*old* BitShare Economic Theory 10 BTC bounty to prove me wrong... paid.

[rethaw]

Does one ever get a mango? What I'm reading is your system allows no trust speculation on the value of other currencies against a bitshare, but how does exchange occur? If I understand correctly I have to find someone willing to take bitshares (EDIT: or crypto-mangos) for mangos, not crypto-mangos, correct?

The Bitcoin exchanges provide liquidity and the single point of trust is the only way I can see exchange across currencies to occur. Check out the last posts of JoelKatz for more on how difficult cross currency exchange is.

[1713487235]

1713487235

[1713487235]

1713487235

[1713487235]

1713487235

[1713487235]

1713487235

[1713487235]

1713487235

[bytemaster]

There are no fundamentals relating the price of cryptoUSD (denominated in shares) to the price of actual USD (denominated in shares).

Say you take x shares and turn them into a cryptoUSD. Now you have a cryptoUSD backed by the discounted present value of txn fee revenue from x shares.

Say you take x shares and hold them. Now you have x shares backed by the discounted present value of txn fee revenue from x shares.

Ignoring liquidity issues, the value of the cryptoUSD should remain constant when denominated in shares. The USD value of a cryptoUSD will go up and down, just like the USD value of a share will go up and down.

You're going to tell me that the cryptoUSD will go up and down in share-denominated price, etc., etc. Why? Suppose it doesn't? Isn't that an equilibrium?

You need to a) create an incentive scheme (e.g. reward issuers that create frequently transferred assets) or b) peg your derivatives to something with a physical interpretation (e.g. PoW difficulty).

I suggest you save your money for a well thought out project.

You overlooked the critical ingredient that makes this work.  

1) when the crypto-USD is created, it is actually created slightly above market value.  New crypto-USD can only be created when there is no one willing to sell existing crypto-USD at the current crypto-USD to BitShare exchange rate.  

2) all crypto-USD is fungible.  So if you only look at 1 transaction, then what you say is true.  But if the price changes between crypto-USD then the 2nd, 3rd, and 4th depositors will demand more and more BitShares to back the new issuance of crypto-USD such that when each of them receives their crypto-USD in exchange for paper-USD then it would be near parity.   Otherwise they wouldn't buy crypto-USD they would just buy BitShares directly.   The reason they buy crypto-USD instead of bitshares is because they know everyone else who also WANTS a USD denominated crypto-currency will demand that it have near price parity before they would give up paper-USD for crypto-USD.

3) If crypto-USD failed to maintain parity with actual paper-USD then there would be no depositors.  Therefore each and every depositor represents a market force that corrects the ratio of dividends between BitShares and crypto-USD.

[bytemaster]

Does one ever get a mango? What I'm reading is your system allows no trust speculation on the value of other currencies against a bitshare, but how does exchange occur? If I understand correctly I have to find someone willing to take bitshares (EDIT: or crypto-mangos) for mangos, not crypto-mangos, correct?

The Bitcoin exchanges provide liquidity and the single point of trust is the only way I can see exchange across currencies to occur. Check out the last posts of JoelKatz for more on how difficult cross currency exchange is.

Someone who is selling mangos could take payment in 3 forms:  1) $USD   2) BitShares  3) crypto-Mangos.   At the time they accept payment for the real mango all 3 forms must have 'equal value' or they will not sell their mango.   

The difference between selling your Mangos for $USD and BitShares is that BitShares are far more volatile and thus the vendor has exchange risk.   If instead the mango vendor wishes to keep his balance in mangos (so he can get a mango next week) he can ask for a crypto-Mango.  Everyone else who is selling mangos for deposits that they want to maintain parity with mangos would also demand crypto-Mangos and demand that the value of crypto-Mangos at the time of exchange was about equal to a real Mango.   With everyone with a real mango that they want to deposit demanding parity, those that covert BitShares to Crypto-Mangos would have to do so at different rates over time as the price fluctuates.  Thus the depositors are the ones that demand crypto-Mangos have parity (while the price of Mangos is going up).

If the price of mangos fall (hence BitShares are going up) then the value of the dividends (in BitShares) paid to holders of crypto-Mangos goes up.   This means that the shorts who exchanged at the 'old rate' would have an opportunity cost.  They could receive more interest by buying back the crypto-Mangos at the new lower price.  This would serve to close out old short positions (reducing the interest paid on crypto-Mangos) and also causing the value of crypto-Mangos to fall with the value of real Mangos. 

How do they buy-back crypto-Mangos?  They have to provide real mangos to the market.

[rethaw]

So is it fair to say this is a system for no-trust speculation on the other commodities as opposed to an exchange?

By exchange I mean a way to put in currency one and get out currency two. The problem you are trying to solve is hard. I think you have arrived at distributed valuation, an interesting, though different problem.

EDIT: For example, for your system to act as part of an exchange:

They have to provide real mangos to the market.

So instead of having to convert BTC to USD to perceived value of a mango, we can use your system to value a crypto-mango in Bitshares directly by seeing what positions are held in the blockchain.

[bytemaster]

Added section to white-paper... (a bit redundant to prior replies, but hopefully useful to people here)

I’m still not convinced that crypto-USD will track the value of paper-USD, can you give me any more reasons to trust the value of crypto-USD?

