Bitcoin Forum

First, I don't think there is anything wrong with bitcoin. However, people have brought up the idea of "backing" bitcoin with gold or otherwise controlling deflation. I'm proposing two types of coins which could be created with a modified version of bitcoin (possibly multicoin: http://forum.bitcoin.org/index.php?topic=24209.0). One would be based on the price of gold, Goldcoin, and the other would be stable with little to no inflation or deflation, Stablecoin.

I would like to see them created. I do not want to see bitcoin changed in any way to support them. I think there are enough other bitcoin like currencies out there that the idea of yet another competing currency would not harm bitcoin in any way.

Both coin programs I'm proposing, would need to include the bitcoin like program, and an exchange. There is not currently an exchange attached to any version of bitcoin to my knowledge. I'm working on a distributed exchange outside bitcoin (https://github.com/macourtney/Dark-Exchange), and you could use a centralized exchange like MtGox or TradeHill, but that would defeat the decentralized aspect of a bitcoin like program. More work and though is needed in this area.

For now, I'm assuming there is both the bitcoin program and an exchange which can be queried for the current exchange rate.

Goldcoin

The purpose of Goldcoin is to create a bitcoin like coin which could be exchanged for a set amount of gold, let's say 1 Goldcoin equals 1 ounce of gold.

Besides a modified bitcoin program (I'm calling the Goldcoin program), we need an exchange. This exchange will only exchange Goldcoin with gold. The price is listed in ounces of gold, and the current price can be queried at any time.

You can mine Goldcoin, and there are transaction fees for sending Goldcoins. However, the transaction fee is not given to the miners directly. The transaction fee and mining award is adjusted to keep the price of Goldcoin as close to 1 Goldcoin equals 1 ounce of gold.

How does it work? The Goldcoin program watches the exchange for Goldcoin and does the following:

If the price of Goldcoin increases above 1 ounce of gold, the Goldcoin program increases the number of coins given as an award to the miners. Also, the transaction fee is lowered, possibly to 0. If 1 Goldcoin is trading for 1.1 ounces of gold, then the Goldcoin program can increase the mining award by 10% (maybe more) until the price of Goldcoin goes back to the target.

If the price of Goldcoin decreases below 1 ounce of gold, then the Goldcoin program decreases the number of coins given as an award to the miners. However, the award for mining can never go to 0 since that would cause all miners to quit. The transaction fees are raised. Any coins captured by transaction fees are not given to minors, instead they would be destroyed. With less Gold coins floating around, the price of Goldcoin should move back up to 1 ounce of gold.

Thus, without an authority (besides programming code), and without using actual gold as a backing, Goldcoin will tend toward 1 Goldcoin equals 1 ounce of gold.

Of course, it would not be perfectly stable with sudden interest causing the price to go up temporarily and loss of interest could bring it down. In either case, you can always assume if you wait long enough, the price will go back to the target price.

One last point. The exchange could use USD instead of gold, and since the price of gold in USD is well known, you could simply target the USD exchange to match the price of gold. You would then have exactly the same outcome as the gold exchange.

Using this method, you could create a coin targeted to almost anything. Which leads me to another coin:

Stablecoin

With Stablecoin, you have the same bitcoin like program and exchange as with Goldcoin with only a couple modifications.

The exchange is in USD. However, instead of targeting USD directly, we target a specific date. Let's say we use the date Stablecoin comes into existence. From then on, we keep track of how much USD has inflated/deflated using something like CPI or the Billion Price Index or even a combination of indices. If USD inflates say 1% we increase the target to $1.01 for one Stablecoin. On the other hand, if USD deflates by 1% we target $0.99 for one Stablecoin. Over time the increase and decreases would be added together to get some strange multiple as the target for Stablecoin.

The ultimate goal for Stablecoin would be to target a value not a price, and thus never inflate or deflate over time.

Stablecoin would be a great benefit to any merchant. They could simply price their goods and services in Stablecoin once, and never have to change the price again. They also know, no matter how long they hold Stablecoin, the value of Stablecoin would never increase or decrease significantly. If it does, they can wait until it makes it back to it's target.

Once Stablecoin is used to price goods and services online, you could then drop the use of CPI and Billion Price index and simply use a basket of goods priced in Stablecoin. If merchants start charging more or less on average for their goods, Stablecoin could adjust the number of coins accordingly.

Stablecoin could become the perfect currency for merchants. Which would make Stablecoin perfect for customers who want to buy from those merchants. Once enough merchants and customers adopt Stablecoin, everyone else would follow.

Making it happen

Unfortunately, I have my own bitcoin project to work on right now which is taking all of my time (did I mention Dark Exchange: https://github.com/macourtney/Dark-Exchange :)). However, I'm sure some goldbugs on this forum would be interested in the Goldcoin idea and take off with it (if you do, please make it open source).

I'm more interested in Stablecoin, and would be willing to put some time into building it. I just don't have the time right now.

Maybe one exchange could work for all of the targeted coin types. If the exchange could be built first (maybe Dark Exchange is that exchange), then I'll work on Stablecoin when I have more time.

Feedback

Please give me feedback on my two proposals. What works? What doesn't/can't work? What have I not thought through enough?

I'm not as interested in whether or not a gold based digital currency or a digital currency with no inflation or deflation is useful. By creating the currencies, they will be able to compete in the open market and the market will decide. I want to know if such coins could be created and how.

Any other feedback is welcome.

Your idea has merit but you could technically do the same thing with bitcoin. Find someone with ALOT of cash (market cap is about 100 million) and have them say "no matter how low the market says/goes I am willing to pay $x for a bitcoin". Your asking to peg bitcoin to something which I believe is the opposite of what bitcoin was made for.

It is not an attempt to peg bitcoin to anything.

It is entirely new coinbases.

However, if you are going to give out more or less coins depending on current prices you are probably chasing a red herring. The number of coins being created soon becomes quite small compared to the number of coins already existing.

So it seems likely that the number of already existing coins people decide to dump or hoard might have far more effect on prices than the number of coins being created by miners.

I do not really see all the alternate blockchain coinbases "competing with" bitcoin at all, I see them as "co-operating with" bitcoin.

It is painful trying to use fiat currencies, exchanges that exchange only between various blockchain-based currencies will I hope be much easier to deal with than trying to exchange with fiat currencies. So I'd like to see a whole bunch of different blockchain-based currencies one can trade between without needing to bother with fiat at all.

-MarkM-

just to clarify, you are not (and have no affiliation with) bitcoin morpheus, correct?

I am bitcoin morpheus.

I don't want to say peg Goldcoin or Stablecoin to a price, since there will always be some volatility in the exchange. More like target a price. It won't be possible to smooth out every sudden jump or drop in price.

However, I think you could increase the transaction cost enough or increase the mining award enough to keep the exchange rate relatively close to the target rate.

On top of that, short term volatility would be smoothed out by speculators who could easily see when the price is above the target or below the target and buy or sell accordingly making a small profit and stabilizing the currency.

Would that not work?

You could try some of your ideas simply by getting in as an early adopter on any new blockchain that is at low difficulty or low price, thus allowing you to becomea major holder of the things. Then in your capacity as a major holder you should be able to stabilise the price, yes?

