rate of loss over time. The flat tax described in the "new musings" proposal is about only doing it when mining is insufficient to replace destroyed tx fees. To avoid tx confirmations from taking too long, the haircut happens and the difficulty gets lowered.

[Red]

I mean a haircut across all the network balances. Transactions are unaffected. Demurrage, at least to me, implies a predictable rate of loss over time.

Ah! You are right. Better to separate the two concepts. I like the term haircut.

The flat tax described in the "new musings" proposal is about only doing it when mining is insufficient to replace destroyed tx fees. To avoid tx confirmations from taking too long, the haircut happens and the difficulty gets lowered.

I guess I should read more of what you've written lately. Are you still equating 1 coin = X joules?

[dacoinminster]

How can we have a decentralized currency with stable prices? This is what I call "The Trillion Dollar Question", and I have spent hundreds of hours thinking and writing about this topic. I wrote a paper about it, and I'll be on a panel at the bitcoin conference in San Jose in May to talk about it.

Here is what I currently tell people (copied from another thread):

You have just asked the TRILLION-dollar question. If someone does manage to create a distributed currency which also achieves good price stability, the biggest barrier to distributed currency adoption will have fallen, and the world will be changed forever.

I have thought about this topic endlessly, and I wrote a paper about it. I have some very important news for you, so please pay attention:

Bitcoin, or something like it, CAN provide price stability. There are three ways this could happen:

• Friction: This is the default answer for most bitcoin enthusiasts, and I believe I have heard Gavin himself espouse this view, although I can't find the quote. If enough merchants and businesses support bitcoin, it becomes harder and harder for the price to swing around wildly, because there are so many goods and services available for buying and selling denominated in bitcoins. Friction stability is a long LONG way in the future, if it ever happens at all.
• Colored coins: If I buy a warehouse full of gold, I could annoint a few Satoshis (the smallest unit of bitcoin) as being worth 1 ounce of gold each, and then sell them. People could then engage in trade, buying and selling things denominated in gold using these "colored bitcoins" I created. I would agree to purchase back these tokens on demand for the current price of gold. I could do the same thing with U.S. Dollars or any other currency or commodity. The disadvantage of course is that people would have to trust the issuer. Incidentally, this is almost exactly the model used by Ripple, but running on top of the bitcoin protocol. I could imagine this working rather well if a reasonably trusted entity such as MtGox started selling them. The development effort for colored coins can be found at Bitcoinx.org
• Self-Stabilizing Coins: Bitcoin lovers don't like centralization, and while colored coins may someday provide stable value and anonymous transactions, they also require centralized trust in a currency issuer who could betray that trust or be raided by a government entity. Consequently, there will eventually be new currency protocols built on top of bitcoin, much like how HTTP is built on top of TCP/IP. These currencies will provide the capability to create self-stabilizing tokens which leverage derivative markets to maintain a target value. There are probably dozens of different ways to go about doing this. I wrote a paper about one of them. However, nobody has ever actually PROVEN that this could work.

The potential huge market for stable coins built on top of bitcoin is why I own bitcoins and have no intention to sell them in the near future. Colored coins WILL happen. Self-stabilizing coins MIGHT happen. Either way, bitcoin price increases are just getting started.

[Etlase2]

I guess I should read more of what you've written lately. Are you still equating 1 coin = X joules?

I'm pretty sure I stopped doing that while you were still around, but maybe it has been that long.


dacoinminster: how have we not ever spoken? Or maybe I've forgotten. What have you written in regards to stable prices? Have you read the decrits and/or encoin proposals?

[Red]

Wow! I thought this thread would be a bust like all the others. But it seems to be collecting a bunch of people who've thought about this subject a lot! Last time I was here posting about it there were exactly TWO of us! (Red, Etlase2)

Last night a couple of people separately PM'ed me to discuss this out-of-band. (I promise I'll reply shortly!)
Now dacoinminster shows up as a new evangelist. Woot!

I wrote a paper about it, and I'll be on a panel at the bitcoin conference in San Jose in May to talk about it.

Here is what I currently tell people (copied from another thread):

Can you link your paper and the other thread you mentioned?

These currencies will provide the capability to create self-stabilizing tokens which leverage derivative markets to maintain a target value. There are probably dozens of different ways to go about doing this. I wrote a paper about one of them. However, nobody has ever actually PROVEN that this could work.

Can you link this paper as well? At least a couple of us want to kick this futures market idea around a bit.

The potential huge market for stable coins built on top of bitcoin is why I own bitcoins and have no intention to sell them in the near future. Colored coins WILL happen. Self-stabilizing coins MIGHT happen. Either way, bitcoin price increases are just getting started.

I think I can make a solid argument about why any reasonably popular self-stabilizing coin will quickly obviate bitcoin. (Shhhh! Don't mention that here!) Actually, I wrote briefly about it a couple of days ago. Anyone else agree?

[Red]

Right now there is no sub-forum for discussing possible stable coin ideas. I'm not even sure there is enough interest to justify asking for one. However, quite a few of us have already written intelligently about the topic in various other sub-forums of this site.

I propose we create a [StableCoin] directory thread in which we can all link our respective ideas and prior writings on the subject. That way we can all get a broad overview of how others are attempting to solve the problem.

Perhaps after that we can prefix new discussion threads with [StableCoin] or some other tag. If we accumulate enough of them we might even ask for our own sub-forum?