Every depositor who wishes to trade paper-USD for crypto-USD in order to receive interest will demand that the value of crypto-USD at the time of exchange is about the same as paper-USD.  If not they would not accept the trade.   This means that as the price of paper-USD goes up (and therefore BitShares go down) the issuers of crypto-USD will have to do so at ever higher exchange rates to maintain parity.   This increases the interest rate paid on all crypto-USD which makes them interchangeable.  Thus every deposit represents a ‘check-point’ in the crypto-USD price.  Because there is a market/demand for a crypto-USD priced at parity the market will provide it.   After all, the buyers of crypto-USD have paper-USD that the holders of BitShares want.  This covers the case of a rising USD value.  What happens if USD falls?

When USD falls then it means BitShares are going up.  That means the value of the interest paid on crypto-USD is going up and therefore the interest rate (denominated is USD) is going up.  The implication here is that the shorts who created crypto-USD at the old exchange rate would have an opportunity cost by maintaining their short position at the old (higher) exchange rate.  They are foregoing dividends on 10 BitShares used to back their crypto-USD issuance when they could now purchase crypto-USD for 5 BitShares after the price change.  This means they could by a crypto-USD for 5 BitShares and start receiving 10 BitShares of interest.  This market force causes them to close out their short-position even though they bet right, BitShares went up and Crypto-USD went down.

[bytemaster]

So is it fair to say this is a system for no-trust speculation on the other commodities as opposed to an exchange?

By exchange I mean a way to put in currency one and get out currency two. The problem you are trying to solve is hard. I think you have arrived at distributed valuation, an interesting, though different problem.

EDIT: For example, for your system to act as part of an exchange:

They have to provide real mangos to the market.

So instead of having to convert BTC to USD to perceived value of a mango, we can use your system to value a crypto-mango in Bitshares directly by seeing what positions are held in the blockchain.

Once you have crypto-USD and crypto-Gold that track the value of Gold and USD then you can trade Crypto Gold and Crypto USD via the blockchain and then withdraw it as Gold or USD.  Thus it is also an exchange.   

[rethaw]

[...] and then withdraw it as Gold or USD. 

I'm missing how that part works.

[bytemaster]

If crypto-Gold has a market value equal to gold, then you should be able to sell it anywhere for gold or something of equal value.   Thus, you could post an ad on craigslist, visit 'local-bitshares' or exchange with a friend or family member.   

Because it has market value of its own (without depending on any counter-party or IOU) it should allow you to purchase gold.

[calian]

OK, I was going to respond to your other thread but I'll start here. From what I've gathered from your initial rebuttals of the first skeptics you're proposing to create a synthetic cryptoUSD, cryptoXAU, etc. based on market forces. However assuming there is no central vault or receiver someone buying one of these with real dollars or gold would immediately be paying out to someone selling one. The synthetic chits would not represent real dollars or gold held anywhere, only the expectation (hope) that someone down the line will need to buy them with the real instruments. Is that correct? If not you should provide a narrative of how a hypothetical person would use the system.

For example, I post up an ad on craigslist offering $100 USD to someone who will meet in person and sell me $100 cryptoUSD? Where would these initial cryptoUSD come from if there's no central backer? There seem to be serious flaws to this even beyond the serious flaws Ripple will encounter. I still don't understand enough about Open Transactions to comment on their model but the idea of federated trust entities (reminds me of the hawala system) at least has a real-world model to emulate. I just don't see your system working in the real world without a trusted entity to hold deposits (think gateways in Ripple).

[thezerg]

Ok, here's my attempt to collect the 10 BTC bounty.  I am going to flesh out out a large issue in your proposal that I mentioned in my first post and hopefully make you see the problem.  Even if not you, it will make everyone else see it.  Or, if we engage in "design-by-answering-questions" I ask you to consider paying me some small portion of your development bounty.  Because good design is a the most important part of development.  I will publicly acknowledge any receipt of funds so others know you are serious with your bounties. 1CKeoT8vBDQDEpMHz5VAswV39pZJ2GTGYd

The issue is fundamentally "There is no way for USD pricing information to honestly enter the system."  Also, there are several smaller-but-still-critical-issues which I'll designate with [N].  So we'll call the above issue [1].  And in the end you'll see that you end up with a currency that tracks some multiple of BitShares, not USD.

Since you seem to think in examples, here is the problem illustrated by example:

Initial conditions.  1 BitShare = 1 USD,  1000 BitShares mined

Person A creates currency Q, mints 100 Q backs it with (say) 100 BitShares.  So 1 Q is worth 1 USD at this point. Now, the dividends he would be getting for the 100 BitShares are now diverted into paying the owners of the Q.  Right now, that is Person A so the system is in equilibrium.  Money is neither gained or lost by minting a new currency because if Q had not been created, Person A would be getting dividends from BitShares.  Other schemes result in runaway currency creation or a draining of the backing (the value) behind Q [2]. 

Now, BitShares like BitCoin are awesome so in 3 months the price goes to 1 BitShare = 10 USD.  Now those 100 Q, backed by 100 BitShares is nominally worth 10 USD each.  So when somebody bids to buy 100 Q for 100 USD (the original dollar-parity price) nobody is willing to sell for that price.  So the currency is created by someone who offers BitShare backing. 