I figured this is what would hopefully happen with bitcoin: major holders would hopefully obtain so much fiat by selling at peak prices that they could from then on use that fait to keep buying back bitcoin any time its price in terms of fiat starts to drop.

This assumes of course that major holders actually intend to stick around and support bitcoin rather than treat it as a ponzi scheme, cashing out to fiat and not using that fiat to uphold the value of bitcoins.

-MarkM-

This could not work if the other blockchain became popular. Eventually it would cost too much to keep the price up or I would run out of coins from the other blockchain trying to sell enough to keep the price down.

I don't think there is a way to target a price for coins from an alternative blockchain without issuing more coins when the price goes up, and destroying coins when the price goes down.

Oh well I wasn't worrying about the early adopters trying to keep the price *down*, I figured in most new blockchains keeping it *up* would be the challenge. If it goes *too far up*, you can simply start yet another blockchain, which is almost certain to start out nice and low in price. :)

Also, if there is any truth in ideas that alternate chains would be *competing* to lower each other's value instead of *co-operating* to make them all in general rise in value, then maybe starting a new chain might help bring down the value of your previous chain. If so you wouldn't have to mess with making more of the previous chain to lower its price, just start up a "competitor".

I still think these alternates should help bitcoin value not lower it though, at least as long as the people or groups behind them continue to treat original bitcoins as the grand-daddy of them all, the blue chip, the most attractive one to use as "reserves" for "backing" your alternates and so on.

-MarkM-

bitcoin needs to compete or fade away on its merits/flaws ... co-opting competing chains only devalues/dilutes its worth by making a mockery of the "only 21 million claim" ...

.... at least the deflation scaremongers can go fo now that the true nature of free market money is made obvious.

After sleeping on it, I've decided that even though I still like the idea of Stablecoin, it would be easier to use multicoin to create a coin like Stablecoin but manually peg the value on the market.

I could do this by issuing all of the coin to myself and only selling it for a set price. I could then keep the money in trust, and buy back the coin if it every drops below the target price. I would sell the coin at a slightly higher price than the target, and buy it back at a slightly lower rate to absorb some of the fluctuation in the currency, while still keeping it targeted at a certain value.

There are a couple problems with that type of coin. The one obvious one is everyone has to trust that I will buy and sell at certain prices to keep the value of the coin stable. This is why I would prefer a market solution where you would only have to trust the market. However, for a test of the concept, it may work until a market driven solution can be made.

The next problem is, if the coin became very popular, I could run out of it and not be able to hold the price down. That would be a good problem to have, but it would still break my original promise to keep the value stable.

Also, the reserve could be stolen and sold on the market driving the price down. I would take precautions to avoid this happening. It's still a risk and requires people to trust me.

I'm going to go with it anyways and see how the market reacts. I'm going to create a block chain called 1971coin which will be pegged to the value of a 1971 dollar. Why 1971 dollar? It's the year the US completely dropped off the gold standard. That seems like a good value to peg.

I'm going to issue as many 1971coins as I can to myself and start selling them 1% above the value of a 1971 dollar (approximately $5.57 current USD). I will hold any money I get in trust, and buy back the coins if they fall below 1% less than the value of a 1971 dollar. I will likely also sell them for bitcoin and other digital currencies.

I will come up with a standard way to calculate the price of a 1971 dollar in current USD. I will the calculation method and the current rate at any given time online. I'm thinking of using the CPI for now. I haven't found a good way to use the billion price index. They don't seem to post the index at any given time on their website.

Once I have the program working, I'll ad 1971coins to the list of currencies you can buy and sell on Dark Exchange.

In a very short time I've come to admire both your work and thought process. And I am trying to think through the essence of the problem you're dealing with, and I suppose that it's:

centralization V decentralization

or stating the same thing another way:

stability V accuracy

With centralization we get currency stability, which is vital for trade, but we lose the accuracy (truth) of the currency's true market worth. We might also say that we lose the integrity, honesty, of the market by exposing the currency to enormous risk vis-a-vis the fiat of centralization powers, people (or person).

With decentralization we get accuracy (honesty), which is vital for integrity. And integrity is the cornerstone of the currency's strength, its durability, its ability to fight off competitors, usurpers, exploiters. Whereas a centralized currency will always be on the knife's edge of destruction, from within or without, because of its targetable point of failure. However, the price for this durability is stability. We lose stability when valuation is turned over to the market.

From a psychological perspective, I suppose the problem could be characterized as one-man's-reality Vs. the reality-of-the-collective...and then all of the inherent pros and cons that flow out from each.

...all of this leads me back to thinking that your darknet exchange is so important.

What's the point in making an accounting ledger (read: bitcoin) to be more than just an accounting ledger?

Bitcoin is already backed by the computing power of good/friendly nodes. Let's just make sure that their number always exceeds the number of rouge nodes. This can be achieved simply by spending bitcoins, rather than hoarding them (as if it's a commodity).

I have an idea very similar to this. I have funneled money into it, and I have an investor who is helping. The point won't be to peg the currency, but to try something similar at a risk of loss to myself as a wager in the future of bitcoin.

This is correct. However...


I don't think this is a good comparison. Especially since I would be basically selling the coin on the open market to force the price to be in a certain range, then the price is still accurate.

I would say, it's more like:

stability V freedom

If I'm controlling the supply of 1971coin, then you have to trust me to only release the coin at a specific rate to keep the price from rising too much, and buy it back up to keep it from falling. I would be the Federal Reserve for the 1971coin with the only difference being that my goal is to remove all inflation and deflation while making my formula for doing this as clear and public as possible. But, you still lose a bit of control and I would have at least some control over you at all times.

In the Stablecoin version, you don't have to trust me, just the code which you can see and audit at any time. The formula for keeping the price stable is written, not in stone, but in code. You can accept that law or not, or even change it to your needs (though you would be forced to use a different block chain). In the end, I don't have any control over you at all, and you always have control over your own finances.


For this reason, my priority is the Dark Exchange. Even though I'm creating 1971coin, I don't want it to take much time from me bringing Dark Exchange to a point where it is useful and used.

I don't think I'm proposing making an accounting ledger any more than it is. I'm just suggesting a way to make the currency in that ledger have a stable value.

And, I'm not suggesting changing bitcoin, but creating a competing currency which would survive only if others found merit in it.

How about Namecoin?

While Namecoin is an interesting concept in the beginning stages of adoption, it doesn't do anything like what I'm suggesting.

However, like Namecoin, I'm suggesting creating a new block chain for these currencies.

I'm a big fan of the "stablecoin" idea, in fact I even used that name earlier in one of my own posts suggesting something similar: http://forum.bitcoin.org/index.php?topic=7945.msg119375#msg119375

I like the concept of pegging to the 1971 dollar value. A central server that took deposits and withdrawals in bitcoins and issued a stable currency for commerce usage would be one way to do it, but then you have all the issues with central control that bitcoins circumvent (one server for the government to shut down, for instance.

A well-used and highly liquid bitcoin futures market would also go a long way toward solving this problem for big players, but wouldn't help much for your average joe.