Either that or we could all go hide in IRC! Any thoughts?

[Edit]
Perhaps I spoke to soon! Other are asking about creating Alternate Coin Sub-Forums at the moment.
https://bitcointalk.org/index.php?topic=178849.msg1864083#msg1864083

[timhuge]

I'm also excited to find this thread!

I just registered yesterday to see if there might be interest in a stable coin proposal: https://bitcointalk.org/index.php?topic=178027.0

[herzmeister]

The reason I think old miners will vote for less reward is that the voting is intended to be valid for everybody, in a similar way difficulty is calculated. The thought is that old miners will most likely already have a decent stash of coins and will more likely be interested in preserving or increasing the value of the stash than accumulating more.

Hmm I don't think this will work. We'd have some kind of prisoners' dilemma situation between the old miners here. Old Miner Joe would be tempted to not forgo his reward "for the common good" when all other old miners are supposed to, just because it's written down on the Golden Plaque of Ye Olde Miners' Guild or something.

How can we have a decentralized currency with stable prices? This is what I call "The Trillion Dollar Question", and I have spent hundreds of hours thinking and writing about this topic. I wrote a paper about it, and I'll be on a panel at the bitcoin conference in San Jose in May to talk about it.

Thanks. Interesting, but for this thread I'm interested in a solution that is a true and native single crypto-currency. Maybe just for the fun of it.  

For the moment I can think of these types of transactions:

1) User buys StableCoins from exchange => money supply should grow.
2) User puts StableCoins into their savings account (paper wallet etc) => money supply should grow.
3) User spends StableCoins for goods and services, or gambles => money supply should remain constant.
4) User gets StableCoins from their savings account intending to spend => money supply should shrink.
5) User sells StableCoins on an exchange => money supply should shrink.

Perhaps, at least in the beginning, users can optionally specify with each transaction which of these types it is. The system would be self-learning. Sure, users can lie, but... this data would be compared against intrinsically determined metrics as well, contradictions would be found, the rate of manipulations would probably also remain constant over time and be filtered out as statistical noise as well.

[ElectricMucus]

The reason I think old miners will vote for less reward is that the voting is intended to be valid for everybody, in a similar way difficulty is calculated. The thought is that old miners will most likely already have a decent stash of coins and will more likely be interested in preserving or increasing the value of the stash than accumulating more.

Hmm I don't think this will work. We'd have some kind of prisoners' dilemma situation between the old miners here. Old Miner Joe would be tempted to not forgo his reward "for the common good" when all other old miners are supposed to, just because it's written down on the Golden Plaque of Ye Olde Miners' Guild or something.

I think you still misunderstand me. Miners won't be able to determine their own individual reward.
The way it is supposed to work is for miners to write the desired reward into the block and the actual reward would be the moving average of the last n blocks.
So nobody would forgo their own reward in a sense that the current vote will be valid for everybody who finds a block.

[bnogal]

The former is an academic paper on the theory behind the issue, not a concrete application, the latter a simple tool to help with optimization tasks.

I get the burning suspicion that you watched the Zeitgeist films a little too often and not much else.
I think we better leave it like that. If you still think your "AI" governed cryptocurrency: Talk is cheap show me the code.

What ever u wanna believe.

Actually i can say i am working 17 hours per day (weekends just 10) so sadly i dont have time to watch movies, i just know in what are working my colleagues. As soon as i have some spare time i will develop an example. I will tell on time and i would lunch as a test. If someone wanna help here i am. If it works then to be fair it will be relaunched as bytecoin.

I just know that using a fix algorithm u need feedback to keep it working "properly", it is working perfect for a fridge, to keep the temperature.... but when the system is too complicated a fix algorithm is not enough. It can react too strong and send the system to the shit...  As 10 ASIC mining bytecoin for some days, and swapping back to bitcoin.

And what ever u can guess today, it can be false in the future. What if in 20 years we develop a fusion power plant that is working properly and the price of the electricity goes to 0. What if a quantum ASIC is developed and is able of produce coins as Finland mosquitos.

[Red]

I'm also excited to find this thread!

I just registered yesterday to see if there might be interest in a stable coin proposal: https://bitcointalk.org/index.php?topic=178027.0

Hi timhuge,

Welcome to the dissuasion. I just read your other thread. It's pretty much the standard thread you get whenever you try to start this discussion. You say "what if" and everyone else says, "can't be done", "you're stupid", "bitcoin's model will make me rich!" etc.

But in reality, you are right. I can be done. It can't be done in a way that bitcoin fanboys will love. But it can be done in a way that the majority of people will love.

[Red]

I don't mean stable against the dollar, I mean stable in itself, as if StableCoin was the national currency of StableCountry. This can be achieved if its money supply can be made to grow (or shrink) by its actual adoption and usage. It will be traded freely at exchanges for other currencies (or baskets of goods, BigMacs or whatever), but with the aim to be stable in exchange value (if the other currency is stable too). The problem of course is we have no economic metrics other than its blockchain (or whatever method is used as a distributed accounting database).

That's the meaning of stable I generally mean as well. I normally phrase it as "if one coin buys a loaf of bread, it will always buy a loaf of bread". However talking about dynamically stabilizing supply vs demand does require something observable to stabilize against. It also requires some mechanism to do the observing.

Those things are worth discussing!