But how many BitShares is the new currency backed with?  For parity with USD, it must be backed with 10 BitShares.[1]  But who decides that value 10? [4]  No voting scheme works because there is no way the holders of currency Q will agree to that backing level, because they lose money.  Regardless, any voting scheme causes people to vote in their self-interest, NOT to vote in a way that causes Q to track USD. 

Within the system, a trade of BitShares for USD simply looks like a BitShare transfer (the USD portion happens outside the network of course -- it could be physical cash).  Even if you add a field to the transaction to "report" the USD transfer amount, there is no proof that the USD actually changed hands [4].  Or even that BitShares changed hands actually.  It would be easy to transfer BitShares to (another account owned by) yourself with a fake USD amount to drive the reported exchange rate where ever you wanted it to go.


[1] Fiat currency pricing information never enters the system.
[2] runaway currency creation
[3] Proven the short term Q's price follows the backing currency.  Point [3] is that there is nothing else for it to follow...
[4] Bitcoin is based on distrust of all participating parties.  Your system requires some trust or centralization at a minimum to inject fiat pricing into the system, but trust/centralization could also solve [2].


A quote and rebuttal:

Added section to white-paper... (a bit redundant to prior replies, but hopefully useful to people here)

I’m still not convinced that crypto-USD will track the value of paper-USD, can you give me any more reasons to trust the value of crypto-USD?

Every depositor who wishes to trade paper-USD for crypto-USD in order to receive interest will demand that the value of crypto-USD at the time of exchange is about the same as paper-USD.  If not they would not accept the trade.

Within the example given above (where the backing BitShares are now worth 10x), why would buyers demand near parity?  Just because the name is crypto-USD?  In fact, if they did there would be NO SELLERS (at that price).  Crypto-USD owners would just sit there receiving the same dividend payment (denominated in BitShares), but those same BitShares are actually worth 10x USD.  How can the dividend payment be reduced to force parity with USD?  This requires that USD->BitShare price information be available within the system (and its not).

[bytemaster]

Ok, here's my attempt to collect the 10 BTC bounty.  I am going to flesh out out a large issue in your proposal that I mentioned in my first post and hopefully make you see the problem.  Even if not you, it will make everyone else see it.  Or, if we engage in "design-by-answering-questions" I ask you to consider paying me some small portion of your development bounty.  Because good design is a the most important part of development.  I will publicly acknowledge any receipt of funds so others know you are serious with your bounties. 1CKeoT8vBDQDEpMHz5VAswV39pZJ2GTGYd

I appreciate your honest attempt with this post.   If I end up tweaking my design in any significant manner from what I have already discussed as a result of 'bugs' you find in it then I will award a tip proportional to the size of the change.  I have already paid 1 bounty (0.5 BTC) for a logo (though he hasn't yet publicly posted it, I have asked him to). 

Are you willing to back your claims that 'everyone else will see it' with any wager?  It would make this even more fun! 

The issue is fundamentally "There is no way for USD pricing information to honestly enter the system."  Also, there are several smaller-but-still-critical-issues which I'll designate with [N].  So we'll call the above issue [1].  And in the end you'll see that you end up with a currency that tracks some multiple of BitShares, not USD.

Getting honest price information is actually a huge and important requirement.  My definition of honest-price information means it cannot be 'calculated' and must always be the result of an intentional trade initiated by a human actor.  All trades imply both parties 'agree' on pricing information and they will only agree if they compare it to everything else in the world they could trade for.  If you can show me a place in my software where I attempt to use an algorithm to 'fix a price' then I will give you a 1 BTC bounty.  I firmly believe that no systems based upon price-fixing (averaging, or otherwise) will work.  (This is why IOU based systems don't in the end... they ultimately attempt to price fix the NOTE with the backed good and the two are never in parity).   

Since you seem to think in examples, here is the problem illustrated by example:

I don't think in examples, I attempt to explain in examples because economic proofs of the austrian variety are based upon Praxeology. http://en.wikipedia.org/wiki/Praxeology    The deductive study of human action based upon the action axiom.   I think on a far more intuitive level (INTJ) that takes time and effort to boil down into something others can take-in.   

That said, I was just about to make a post requesting that people prove to me how logical humans making decisions in their best interest would result in price variance.  So thank you for your example.  I will now proceed to show you where you assume a human actor will make an illogical (unprofitable) choice and as a result debunk your counter example.

Initial conditions.  1 BitShare = 1 USD,  1000 BitShares mined

Person A creates currency Q, mints 100 Q backs it with (say) 100 BitShares.  So 1 Q is worth 1 USD at this point.

 Now, the dividends he would be getting for the 100 BitShares are now diverted into paying the owners of the Q.  Right now, that is Person A so the system is in equilibrium.  Money is neither gained or lost by minting a new currency because if Q had not been created, Person A would be getting dividends from BitShares.  Other schemes result in runaway currency creation or a draining of the backing (the value) behind Q [2]. 

Now, BitShares like BitCoin are awesome so in 3 months the price goes to 1 BitShare = 10 USD.  Now those 100 Q, backed by 100 BitShares is nominally worth 10 USD each.  So when somebody bids to buy 100 Q for 100 USD (the original dollar-parity price) nobody is willing to sell for that price.  So the currency is created by someone who offers BitShare backing. 