What would be REALLY cool would be if some sort of automatic behind-the-scenes "stabilization option" could be built right into the bitcoin client. Imagine a radio button selection where I could choose to store my value in "1971 USD", "bitcoins", or "HyperBitcoins". The 1971 folks would be paired with the HyperBitcoin folks using contracts in the block chain. The contract allows the 1971 user to spend their bitcoins as if they have a set value in 1971 USD, and the hyperbitcoin contract holder is obligated to add bitcoins to the transaction if bitcoins are worth less, or pockets the extra bitcoins if bitcoins are worth more.

Edit: I created a separate thread for the HyperBitcoin idea above: http://forum.bitcoin.org/index.php?topic=30741.0

Your hypercoin idea seems complicated and could easily be handled with bitcoin options. It seems almost like hypercoin is a bitcoin option of sorts

My original idea for Stablecoin was to create a new block chain for it, and have the program increase the mining reward when the price goes up, and increase the transaction cost (then destroy the coins) when the price goes down.

Since I don't have the C++ skills to create Stablecoin, I planned to use the Multicoin client to create my own currency, 1971coin, which I could then issue and buy back to keep stable at the 1971 dollar value. Unfortunately, I seem to be having trouble building Multicoin, and the creator of Multicoin doesn't seem interested in creating pre-built binaries for it.

I've also looked at using Open Transactions to issue my own currency, but that seems to be even more confusing than Multicoin (though they do seem to have pre-built binaries posted).

I still have a few things to try.

BTW, a 1971 dollar is worth about $5.50 now which tells you something about inflation recently. One bitcoin now, would be a bit more than two 1971coins at the current rate.

Lets say I mine 50 Goldcoins, who is going to provide the gold?

Nobody provides you the gold (directly anyways). Since the mining reward is variable based on the price of Goldcoin, then you would be able to sell the 50 Goldcoin on the open market for 50 ounces or more of gold (or equivalent).

You are right that the hyperbitcoins idea accomplishes something like options, but nobody understands options except the few people who trade them for a living. :-/

Grandma isn't going to buy put options, but she might click the button on the bitcoin client to store value in 1971coins.

For your goldcoin blockchain, I think it would be better to piggyback on existing bitcoin mining, and create Goldcoin/Antigoldcoin pairs as described here: http://forum.bitcoin.org/index.php?topic=31032.msg390372#msg390372

I would also like to suggest UnstableCoin where the value swings as wildly as possible. This could be accomplished in a few means e.g. by a very funky and rapidly changing way of a deciding difficulty and giving a random number of coins at the completion of each block.  

I have a very similar idea, but much prefer to peg generation and destruction rates to difficulty growth rather than market prices.

My issue with pegging generation/destruction to market prices is that it requires a third party to supply accurate price information to the system forever. Two issues with this:
a) this is very centralized; take out the third party and you take out the coin
b) the third party could benefit from supplying inaccurate information.

An ostensibly attractive option is to allow users to vote on prices, but this does create incentives to supply truthful information. A voting system can ensure that everyone supplies similar information, but it cannot ensure that this information is truthful. There is no apparent economic reason why people will form a consensus around the truth instead of forming a consensus around a lie.

Pegging generation and destruction to difficulty growth (for example to achieve 50% annual difficulty growth) does not require the supply of outside price information. The relevant information is in the blockchain already. Linking coin generation and destruction to difficulty would ensure that the coin price approximately tracks the electricity price (I am assuming Moore's law will continue to hold).

There are certain features of this idea which would make it very difficult for anyone to profitably manipulate difficulty, but I don't want to go into the details because most people have short attention spans.

If you release stablecoin it should be in EUR cause EUR is more stable, and most important, I work with EUR :p

Could you blend it with dark exchange?

I am just absurdly interested in these "pegged" coins of all type, as you can tell from my own threads. I was wondering how you would handle the case when the coins were too plentiful and their prices too low, so I'm glad you described your plan.

I am concerned that destroying coins in high transaction fees would lower interest in the coins, leading to lower prices for existing coins, which would force you to raise transaction fees, which would lower interest in the coins, leading to lower prices, . . . you get the picture. Basically, this could result in a divergence which never recovers.

Another concern I have is that you don't want to drive the price too forcefully, or you will limit the impact of goldcoin traders on real-life markets. As long as the prices converge over the long-term, short-term swings are not much concern, and they present arbitrage opportunities.

You might want to consider combining your feedback system with the one I came up with: charging variable transaction fees. If you have a distributed exchange, you can charge no transaction fee when trading at the external spot price for gold, and an increasing transaction fee the further you trade from that spot price. This continually nudges prices toward the external spot price, and helps you not rely so heavily on destroying coins.

If the price of the coins drops too low, then the mining reward will be dropped (not all the way to zero which would stop people from mining), and the transaction fee would be raised. The lower the price of the coins, the higher the transaction fee. All coins collected from transaction fees which are not given to miners are destroyed.


I don't think we can know exactly what will happen with this setup without the actual stable coins in existence. My guess is, the higher transaction fees would cause people to hoard the coins. This will then remove more coins from existence than the higher transaction fees, and thus cause the price to rise.

Anyone desperate enough to sell below the fair value, would also likely be desperate enough to pay the transaction fee. Hoarders would then have an opportunity to make money through arbitrage, and coins would be destroyed at the same time.

In other words, I think the threat of higher fees or higher mining rewards would keep the price fairly stable without the actual higher fees or higher mining rewards going into effect that often.


I agree. I would have to think carefully about how to set the fee structure. A little bit of price fluctuation is fine as long as it converges to a stable price over time. However, that does seem to go against the name "Stablecoin". Maybe, "Targetcoin" is a better name. :)


The problem with a distributed exchange is, there is no central authority to enforce trading fees. Otherwise, it would be a good idea.

I will be so, so happy when this hypothesis actually gets tested by someone :)


Dang. You are right. I don't know why I didn't see that - even if everyone somehow used the same rules on the distributed exchange, people would just trade outside of the distributed exchange.

I'm increasingly convinced that you have the right approach here. I'm just worried about the coin destruction driving down coin prices rather than driving them up. If you can handle the "doomsday scenario" where 90% of people holding these coins panic that the coins won't hold their value, I'll be totally convinced.

People don't behave rationally during a panic, and I don't think you can rule one out.

I'm having too much trouble with multicoin and will look at modifying bitcoin directly (creating a new block chain and all). I'm not a C++ maven, so don't expect anything soon.


I'm sure panics will happen, but people who behave rationally would earn a lot of money from those who panic. I think it will sort itself out without destroying the currency. However, we won't know until the coin is created.

I had a similar idea (https://bitcointalk.org/index.php?topic=31032.msg398699#msg398699) but using demurrage instead of destroying the currency through fees.
The problem is how you get from starting zero to the target value just by controlling the quantity of coins.

The pegging idea may be more viable. Assuming that voting to obtain the price index in terms of a bitcoin-like currency works, I have a completely decentralized solution for a distributed reserve.
I'm still not sure that the system would be resistant to the depreciation of the currency in reserve.

The idea is based on the decentralized exchange for coins (https://bitcointalk.org/index.php?topic=31643.0).
Anyone can "destroy" (put in the network reserve) middlecoins to issue stablecoins or destroy middlecoins to get back middlecoins from the reserve.
The number of stablecoins depends on the market, but the number of middlecoins should be stable to prevent its depreciation.
Stablecoins can also be traded with other users at the price they like, for middlecoins, for bitcoin and/or other currencies if the protocol makes miners know about other chains.
Middlecoins can also be traded for any of the supported currencies.

Maybe in this case, escrowcoin would be a better name than middlecoin. It can be in another chain, but it has to look inside whatever chain want to use it as reserve and make sure that no more escrowcoins are created in the other chain than were destroyed previously (to put them in the reserve).

morpheus,

While pondering this topic this morning I suddenly realized there is a fatal flaw in your plan as stated. I'm very sad to have to point this out, because I really want something like this to work because it is so beautifully simple.

Your plan to destroy coins will not work as stated. You are right that destroying coins will give people incentive to not transfer them between wallets, but they will simply sell whole wallets loaded with coins instead. Once some people start doing this, anybody transferring the normal way will be at a disadvantage, and everyone will have to start doing it, then *poof* you completely lose all control over reducing the coin supply.

Personally, I have to return to pondering the hideously complex ideas in my proposal for the second bitcoin whitepaper for now, but I'm hoping you have a good answer for this because I liked your idea better.

You could use demurrage instead of fees to destroy the currency. With demurrage the velocity of circulation would also be more constant and higher, thus the changes in the monetary base affecting more (http://en.wikipedia.org/wiki/Equation_of_exchange) to the value of the currency.
I still see the problem of how do you get to the target value in the first place. The system can't control the value of a currency only by controlling its monetary base, and you have limits on how fast you can destroy and create. At the beginning you need to increase its value and also create much more than it is destroyed. I guess the value targeting system should be "switched off" until the currency reaches certain value.
Although I'm very curious about the feasibility of a decentralized stable currency, I'm not sure it would be desirable "for the economy".
The hard technical problem of a decentralized price index is still there. It could be very useful to solve it even without the stable currency.

I've updated the list on my sign to include stablecoin and reservecoin (decentralized beertoken/1971coin backed by escrowcoins).

I don't have a great answer for this problem, but...

If someone wanted to sell 100 Stablecoin this way, they would have to create a new wallet and put 100 Stablecoin in it. If they did this while the price of Stablecoin is low, then they would have to pay the higher transaction fee. Thus solving the problem in this case. If they moved the coin while the price was high, they wouldn't have to pay the transaction fee, but they could just sell the coin for a profit instead. The only way this would work is if they bought the coin when the price is low, then moved the coin into separate wallets when the price is high, then sold the wallets when the price is low again. Not only is this complicated and risky, but the initial purchase would pay the high transaction fee.

An alternative to the above is if someone bought a bunch of Stablecoin at some point and doesn't care about selling exactly 100 coins. They could sell their entire wallet at once with some random number of coins inside. The problem with that is, the receiver couldn't pull the coins out of the wallet without paying the high transaction fees. The receiver would have an incentive to hold the wallet until the prices come back up. He could then sell the coins at a profit and may not have to worry about the transaction fee. Again this encourages hoarding which would drive the price of Stablecoin back up.

The big issue is an exchange like Mt Gox. If there were an exchange like Mt Gox which held all of its Stablecoin in a wallet and let people buy and sell on its open market, then the price of Stablecoin would be independent of transaction fees. Of course, anyone who bought a bunch of Stablecoin while the price is low would not be able to pull it out of the market without paying the transaction fee. This again would encourage hoarding which would drive price back up to the fair market value.

There may be issues in practice, and the transaction fee/miner reward may have to be adjusted at times. I still think it could work and I'm slowly working on an implementation.

If anyone has any info on creating your own bitcoin chain, let me know. I'm looking for a tutorial if one exists. Or maybe a tutorial could be added to the bitcoin wiki.

I'm especially worried about the scenario where somebody writes a client which stores your coins in hundreds of little wallets, and sends and receives wallets instead of coins, completely bypassing the transaction fees and coin destruction. If everybody switched to that method, including the exchanges, then no coins would ever be transferred through the protocol, nor would they ever be destroyed. And people would definitely have a big incentive to move in that direction, even if it didn't get that extreme.

That's not going to happen. When you transfer a wallet you can keep a copy of it. The recipient must move them through the block chain or trust you. And if they need to trust you, you've lost the main advantage of scarce moneys. Now people have to trust each other and not be completely anonymous: people probably prefer Ripple for that. You can denominate IOUs in stablecoins, but again you need the decentralized price index system.

The destruction of money through transaction fees has other problems:
Is the transaction fee voluntary?
Why would the merchants include a transaction with a big fee sooner than one with a small one?

I think you need demurrage for this system even if you don't like it.

You are quite right that only trusted transactions would skip the fees and/or coin destruction. I hadn't thought that through all the way. So buying coins from a trusted exchange could potentially be free, but selling coins to the exchange would definitely require paying the fee since the exchange doesn't trust you.

If you want widespread adoption, there's a huge marketing advantage to getting rid of demurrage. I think morpheus' transfer fee idea has all the advantages of demurrage without the drawbacks. When the price of coins is at or above the target, there would be no fees or penalties of any kind.

Demurrage has many advantages for its users than just providing a means of money destruction.
It lowers interest rates, drives transaction fee prices down (or improves security) and recovers lost coins.
Merchants should accept a currency that his customers are willing to spend. But let's discuss these other advantages in the freicoin thread if you don't agree or don't see them as advantages.

If you want to influence prices by changing the monetary base, changes will have more effect if there's demurrage, because demurrage increases the velocity of circulation, and having a high V multiplies the effects of increasing/decreasing M.

M * V = P * Q

You can change both, the amount of newly created coins NCC and the demurrage fee rate.
When you have stable prices, you have NCC = (coins destroyed by demurrage) CDD to keep M constant.
When you have inflation you want NCC < CDD to decrease M
When you have deflation you want NCC > CDD to increase M
When you have inflation you want to decrease CDD to decrease V
When you have deflation you want to increase CDD to increase V
So finally,

With deflation you increase both NCC and CDD, but NCC more than CDD
With inflation you decrease both, but you decrease NCC more than CDD

If you destroy the money through transaction fees, you don't have direct control over coins destroyed by transaction fees CDT, you just can change the mandatory transaction fee rate MTR.
Even worse, when you increase MTR to fight deflation inflation, V goes down, reducing CDT, making you increase MTR even more...a positive feedback that takes fees to the sky and stops transactions completely. It is better to have deflation inflation than no trade at all.
I don't think that morpheus's proposal can work as it is.

Edited

I don't think destroying coins through higher MTR will cause the spiral you predict.

If the transaction fees are increased to avoid inflation, then it will cause people to avoid transactions. However, it would also cause people to ask more for their coins to cover the transaction fee which would cause deflation which is the goal of increasing the transaction fee in the first place. The would break the transaction fee death spiral and stabilize the price of the coin.

Actually, the transaction fee may never get paid if the threat of a transaction fee causes people to hoard Stablecoin and drive prices back up.

Maybe you're right and this can work only with fees.
With demurrage you still have direct control of how much will be destroyed while with MTR you don't.
Also the increased velocity amplifies the effects on price that money creation/destruction produces. So you have to create/destroy less quantity to achieve the same effect.