But how many BitShares is the new currency backed with?  For parity with USD, it must be backed with 10 BitShares.[1]  But who decides that value 10? [4]  No voting scheme works because there is no way the holders of currency Q will agree to that backing level, because they lose money.  Regardless, any voting scheme causes people to vote in their self-interest, NOT to vote in a way that causes Q to track USD. 

You are almost right, but have already missed something critical. Your statement is only true for the very first 'minter'.  What is the reason that Q's are minted in the first place?  Is it not because someone, somewhere, wants a Q == USD?  In order to trade a Q for a paper-USD then Q must be equal to a USD.  If Q were not equal, the minter would be unable to sell the minted Q for a paper-USD.  If the minter is unable to sell Q for USD then they are left with only one option, to redeem it and re-mint it at a proper value.  This would involve transaction costs and therefore a loss.

As a result, the person who decides the price at which new Q is minted for USD is the person looking to sell USD and buy Q.   No voting [4], No algorithm, No pegging, No Price Fixing.  An honest price appraisal as a result of an intentional trade between two people.

The second 'minter' can only mint when the BitShare backing(dividends) is worth more than the new Q that would be issued by the buy-sell spread of Q vs BitShares. Thus, the act of printing the second Q increases the interest rate paid to all holders of Q.  The person buying Q with paper-USD would only do so if they had an expectation that everyone else who also wanted a BitShare-Bond with an interest rate that made Q equal to USD would also demand crypto-Currency Q.   I think we can assume that everyone on this board who wants crypto-USD would be able to come to a consensus and agree that 'Q' is the name of the sub-currency that we will ask for.  Anyone else who wanted Q to be gold would lack consensus and thus the price wouldn't match gold.

Within the system, a trade of BitShares for USD simply looks like a BitShare transfer (the USD portion happens outside the network of course -- it could be physical cash).  Even if you add a field to the transaction to "report" the USD transfer amount, there is no proof that the USD actually changed hands [4].  Or even that BitShares changed hands actually.  It would be easy to transfer BitShares to (another account owned by) yourself with a fake USD amount to drive the reported exchange rate where ever you wanted it to go.

There is no need to report anything to the system.  The system does not enforce 'IOUs' except the conversion between BitShares and Q which are entirely within its control.  Reporting of 'prices' is not a valid means of determining price.  Only actual exchanges matter.

I have addressed this before, but I will restate it here again.  In order to 'issue' new currency you must first place a bid to BUY that currency.  If your bid is too high, then it will come from the existing stock.  If your bid is too low, then there would be no takers and issuing against your bid would be doing so at a rate no one else would accept and therefore you would be unable to re-sell your newly minted Q for the price you minted it at.  As a result you only choose to mint Q if you have a buyer (not yourself) already lined up to pay that new price with actual paper-USD.   If you attempt to manipulate the price in the other direction by placing a bid that you hope to issue against.. then your bid will be filled by someone redeeming an existing Q and as a result you would end up covering any existing short position rather than growing your short positions.

[1] Fiat currency pricing information never enters the system.

 
It enters the system from those who hold USD that will only trade their paper-USD for a bitshare-bond denominated as USD with current market value equal to USD.

[2] runaway currency creation

No currency is created nor destroyed and no single actor can profitably issue an arbitrary number of Q

[3] Proven the short term Q's price follows the backing currency.  Point [3] is that there is nothing else for it to follow...

Future buyers wishing to purchase a crypto-currency denominated in USD will demand that the price track or they won't sell.

[4] Bitcoin is based on distrust of all participating parties.  Your system requires some trust or centralization at a minimum to inject fiat pricing into the system, but trust/centralization could also solve [2].

Every exchange made in my system is based upon trading equal value for equal value as judged by both participants in the trade.  Therefore, there is no debt before or after the trade.  All trades are final.  The system is setup such that there is 0 trust and you can assume the most evil individual in the world will be unable to shake the system (manipulate the price) any more than they could manipulate the price in a market of purely physical goods.   

A quote and rebuttal:

Added section to white-paper... (a bit redundant to prior replies, but hopefully useful to people here)

I’m still not convinced that crypto-USD will track the value of paper-USD, can you give me any more reasons to trust the value of crypto-USD?

Every depositor who wishes to trade paper-USD for crypto-USD in order to receive interest will demand that the value of crypto-USD at the time of exchange is about the same as paper-USD.  If not they would not accept the trade.

Within the example given above (where the backing BitShares are now worth 10x), why would buyers demand near parity?  Just because the name is crypto-USD?  In fact, if they did there would be NO SELLERS (at that price).  Crypto-USD owners would just sit there receiving the same dividend payment (denominated in BitShares), but those same BitShares are actually worth 10x USD.  How can the dividend payment be reduced to force parity with USD?  This requires that USD->BitShare price information be available within the system (and its not).

There would be sellers, new issuers! 

I suggest that we turn this this debate into an experiment that anyone can participate in.  I will start a new thread where we will simulate the dynamics of my network and individual actors on this board will attempt to manipulate the price of crypto-USD away from actual USD.  The rules of the game will be such that everyone is attempting to maximize the USD dollar value of their position.  I will come up with a way to manually do the accounting.

Anyone can play the game and if someone devizes a strategy that can 'cheat' the system then they will claim the bounty.  I will use my 10BTC as the initial backing of this 'toy' network.   If you can cheat me out of my backing by following the rules I gave for the network then the bounty will be yours.