I very much like the idea of stablecoin. This would be a version of bitcoin that's better suited for commerce than the original. In order to use a currency in commerce, stability is very important. No merchant wants to constantly have to change prices.

What I do not like in your proposal (and in your subsequent amendment to 1971 coins) is the proposal to target a certain dollar value. Someone said Euro would be better, but I think we could do away with targeting a fiat currency value altogether.

We would need a good way to release more coins into circulation (which is available in the bitcoin software, as it is).

We also need a good way to lower the number of coins in circulation, in case the user base contracts. jtimon has proposed to do this by introducing demurrage, i.e. to take a small percentage of the amount transferred at the moment of transfer, and to destroy that small percentage. The percentage would increase depending on the time the coin had been held onto before being spent, making demurrage a powerful incentive to not hold on to stablecoins but to actually use them in commerce.

How to achieve stability without external input

The way to have a stable currency without the need for either a price index or a steward of the currency holding the value constant, is to target a fixed ratio between the size of the stablecoin economy and the total number of stablecoins extant. It is not really important what value that ratio produces, the important thing is that the ratio can be kept stable with only internal inputs.

The inputs we would need are:

- volume of trades in a determined time interval (I believe the system can supply that data).

- number of coins existing at the moment (also a system internal datum)

Using a fixed ratio (I am saying just for example 1:10 but it could be anything that's decided) we can now calculate the target amount of coins. 5,640 in trade would give us a target of 56,400 coins. Please don't hold me to the numbers, it's just a made up example.

Since demurrage is constantly lowering the amount of coins in existence, we can now adjust the degree of difficulty of creation of new coins to get us as close to the targeted coins total as possible.

Since we have the volume of trades as an internal input, we can constantly update the targeted coins total.

Since the value of a currency depends on the number of coins times velocity of circulation in relation to the volume of trade, we have a reasonably stable currency that does not depend on external inputs.

We do not know what the value of each coin will turn out to be before putting this in practice, but it is of no concern because all we want is stability of the value of each coin over time.

Could that work?

It might work technically, but I don't think it would work from a marketing perspective. That is, holding coins denominated in gold, oil, or 1971 dollars is a very simple and appealing concept.

Morpheus, I've been thinking and thinking about your proposal, and I am now completely convinced that your way is the best way. It is simple to explain and seems reasonably straightforward to implement.

The biggest programming challenge is not deciding the rules of the new currency, nor is it importing public data on inflation, the price of gold, etc. The biggest problem is determining the current market value of the coins in circulation.

Once there are dozens of exchanges running which trade your new coins, this gets a lot easier, as it is just another public data source that the miners import, and any client can reject a block that doesn't have the right exchange rates encoded. But until then, figuring out how many new coins to distribute in a new block in a way that all the clients can agree on is a tricky problem.

There are a couple ways I see that the network can get this data in its infancy:

1) Launch an exchange for your coins at the same time you launch your new client
2) Build in a distributed exchange between your block-chains and the bitcoin block chain into your software, then you know the ratio of prices between your coins and bitcoins, and the price of bitcoins is of course public data.

I'm not sure which option is harder to implement, but I note that the first option is a single point of failure while your coins are new. I also like the second option because it allows bitcoins to still have a role in this new (much bigger) economy, at least for awhile.

I assume you are planning on riding along on the bitcoin block-chain using merged mining?

I don't think so. Read this:

https://bitcointalk.org/index.php?topic=26380.msg330368#msg330368

An alternative way to achieve this would be a method I call "one block, one vote". Under this system, everytime a block is mined, the miner may specify the amount of coins that will be received by a future miner a given number of blocks into the future. To prevent wild oscillations, the amount that can be specified would be limited to some range around an average of previous blocks. Under such a system, the price level would be determined by the equilibrium between new miner hash power trying to push up the inflation rate and existing wealth holders trying to push the inflation rate down. My hypothesis is that such an equilibrium would lead to roughly stable prices.

I am currently working on a patch for Multicoin ( http://bitcointalk.org/index.php?topic=24209.0 ) to enable blockchains to use this method. Since I am neither a professional programmer, nor an expert on the bitcoin source code, I welcome help from those interested in this topic.
 

If you have a look at how the bitcoin exchange rates changed over the last months without any changes to the mining awards having been done, you will notice that the amount of coins being created is only a minor factor towards stability. When will people start creating actual services using bitcoins and thereby stabilizing the currency in the most natural way?

With predictable mining awards, we can at least ensure that the computing power that keeps the network running is stable. When we take away this basic (computational) stability, I wonder how a decentralized currency is supposed to work at all.

I do not get Goldcoin. I mean, I understand the wildly awesome idea of selling free coins for the price of Gold. However, I do not understand why anyone would buy them instead of Gold itself. There is tremendous risk that Goldcoin would not actually match the price of Gold. There is exactly 0% risk, however, that Gold will not match the price of Gold. Saving transaction fees is not good enough to make up for the risk. I store gold. I do not pay transaction fees anyway.

So, lets say the first block of GoldCoins come out. What if nobody wants to buy them on an exchange? Simply reducing new supply will not help. It would be Dead on Arrival. Destroying coins will not work either. Why would I buy coins that can be destroyed? Why would I use transactions that have fees significant enough to make any difference on overall inflation/deflation rates? They would have to be huge. No thanks.

Even if the first few mined coins could be sold, the problem is new on every new block. If there are not enough buyers, GoldCoin would drop and may never come back. Price fixing on the exchanges cannot create demand, either.

I suppose you could destroy coins uniformly to get the price up. But, remember, you are destroying someones real wealth when you do that. If you need to delete half my GoldCoins to get the price up, how exactly does that help me? I paid for 10oz of gold and if I now have 5oz that means I lost money.

I am not saying Goldcoin is impossible. I just think that the equilibrium that keeps the price right would likely be a very small population of true believers. (You might be generating 1 goldcoin per day to avoid oversupply and God forbid if demand drops too fast.) Everyone else would find real gold much less risky and coin shops more convenient than Paxum/MtGox or whatever it takes to get GoldCoin. Or they buy bitcoin if they want a new currency. At least that has potential upside.

I prefer using difficulty rather than txn volume to establish the target. Difficulty is directly related to coin value. You will see this in months to come now that difficulty levels are approximately in equilibrium. The ratio of txn volume to value depends on what share of people are hoarding coins rather then spending them. This will likely be somewhat unstable. Nevertheless, I think this idea is uch better than goldcoin and stablecoin. These external data-pegged coins are similar to liberty reserve and egold. They drop decentralization which is perhaps the key tech advance in bitcoin
 i see this as moving backwards.

the objection here is that txn data is meaningless. You can easily fix this by imposing a 0.1% tax on txns. Suddenly tax collection volume becomes a meaningful measure of txn volume. You need a tax like this anyways to keep mining sustainable. Two birds with one stone.

I was thinking in free transactions. But I guess that could work.
I still think that you can keep mining sustainable with demurrage, maybe even only with voluntary fees.

Merchants will not give up even the smallest percentage of their profit margin unless they have to. A mandatory TX fee is the only way to extract blood from that particular stone.