Anyone willing to give it a shot?

[Skrapps]

So I’ve read through most of your posts on this. Let me see if I understand this model.

   IRL-X <-> IRL-market <-> Bitcoin <-> BitShares <-> Bit-market <-> crypto-X

   IRL-market - In real life market (fiats, golds, silvers, mangos...). Includes bitcoin, as its established as of this moment
   IRL-X - That which is in existence (paper fiat, gold, silver, mangos)
   Bit-market - Digital exchange of Bitshares and crypto-X

 1)

Shares derive their value from the same sources as Bitcoins.

The native currency will be called a share and is mined into existence on the same schedule as Bitcoin.

Shares will pay dividends from half of the mining fees and rewards

So say 25 BitShares (same as bitcoin) are mined in block 1, by a pool A of 25 people. Each person has 1 BitShare now (which may or may not equal 1 bitcoin, this depends on IRL-market). In block 2, 12.5 Bitshares are distributed to among the block 1 shares, and 12.5 is given to a new pool B. Block 3 etc..
   Pool A = 37.5 BitShares or 1.5 BitShares per person | block 3 gives .25 for total of 1.75 per person

   Pool B = 12.5 BitShares or .5 per person | block 3 gives .25 for total of .75 per person

   Pool C = 12.5 BitShares, .5 each
I excluded mining fees, since they seemed negligible: http://blockchain.info/stats

2)

Users may issue new sub-currencies by ‘shorting the sub currency’ and backing the short position with dividend payments from a defined number of their Shares

I’m going to ignore the ‘shorting’ because I don’t completely understand it in this Bit-market:

So from User 1A has 1.75 BitShares, looks at IRL-market (say bitcoin = $100, and assuming IRL-market values 1 BTC = 1 BitShare), says “I’m going to issue 10 crypto-X with a 10:1 ratio of 1 BitShares, to which their dividends go towards”, and he finds a buyer of 10 crypto-X for 10 IRL-X and IRL-market.

So what does buyer receive? A dividend address and key? But not the BitShares you said, right? So 10 crypto-X that receives 1 BitShares worth of dividends - 12.5/(Total number of BitShares).
So 1 of the buyers 1A crypto-X is worth = (.10*[12.5/(Total number of BitShares)])

Is that correct? If User 1B issues 10 crypto-X at 9:1, then crypto-X-1A is different than crypto-X-1B, right? Are these differences and histories encoded in the blockchain? Isn’t the crypto-X only similar in value to IRL-X at the time of issue and purchase, and that value is then is set in stone, or at least until User 1A buys it back to un-issue it? Wouldn’t you need to constantly be trading to maintain crypto-X’s similar value to IRL-X?

3)
If this is correct, then what is the point of issuing or calling anything crypto-X (USD, Gold, mangos) when the worth is in BitShares, and BitShares are *supposed* to be valued the same as Bitcoins? Crypto-USD and Crypto-Gold only serve psychological functions, in name and language only. If the main factor for acquiring crypto-X is interest, then why not own the BitShares its determined by, which has more use, liquidity?(-is that the right term)? Changing crypto-X to IRL-X won’t emulate IRL value, but will follow the BitShare fractional dividend tied to it and the BTC price.

4)
It seems like this ‘shorting’, ‘interest rate’ and price parity stuff could work, again I don’t understand it completely in application here. I think you made one or two strong connections and associations with that system to this system of peer exchanges and Bitcoin, however I think you got too excited and made some leaps in logic or other implications, and you are trying to hard to smash them together. I can see this ‘shorting’ and ‘interest rates’ influencing a market, price and being the main factors or incentives, but I don’t see why any crypto-X is worth having. Can’t you make this work with bitcoin and BitShare alone (each of those could be exchange for IRL-X, can’t BitShare’s dividend just add a little value to something already existing?)?

5)
IRL-X <-> IRL-market <-> Bitcoin <-> BitShares <-> Bit-market <-> crypto-X

It seems like you want a complete circle/cylce where crypto-X = IRL-X, but crypto-X = BitShare, and Bitshare ~ Bitcoin; and further Bitcoin ~ IRL-market and IRL-X. It seems such a circle already exists with IRL-X <-> IRL-market <-> Bitcoin (though each step has unique bottlenecks) and Bitshares can fit into this cycle or at least complement it, but  Bit-market <-> crypto-X seems to be an appendage — I don’t see it easily, openly flowing into any other aspects, the use function and travel between seems rigid and complicated for whatever end you are trying to accomplish (forcing crypto-X into existence? easier IRL exchanges of currencies and BTC? digital fluidity?). Unless you somehow peg crypto-X to IRL-X — I have no clue how, making the blockchain obey/enforce IRL prices with outside sources, data? But then that makes it a decentralized exchange with a centralized or regulated factor of price determining it; Bit-market would then be subservient to IRL-market. Also, is this ‘shorting’ occurring completely inside the Bit-market, or is the User/Issuer bouncing between IRL- and Bit-markets?

6)
Too, it doesn’t seem like the power is balanced. When block 4 is mined, Bitshares from block 1 have almost doubled (~1.91 BitShares) by their dividends and represent  ~48% of the BitShare total, and BitShares from block 4 are 0.5 BitShares and only 12% of the BitShare Total. (Assuming the 25 users per pool from above, with each new block being earned by a new pool of new users). How are the Bitshares from later blocks suppose to compete with the earlier blocks?