I think demurrage could work, but that users would prefer txn fees. Simply because people won't like the idea of money they have forgotten about slowly being drained away.
It is like signing up for a credit card with annual fees. I would much prefer that they gouge me only when I purchase something, rather then when I am not paying attention.

Merchant are also giving a percentage of their margin profits when customers pay fees. Demurrage fees are part of the merchant costs and customer will pay them indirectly just like they pay for bank fees.
And demurrage can be much cheaper then those.
It is possible that users prefer tx fees, though. Specially if they're not aware how much they will save thanks demurrage.
Most people think that they don't pay interest if they don't borrow money, but that's not true. Interest is factored in the price of every item you purchase.

As of today, you are wrong: many bitcoin merchants still change prices very often, even though this hurts stability. When they start to grow up and offer their services for dependable prices (I would call it that way at the point where they don't change BTC prices more often than e.g. USD prices), there will soon be no market for bitcoin adaptions with an additional single point of failure added, like stablecoin or goldcoin are.

I would not see them as an alternative to bitcoin, and neither would most of the people I work with, as these approaches backdoor the dependability which bitcoin achieves through decentralization.

Stablecoin is not beertoken. It tries to achieve stability within a decentralized system.
Maybe it can't work, but it has no single point of failure.

The source of info on exchange rates is the single point of failure. Ever heard of the mtgox flash crash. Yup, that's failure.

You are assuming that the protocol would dump a ton of coins in the first block. But what if rather than dump 50 ounces worth of gold coins, it only gives out 0.00001 ounces? That is worth about $0.02, and I don't think you would have trouble with oversupply.

I believe the protocol will need to err on the side of scarcity, especially early on. Artificial scarcity won't generate a massive bubble, since everyone knows the price will eventually converge with the underlying commodity, so it would be silly to pay 2x what they will be worth in a couple years. However, it will give people more confidence to hold these coins if there is more demand than availability.

I predict USDCoins (not the stable 1971 variety) will actually be the most popular one among the masses here in the states, not because it is a good investment, but because it will be easiest to understand. In other parts of the world I expect Eurocoins to be equally popular.

One thing I love about this idea is how extendable it is. If people want to trade coins denominated as Google stock, boom, Googlecoins are born. You just need agreement among all the clients on what data sources to trust and how to deal with data sources getting hacked.

It would be sweet if the integration with bitcoin will allow me to hold my money as goldcoins, stablecoins, etc, but then spend them to any bitcoin or goldcoin address. If it is a bitcoin address, then behind the scenes, the client would exchange the gold coins for bitcoins, and then send the bitcoins.

Any coin traded is going to need multiple independent data sources, and some way to deal with data sources getting hacked. Keep in mind that the external price only sets a target which is approached over time by controlling supply. The actual price of the coins is set by supply and demand. A flash crash isn't going to change what people are paying for goldcoins. A sudden change like that might lower the amount of coins being generated, but it would not raise transaction fees - I expect they would come into play very slowly over a long period of time.

If the protocol sees that its data sources have wildly different prices, it won't know which target to use, and the rules to deal with that will need to be flexible enough for everyone using the coins to agree that "at block 1024, we'll stop using this compromised data source and add these two new sources". If your client doesn't vote on a decision like that, you get stuck with the new target that everyone else decides they want.

What if two of your three sources for gold prices get hacked, and there is significant disagreement on which new sources to use? In that case, you might split off a new block chain so you have two chains which recognize the old coins, but each chain uses their own rules for new coins (and new coins wouldn't be equivalent). Any old coin would get onto one of the new chains the next time it was spent.

I can only think that only a real market can control a price of any commodity or currency and the a central bank that dumps and buys as needed to control the price would be the only way I know to make a stable currency possible.  as I said before you can peg the value of a currency to anything you want.  It's "The Trust" holders that should decide what they want to hold as there base of value of the asset holdings and just use the crypto coins to pass the value for P2P transactions.  I have almost completed the infrastructure to make all that possible in the BeerTokens model http://bitcointalk.org/index.php?topic=9493.msg138247#msg138247.  The first beerA coins have already been minted with the new merge mining feature using MulitCoin-exp https://bitcointalk.org/index.php?topic=24209.msg300830#msg300830.  With MultiCoin  we also can setup secure exchanges with escrow deposits to prevent third parties from stealing from a central exchange.  I will continue to develop what I feel is missing in the infrastructure to make what  sounds some of you want in the near future.  The basic concept is that the holders of "The Trust" decide how they want things to be and what needs to be changed, Not the developers and miners.  Each holder on record has a voice.  You just have to make it heard by being a part of it.

We don't know how long it will take for bitcoin to get stable, or how stable it will be. Stablecoin gives us known stability right now. Also, this method of tracking an external price has a HUGE number of applications. Stablecoin is only one of them.


Like you, I started with the idea of a fund which would be collectively held to back up the value of new coins/tokens released. I eventually became convinced that Morpheus' idea was better due to its incredible simplicity.

I haven't tried to use Multicoin, but morpheus stated he was unable to get it to work. The future of distributed currencies is definitely going to be one client which is able to support multiple block chains, and hopefully even a distributed exchange built in for converting between them. I suggest you put some effort into getting morpheus set up to use multicoin!

So many questions awaiting empirical answers!  I strongly suspect that someone can design a better cryptocurrency, but I set high standards and look forward to seeing these and many other proposals in competition.

Some good software could open the field of currency design to anybody with an idea to test, not just C++ programmers with large time budgets.  Namecoin and MultiCoin have made good progress, but we have a long way to go.

I am thinking specifically of:

• Bitcoin client plugin interface: plugins can define block acceptance rules.
• Bindings to facilitate plugin development in high-level languages.
• Set of highly configurable plugins extending MultiCoin's advanced configurability.
• Multiple currencies handled by one client process.
• Multiple currencies on one p2p network with a facility for negotiating which chains to relay to which peers.
• Further development of merged mining:
  
  • Multiple auxiliary chains (http://dot-bit.org/Merged_Mining) mined together.
  • Multiple proof-of-work formats per chain.
  • Effective difficulty able to vary according to proof-of-work size/complexity.
• New type of "inv" message for broadcasting a new currency definition (plugin code and parameters).
• Block chain browser (Abe) and exchange (Bitcoin Central?) automatically supporting new currencies.
• Graphical wizard/tools to put it all together: plugin architecture, configuration, testing, and launch to the network.  Marketing is still a Phase 2 feature.  ;)

I have the C++, network, database, and thread programming background, and I will eventually do this if I live long enough, but I can not donate enough time to it in the foreseeable future.

Will you help me elaborate this vision until someone picks it up and runs with it?

I think this is done in multicoin: "Multiple auxiliary chains mined together."

In the future, you may want to read the OP carefully in order to avoid disqualifying yourself like you did in this case. To quote the OP on stablecoin: "From then on, we keep track of how much USD has inflated/deflated using something like CPI or the Billion Price Index or even a combination of indices". Have a look at the wording: The idea is to introduce a single point of failure, either through CPI or Billion Price Index. "Or even" multiple points of failure through a combination of indices. But a stable, decentralized system is completely out of the question. Basically, stability is completely sacrificed, but ironically it is put into the name...