7)
Finally, is this suppose to be accessible to the average person and user, or just to hardcore financial users? Bitcoin is hard enough to make accessible. If your user aim is the general user, then why rush this project? Early implementation could harm the overall idea and drain your money getting it to work.

I totally acknowledge that I might be wrong and that this whole thing may be over my head, but I technically count as an average user, and it the average user is your aim then perhaps this will help both of us. Overall, I think you should be a little more careful with explaining this concept. You seem to jump between the technical workings and aspects of the system, to multiple economic schools of thought, to psychological reasonings determining behaviour of users, and in between.

[thezerg]

Ok, here's my attempt to collect the 10 BTC bounty.  I am going to flesh out out a large issue in your proposal that I mentioned in my first post and hopefully make you see the problem.  Even if not you, it will make everyone else see it.  Or, if we engage in "design-by-answering-questions" I ask you to consider paying me some small portion of your development bounty.  Because good design is a the most important part of development.  I will publicly acknowledge any receipt of funds so others know you are serious with your bounties. 1CKeoT8vBDQDEpMHz5VAswV39pZJ2GTGYd

I appreciate your honest attempt with this post.   If I end up tweaking my design in any significant manner from what I have already discussed as a result of 'bugs' you find in it then I will award a tip proportional to the size of the change.  I have already paid 1 bounty (0.5 BTC) for a logo (though he hasn't yet publicly posted it, I have asked him to). 

Are you willing to back your claims that 'everyone else will see it' with any wager?  It would make this even more fun! 

This is not about a wager.  This is about me putting a h*ll of a lot of time and effort into understanding your system which, frankly nobody is understanding and everybody seems to be giving up on because it is either not clear or not workable.

The issue is fundamentally "There is no way for USD pricing information to honestly enter the system."  Also, there are several smaller-but-still-critical-issues which I'll designate with [N].  So we'll call the above issue [1].  And in the end you'll see that you end up with a currency that tracks some multiple of BitShares, not USD.

Getting honest price information is actually a huge and important requirement. 

Yes, a fact that you have completely failed to address until I brought it up.  If you are going to design-by-question-and-answer please start admitting it.

My definition of honest-price information means it cannot be 'calculated' and must always be the result of an intentional trade initiated by a human actor. 

You have not previously stated anything about price information.  That is my point.  Fiat currency price information is not entering the system.

All trades imply both parties 'agree' on pricing information and they will only agree if they compare it to everything else in the world they could trade for.  If you can show me a place in my software where I attempt to use an algorithm to 'fix a price' then I will give you a 1 BTC bounty. 

There IS no trade from the point of view of the system.  There is only "I give you 100 Q".  The fact that you gave me 5 twenties from your pocket for them cannot be encoded honestly into the system.

I firmly believe that no systems based upon price-fixing (averaging, or otherwise) will work.  (This is why IOU based systems don't in the end... they ultimately attempt to price fix the NOTE with the backed good and the two are never in parity).   

Since you seem to think in examples, here is the problem illustrated by example:

I don't think in examples, I attempt to explain in examples because economic proofs of the austrian variety are based upon Praxeology. http://en.wikipedia.org/wiki/Praxeology    The deductive study of human action based upon the action axiom.   I think on a far more intuitive level (INTJ) that takes time and effort to boil down into something others can take-in.   

That said, I was just about to make a post requesting that people prove to me how logical humans making decisions in their best interest would result in price variance.  So thank you for your example.  I will now proceed to show you where you assume a human actor will make an illogical (unprofitable) choice and as a result debunk your counter example.

Initial conditions.  1 BitShare = 1 USD,  1000 BitShares mined

Person A creates currency Q, mints 100 Q backs it with (say) 100 BitShares.  So 1 Q is worth 1 USD at this point.

 Now, the dividends he would be getting for the 100 BitShares are now diverted into paying the owners of the Q.  Right now, that is Person A so the system is in equilibrium.  Money is neither gained or lost by minting a new currency because if Q had not been created, Person A would be getting dividends from BitShares.  Other schemes result in runaway currency creation or a draining of the backing (the value) behind Q [2]. 

Now, BitShares like BitCoin are awesome so in 3 months the price goes to 1 BitShare = 10 USD.  Now those 100 Q, backed by 100 BitShares is nominally worth 10 USD each.  So when somebody bids to buy 100 Q for 100 USD (the original dollar-parity price) nobody is willing to sell for that price.  So the currency is created by someone who offers BitShare backing. 

But how many BitShares is the new currency backed with?  For parity with USD, it must be backed with 10 BitShares.[1]  But who decides that value 10? [4]  No voting scheme works because there is no way the holders of currency Q will agree to that backing level, because they lose money.  Regardless, any voting scheme causes people to vote in their self-interest, NOT to vote in a way that causes Q to track USD. 

You are almost right, but have already missed something critical. Your statement is only true for the very first 'minter'.  What is the reason that Q's are minted in the first place?  Is it not because someone, somewhere, wants a Q == USD?  In order to trade a Q for a paper-USD then Q must be equal to a USD.  If Q were not equal, the minter would be unable to sell the minted Q for a paper-USD.  If the minter is unable to sell Q for USD then they are left with only one option, to redeem it and re-mint it at a proper value.  This would involve transaction costs and therefore a loss.