I know how stablecoin would work. The source of conflict may be our definition of "single point of failure".
To simplify, let's use goldcoin. Miners would report the price in goldcoins of gold (getting information from various exchanges and probably averaging them).
You may think that gold and goldcoin exchanges are the points of failure in this case, but it can be discussed.
As I see it, instead of a CPI index miners would report spot prices from various exchanges and then the price index would be calculated from those spot prices.
I don't think that you can target the price of any commodity or that you can have a totally stable coin just through changing the money supply, but it could be more stable.
Don't know what the OP is, but if you mean this thread, yes I've read it.
Maybe you didn't understand me because some of us have been discussing some topics of this thread in other threads.

This is amusing since I read somewhere that stablecoin is supposed to be attractive for merchants. I would like to meet a merchant who prefers a currency that will block the market from time to time...

The points of failure remain, no matter what verbal contortions are put in this thread, in other threads, or anywhere else in order to disguise them.

You cannot have a totally stable coin, period.

The value of any currency is still dependent on how well the economy works. When the economy suffers, you can't do as much with your money, no matter how it is called.

But you can limit the effects, which is what bitcoin does: Other than a bad economy, it can't be hurt much. However, most ideas to "improve" bitcoin which came up lately just introduce more potential weaknesses: external currencies which may be wrecked through currency wars or simply by incompetent politicians, external commodities which may lose in value, external data sources to gather the value of the currencies or commodities, which may be manipulated, may fail etc. Essentially, you are making a mere toy from an experiment that may succeed (bitcoin).

If you really want to remove points of failure, you may want to make every node an exchange. But then, there is no point in weird coin supply strategies including artificial stagnation through transaction fees. Just make the exchange 1:1, and you have a distributed currency or commodity transfer network, which might be very useful. Should it manage to fly, I might even want to use it. But in terms of long-term stability, I still think it would not be able to compete with bitcoin.

The points of failure remain. You may have found a couple of people who don't mind in the other threads you mention, but it won't change the fact that those who care will still see the flaws even if you plainly neglect them.

I see that a decentralized system for tracking exchange prices has the points of failures in the exchanges miners watch, but still is an interesting concept. My proposal for the best use of this system is to define a nonexistent reference currency that is just useful for contracts. But it lacks a system to reward miners so it should be inside another chain. Another use could be a a currency for decentralized option trading. You need the spot prices in that chain anyway.

So how about we contemplate the doomsday scenario for stablecoin:

• stablecoin starts out wildly successful, miners get rich, everybody is happy
• Stablecoin 2.0 comes out, and nobody wants the original stablecoins anymore
• Everyone tries to sell their stablecoins all at once, price drops 99%
• All confidence in the protocol is lost except a couple hardcore believers holding out hope that the protocol will somehow correct itself

What should the protocol do to transfer fees in this case? Is there a 90% transfer fee? 50%? 10%? If the transfer fee is too high, who is going to want to buy stablecoins? Transfer fees penalize buyers as well as sellers. If nobody wants to buy OR sell, there is no incentive for prices to go up. Stablecoin simply dies.

It would seem that the transfer fee needs to be pretty small even when the price diverges by a large amount from the underlying asset, otherwise the new currency is completely ruined. On the other hand, if the coins cease to track the underlying asset, nobody wants to hold the coins anymore, and the currency is completely ruined.

Morpheus - have you considered how to handle a doomsday scenario like this?

I think a more direct control over the supply of coins is required. Some mechanism needs to be created that allows the user base and interest in stablecoin to shrink by a large amount while still maintaining stablecoin values.

What if stablecoin sells shares like a company? Instead of distributing all new stablecoins to miners, some of them could be distributed to shareholders.

IN ADDITION to a transaction fee when the stablecoin prices drop, the protocol could sell new shares in exchange for stablecoins, taking stablecoins off the market.

Obviously if NOBODY wants stablecoins, there is no way for it to survive, but I think an additional measure such as this could help it survive a much bigger hit.

What do you guys think? Morpheus, I'm especially interested in your thoughts, since this is your baby.

Another note on the doomsday scenario:

Even if you have shareholders as described above, a catastrophic loss of interest in stablecoins would eventually send share prices to zero, and there would be no further leverage to take coins off the market.

At this point, the shareholders are bankrupt, and it would seem wise for the protocol to "declare bankruptcy", wipe out the old shares, wipe out the excess supply of stablecoins (replacing them with some new post-bankruptcy shares), and go from there.

It sounds as if whatever it is that your coins are "shares" of has no assets.

If a corporation, for example maybe General Mining Corporation, actually owned assets, such as, for example, a bunch of mines, a bunch of mining rigs, or even a bunch of shares in various publicly traded sold, silver, platinum, paladium, iron, copper, etc etc etc mining corporations, and used a blockchain whose genesis block issued all coins ever to be issued of that blockchain, issuing them to GMC's treasury... then GM could use its actual assets to buy back any GMC coins it had issued at the same price it had issued them at or close to that price.

-MarkM-

The asset is a share in the stream of future coins that need to be distributed as the network grows. Miners would be paid in shares, which they could sell on the open market. When more stablecoins need to be issued to keep prices down, the shareholders would get them to keep or sell.

A share in a revenue stream does have value. The exact value I would leave up to speculators!

Edit: removed option expiration.
How about a very limited type of exchange within the main chain:

Every block contains a price inserted by the miner.  A block solution must contain a signature on a standard contract using the coinbase output keypair.  The contract is an option to:

    buy at the quoted price plus a standard spread (e.g., 1%)
    or
    sell at the quoted price minus a standard spread

an amount of the underlying commodity equal to the block reward.  The option can only be exercised by solving a block with this block as its parent.  To exercise such an option, a miner must set a special flag in the next block.  The two miners have to find each other and settle up (no need to specify how) before either can claim his block reward.  They must both sign something using their coinbase keypair and submit the signatures to the network before the network will recognize their block rewards.

Perhaps the spreads don't have to be standard, the contract can specify any prices straddling the quote.  Miners would be motivated to offer good terms in order to encourage the network to accept their blocks and thus validate their rewards.

I'm not sure what would happen in a total doomsday scenario, but I doubt the above would happen. I don't think a new "better" currency would cause everyone to suddenly jump all at once away from Stablecoin. Instead, it would be a gradual migration where the price of Stablecoin would be consistently under the target. The transfer fee would kick in to destroy the Stablecoins at a regular rate which is exactly what we would want.

A doomsday scenario which is very likely to occur would be like the Mt Gox hack. What happens if a bunch of Stablecoin was stolen and the thief doesn't care about the exchange rate and just dumps the coins on the market. Since something like that has already happened with bitcoin, we can assume this would happen with just about any digital coin.

This scenario will cause some issues, but I think Stablecoin would survive. Lets look at each step of the thief's sale to see how Stablecoin would react.

1. The thief initially tries to dump all of his stolen coins on the market. All of the bids are cleared from the market and the price of Stablecoin drops. If the thief is able to dump all of his coins, he gets away with the money, and a bunch of others buy really cheap Stable coins. If not, He has to wait for more bids.