My example starts with minting at Q==1 USD.  At that point someone wanted Q==1 USD.  But not now.  Now, the holder of Q wants it to diverge from USD and become infinitely valuable so he can buy a private island.   So I missed nothing.  And all subsequent minters want to mint Q for NO backing BitShares at all.  That way the second minter gets the dividends from the backing BitShares put up by the first! 

But if the first minter has any say whether to allow minting, he'll vote NO for any backing less than his original.

As a result, the person who decides the price at which new Q is minted for USD is the person looking to sell USD and buy Q.   No voting [4], No algorithm, No pegging, No Price Fixing.  An honest price appraisal as a result of an intentional trade between two people.

Great.  I'm looking to sell USD and buy Q.  So I'll decide the price and buy 1000000Q for USD .01 in BitShares.  Now there are 1000100 Q backed by 100.01 BitShares.  I get 99.999% of the dividends of 100.01 BitShares and paid .09% of the cost. 

This is not trade between 2 people, this is minting new Q!

The second 'minter' can only mint when the BitShare backing(dividends) is worth more than the new Q that would be issued by the buy-sell spread of Q vs BitShares. Thus, the act of printing the second Q increases the interest rate paid to all holders of Q. 

Ok what you are saying now contradicts or at least limits the above.  So your rule sounds to me like I can't mint unless I offer a little bit more backing then the previous guy (and nobody sells me Q for that).  That is fine, and it means that Q will rise slightly in value relative to BitShares.  Since in my example BitShares has increases 10x relative to USD, Q will now be worth MORE then 10x USD.  So minting cannot bring the price of Q into line with that of USD.

You need to pick one rule or the other.  If you let me mint for less backing, I will print Q to worthlessness.  If you do not let me mint for fewer backing BitShares, Q will never descent to USD parity if BitShares rises relative to USD.  You cannot let me mint for backing USD because USD is not in the system.  You cannot let me mint for a backing of BitShares-equivalent-to-N-USD because BitShare-per-USD pricing information cannot be honestly inputted into the system.

If my description confused you, just note that you said nothing about USD in the above rule.  As I said before, USD pricing information is not entering the system so it CANNOT be part of whatever equation governs Q.  So Q tracks BitShares not USD.

The person buying Q with paper-USD would only do so if they had an expectation that everyone else who also wanted a BitShare-Bond with an interest rate that made Q equal to USD would also demand crypto-Currency Q.   I think we can assume that everyone on this board who wants crypto-USD would be able to come to a consensus and agree that 'Q' is the name of the sub-currency that we will ask for.  Anyone else who wanted Q to be gold would lack consensus and thus the price wouldn't match gold.

Within the system, a trade of BitShares for USD simply looks like a BitShare transfer (the USD portion happens outside the network of course -- it could be physical cash).  Even if you add a field to the transaction to "report" the USD transfer amount, there is no proof that the USD actually changed hands [4].  Or even that BitShares changed hands actually.  It would be easy to transfer BitShares to (another account owned by) yourself with a fake USD amount to drive the reported exchange rate where ever you wanted it to go.

There is no need to report anything to the system.  The system does not enforce 'IOUs' except the conversion between BitShares and Q which are entirely within its control.  Reporting of 'prices' is not a valid means of determining price.  Only actual exchanges matter.

I have addressed this before, but I will restate it here again.  In order to 'issue' new currency you must first place a bid to BUY that currency.  If your bid is too high, then it will come from the existing stock.  If your bid is too low, then there would be no takers and issuing against your bid would be doing so at a rate no one else would accept and therefore you would be unable to re-sell your newly minted Q for the price you minted it at.  As a result you only choose to mint Q if you have a buyer (not yourself) already lined up to pay that new price with actual paper-USD.   

No, I'll bid low and then choose to mint Q at a low price and sit on it for the dividends paid by your minting it at a high price (see example above). 

Now you'll argue that every minting has unique BitShares that back it.  So now we actually have N different currencies that are called the same name.  In other words Q is not "fungible"; it breaks a key property of a currency.  My low-backed Q is not as valuable as your high-backed Q.  Anyway, I'll still choose to mint it at a low price and sell it to some sucker who doesn't know the difference.  This will happen until Q is valueless which is why "fungibility" is an important property.

If you attempt to manipulate the price in the other direction by placing a bid that you hope to issue against.. then your bid will be filled by someone redeeming an existing Q and as a result you would end up covering any existing short position rather than growing your short positions.

[1] Fiat currency pricing information never enters the system.

 
It enters the system from those who hold USD that will only trade their paper-USD for a bitshare-bond denominated as USD with current market value equal to USD.

No it doesn't because the exact price paid in paper-USD cannot be honestly recorded in the transaction.

I suggest that we turn this this debate into an experiment that anyone can participate in.  I will start a new thread where we will simulate the dynamics of my network and individual actors on this board will attempt to manipulate the price of crypto-USD away from actual USD.  The rules of the game will be such that everyone is attempting to maximize the USD dollar value of their position.  I will come up with a way to manually do the accounting.

Start by actually posting how you do the accounting so we can put this to bed once and for all.

Anyone can play the game and if someone devizes a strategy that can 'cheat' the system then they will claim the bounty.  I will use my 10BTC as the initial backing of this 'toy' network.   If you can cheat me out of my backing by following the rules I gave for the network then the bounty will be yours.