2. The thief still has more coins to dump and some of those with cheap coins may want to sell at a slightly higher price than they bought them at. If the transaction fees haven't change yet, this is really step one with the addition of more people selling below the fair value. When the transaction fees change, the program will now calculate a very low value for Stablecoin and end up with really high transaction fees. If the thief tries to sell, he will lose much of his stolen coins to transaction fees. Also, anyone who purchased cheap Stablecoins and wants to sell out, or anyone panicking will also pay high transaction fees. Some people will choose to pay the transaction fees, others will not.

There's the important distinction, choice. Some people will choose to return some coins to the either because the coins didn't cost them anything to begin with or they have lost faith in Stablecoins. Either way, coins are destroyed. Hoarders can hoard their coins until the price comes back up. Of course, with so many coins being destroyed, the price will come back up.

Also, the transaction fees are a disincentive for the thief to dump his coins all at once. If he does, he won't make as much money as he would if he sold them slowly. There will likely be some level where he doesn't care about the transaction fee and just wants to dump the coins. But, that would just allow some people to buy cheap coins for a while and allow some of the stolen coins to disappear.

3. Finally, some coins are destroyed, others end up with some cheap coins. The big downside, Stablecoin isn't as stable as the name suggests. It will be influenced by the market, but the goal is to always go back to a specific price. I believe it will do that even in this doomsday scenario.

There is, I believe, a much simpler way to exchange between bitcoins and stablecoins.

If the stablecoin protocol has visibility into the bitcoin network, then stablecoin users can mark their coins as "for sale" in the network, along with a bitcoin address and price.

When the stablecoin network sees bitcoins go to the address, it releases the stablecoins to the buyer.

There are a couple issues to work out, but I believe it would work. One issue is what happens if two bitcoin users try to buy the same stablecoins at the same time (answer: a buyer could send a tiny amount of bitcoins to "lock" the coins before sending the bulk of the coins. The first person to lock the coins has the right to buy them for a set amount of time.)

Sure.  My proposal is about linking to prices outside the cryptocurrency world.  The bootstrap currency, if you will.  E.g., Goldcoin.

I think only a real-life experiment can determine if you are right. How high should the fees go, in your opinion?

One thing I don't like about high fees is that they discourage commerce if stablecoins happen to be priced too low rather than too high. If you want the ideal coin for commerce, this doesn't work as well, since 50% of the time there are fees and 50% of the time there are not once equilibrium is reached. I agree it might work pretty well if it is mostly seen as a store of value.

One other thing that bothers me is the user who just wants to store value, and access it later when they need it. If they happen to need the money while the coin-base is shrinking, you have effectively cut off access to their funds unless they want to surrender a big portion of the supposed value of the coins. I realize that is the point (to discourage people from cashing out when prices are too low), but from a marketing perspective adoption would be much more widespread if someone else (shareholders is my vote) was bearing most of the shrinkage risk.

You may be completely right that this will work with only tiny fees and gradual destruction of coins when necessary, but any violent price swings are going to make these coins less ideal for both commerce and store of value.

If that volatility risk can be transferred somehow as I described, these coins become much better behaved, both for commerce and for storing value. The "bankruptcy scenario" I described above could fail gracefully into the fee-based system you describe, but during any normal market, you would get nearly perfect stability without any fees.

I realize I'm all over the place here, but here is another way to decrease they coin supply temporarily: have the protocol sell bonds. In this way, the protocol can borrow coins from users (reducing the coin supply), then make interest payments, and return the money at maturity. This would be a good way to handle a temporary oversupply of coins, but not a good way to handle a coin supply that needs to keep shrinking (since more coins are ultimately introduced by these bonds through the interest payments).

Maybe such people are what futures and options are for?

If you want a loaf of bread in five years time, and are not sure how the price of bread might fluctuate by then, you buy a five year future loaf of bread.

If you want a loaf of bread anytime during the next five years but are not sure exactly when during that span of time you will want it, you buy a five year span open option on a loaf of bread.

Is that not how futures and options work?

-MarkM-

So people would have an option and mechanism to buy the bonds?  I kind of prefer a dynamic demurrage-like factor that just multiplies every coin's value by a number based on the linked commodity price/index.  People are already used to seeing their account balance go up and down by small amounts ("interest" and "fees" or perhaps "taxes").

Yeah. It was just a crazy thought that crossed my mind. I doubt it would be as useful as having shares.

There's one guy in particular who is really hot on demurrage. I don't think anybody will ever hold coins with demurrage though if there is another option available. It's a non-starter from a marketing perspective.

Quite true. However, I think the whole point is to bypass and/or hide the complexity of futures/options from the user. A client that tells you that "your goldcoins = 4.31 ounces" is very appealing if you can provide enough confidence that they will STAY worth that much.

How about we create a very simple system wide program with the sole purpose of keep the price of these new coins stable.  For fun we could call it the "Feral Reserve Program".  If the value of the new coins is dropping because there are too many of them in circulation then this hypothetical program would sell some sort of AAA rated bond or debt instrument like US treasuries – oops those are not AAA anymore - in order to reduce the number of coins in circulation.  And when there are not enough coins in circulation the program would buy back the debt instruments in order to increase the number of coins in circulation.  Hmmm, this is all starting to sound vaguely familiar…

I think people who clip coupons, at least - and they are imagined it seems to be normal soccer-mom types, aren't they? - should be able to understand things like "pay to the bearer upon demand one loaf of bread" and "pay to the bearer on 25th December 2016 one loaf of bread", and that is futures and options right there, isn't it?

-MarkM- (Admittedly I don't clip coupons - too complicated per dollar-per-hour that it pays...)

I agree high fees will discourage commerce, but I don't see any way around it. However, I don't think high fees will be all that common. The only way to find out is to create the coins and see what happens.


I don't see how you're going to avoid volatility entirely. Also, I don't get your shareholder proposition either. I think a certain amount of volatility will be necessary to live in a decentralized world. You could remove the volatility completely if you have an issuing authority, but then you have to rely on the issuing authority to do the right thing all of the time. If you're fine with that, just go with USD and the Fed.

The problem with demurrage is apparent with the thief scenario. If a thief steals a bunch of the coin and sells it for a very low price, the demurrage would take the coin from everyone based on how much they are holding. This gives the thief and incentive to dump their coin as fast as possible to avoid the demurrage. It also gives others hoarding coins an incentive to sell out before the higher demurrage kicks in. You could suddenly have a spiraling inflationary problem as everyone tries to dump their coins before demurrage.

The opposite could happen if someone buys a bunch of coin at once and a reverse demurrage kicks in. People would then try to buy as much coin as possible to get the most reverse demurrage fees, thus driving a deflationary spiral.

Good point.

Instead of that dynamic interest/demurrage you could just increase/decrease people's balances automatically and instantly (https://bitcointalk.org/index.php?topic=11614).

In my proposal for using demurrage instead of destructive tx fees, the demurrage is reduced when the price drops and gets higher when the price rises, so it doesn't have this problem.

Morpheus,

As you probably know, I've described something very similar to goldcoin/stablecoin targeting mechanism in my own thread. I give you some credit here: https://bitcointalk.org/index.php?topic=36453.msg451254#msg451254

I'd be very interested in your opinion if you have a moment to post in that thread.

Further elaboration of the GoldCoin / {Insert Commodity}Coin idea in this thread: https://bitcointalk.org/index.php?topic=24714.msg523685#msg523685