Anyone willing to give it a shot?

Sure...

[mmeijeri]

Just another idea to throw into the mix: it is in fact possible to get fiat data into the system by playing with signed data from an oracle, as described in the documentation about scripts. Of course, that creates a dependency on a trusted third party, but thinking about this may give us new ideas.

[greBit]

I suggest that we turn this this debate into an experiment that anyone can participate in.  I will start a new thread where we will simulate the dynamics of my network and individual actors on this board will attempt to manipulate the price of crypto-USD away from actual USD.  The rules of the game will be such that everyone is attempting to maximize the USD dollar value of their position.  I will come up with a way to manually do the accounting.

Anyone can play the game and if someone devizes a strategy that can 'cheat' the system then they will claim the bounty.  I will use my 10BTC as the initial backing of this 'toy' network.   If you can cheat me out of my backing by following the rules I gave for the network then the bounty will be yours.

Anyone willing to give it a shot?

This sounds like fun, im in!

One could even make a pretty rudimentary game, perhaps a simple web-app, to simulate your currency...

[greBit]

You need to pick one rule or the other.  If you let me mint for less backing, I will print Q to worthlessness.  If you do not let me mint for fewer backing BitShares, Q will never descent to USD parity if BitShares rises relative to USD.  You cannot let me mint for backing USD because USD is not in the system.  You cannot let me mint for a backing of BitShares-equivalent-to-N-USD because BitShare-per-USD pricing information cannot be honestly inputted into the system.

If my description confused you, just note that you said nothing about USD in the above rule.  As I said before, USD pricing information is not entering the system so it CANNOT be part of whatever equation governs Q.  So Q tracks BitShares not USD.

Im gonna make an attempt here...

So we have already established that there exists a market for USD/Bitshare which is likely to be fairly volatile. So we have USD/Bitshare information. Everyone has it.

No one will buy Q from you if you back it with a tiny amount of Bitshare. The buyer knows how much Bitshare he wants backing his Q.

If you mint and fail to sell, you lose out due to transaction fees etc - you should have just kept hold of your Bitshares.

EDIT:

So this is how the system motivates minters to mint at a lower backing than already exists, but not too low as there will be no buyer available.

Am I on the right track Mr Bytemaster?

[mmeijeri]

If my description confused you, just note that you said nothing about USD in the above rule.  As I said before, USD pricing information is not entering the system so it CANNOT be part of whatever equation governs Q.  So Q tracks BitShares not USD.

That's what my intuition says too. Mind you, it's not impossible to back a crypto-USD with BTC, but doing it without a trusted third party seems to be problematic, and even with a third party there's no getting away from exchange rate risk.

[bytemaster]

Simulation Game Thread / Rules:

https://bitcointalk.org/index.php?topic=218507.msg2297712#msg2297712

[thezerg]

You need to pick one rule or the other.  If you let me mint for less backing, I will print Q to worthlessness.  If you do not let me mint for fewer backing BitShares, Q will never descent to USD parity if BitShares rises relative to USD.  You cannot let me mint for backing USD because USD is not in the system.  You cannot let me mint for a backing of BitShares-equivalent-to-N-USD because BitShare-per-USD pricing information cannot be honestly inputted into the system.

If my description confused you, just note that you said nothing about USD in the above rule.  As I said before, USD pricing information is not entering the system so it CANNOT be part of whatever equation governs Q.  So Q tracks BitShares not USD.

Im gonna make an attempt here...

So we have already established that there exists a market for USD/Bitshare which is likely to be fairly volatile. So we have USD/Bitshare information. Everyone has it.

No one will buy Q from you if you back it with a tiny amount of Bitshare. The buyer knows how much Bitshare he wants backing his Q.

Ok, is Q "fungible" or not?  That is, if I have 1 Q and you have 1 Q and we trade them did end up with the same value?  When you say "...backing his Q."  You are telling me that Q is not fungible.  If it's not fungible, its not a currency... Its a whole bunch of different currencies called the exact same name.  How many Q does a Mango cost? You can't price it because my Q is different then yours.

Ok so if we assume Q is fungible, then the sum of all the backing BitShares must apply evenly to all Q notes.

If you mint and fail to sell, you lose out due to transaction fees etc - you should have just kept hold of your Bitshares.

So if Q is fungible this is what happens:

You mint 100 Q and back by 100 BitShares.  Then I mint 1000000 Q and back them by .01 BitShares.  Now I don't need to sell.  I just sit on my 1000000Q and pull in the dividends coming from the 100.01 backing BitShares.  Your original 100 Q gets you (your ownership/total Q issued) 1/10001 th of the total dividends.  You get almost no dividends.  My 1000000Q gets me 1000000/1000100 (or almost all) of the dividends.

So Q is not a viable currency if the next guy can mint for less then what the previous guy paid.

However, if BitShares rises relative to USD, then to get the price of Q down relative to BitShares (on par with USD) I must mint for less backing.

Therefore Q is either not a viable currency or cannot track USD when BitShares rise relative to USD.

[bytemaster]

All sub currencies are fungible.  All units of the same currency pay the same number of BitShares per unit of sub-currency.

If User A  mints at 10:1 and User B mints at  5:1 then the resulting dividend will be 15:2.