Topic: *old* BitShare Economic Theory 10 BTC bounty to prove me wrong... paid. - page 7. (Read 10089 times)

The mechanics of what you're trying to do are completely over my head, but Austrian economics is most definitely not.

For this comment alone, I'm watching this carefully. (and hopefully I'll learn something useful to both of us along the way!)

When I was a kid, we called it "saving", and those who did it did better at life.

I already mentioned this problem some time ago ... but apparently the rules are such that this is not possible because minting can only be done when no-one else is trying to sell Q.  Thus if we have a market and Q has any value at all (some bid offers) basically never.  This is what's my current problem with the proposed system.  And yes, I also do agree that I *still* haven't really understood all the rules; also the whitepaper doesn't clearly define every detail as I think a proper definition should.  (But maybe as a mathematician I have different assumptions on what a "definition" should be than other people.)

dacoinminster:   thanks for your encouragement. 

I would recommend you reconsider your 2nd requirement: not have accurate 'prices' via averaging, voting, or any other means of calculating them.  All prices must be derived from an intentional decision of two people to trade.   If you did have a 'ticker' someplace, then that ticker would have to be sourced from somewhere based on real trades (not reported trades).   Unless there is an active 'auction to the highest bidder' then prices can be easily manipulated.  Therefore, I do not see how a decentralized trading system could depend upon a centralized datasource producing real, verifiable trades that occur at auction.

Someone earlier mentioned that something that could build on top of bitcoin would 'outstrip' other currencies.  Presumably because of the network-effects of bitcoin and brand-loyalty. 

If I wanted to build on top of bitcoin without breaking backward "compatibility" it would be almost impossible.  You would need to add new types of scripts that older clients do not support.    You would end up needing to post 'colateral' in bitcoin and that collateral may be insufficient.  You would end up with 'escrow' services and 'counter parties'.    The closest thing I could think of would be something along the lines of my version 1.0 proposal:

1) someone could post  2x the bitcoin collateral to issue a 'colored' coin.  A certain amount of that collateral would be paid out to the holder of the colored coin over time.   The problem is that you are price-fixing both the margin requirements and the term length.  If you don't standardize on those two things then the colored coins wouldn't be fungible.

I really think that there is nothing wrong with creating a new chain as a means of moving forward, especially if the API was 99% compatible.  Value is easily transferred from one chain to the other and you would have just as smooth an upgrade path.   Sure the miners may not like it if they have a lot invested that isn't transferrable, but that is just the nature of markets.   

I am going to make another attempt at proof that does not rely upon 'examples'.  

The first thing I will do is clear up any idea of 'backing' or 'collateral' and instead we will do some comparisons based solely on revenue streams.

Given two revenue streams, one that pays 2 BTC / year and one that pays 4 BTC / year, how much more valuable is one stream than another? Regardless of the price volatility of BTC one revenue stream will always be twice as valuable as the other revenue stream.

When you combine 2 bitshares to create 1 crypto-Q (cQ) you are really creating a new revenue stream cQ that will always be worth 2x the value of 1 BS.

If you combine another 4 bitshares to create a second cQ the value of all cQ will now equal 3 BS.  

Now all I need to do is insure that the variance in the revenue stream of crypto-Q vs BS tracks the variance in the real world price between BS and Q.

So I create a '0-sum' game between BS and Q.   To create crypto-Q you must give up current market value of Q in BS.   To buy Q with crypto-Q requires they have the same value.  Therefore, as long everyone is buying crypto-Q with Q (and no one is selling crypto-Q), the price of crypto-Q must, by necessity, follow the price of Q up or down. If it didn't the creators of new crypto-Q would be unable to buy Q with it or they would be paying more than they need to.

Now what happens if we reverse the process and only have sellers of crypto-Q.  Who are the buyers?  The only buyers at this point would be the 'shorts' because there is no one else in the market.  How much would the shorts be willing to pay?  They would be willing to pay anything below their creation costs and still make a profit. The more below the creation costs the greater the profit.     But the shorts are competing against one another for this profit opportunity and who ever was willing to accept the least profit would win.  So, how does a short buy crypto-Q?  They must provide the seller of crypto-Q with Q.   So all of the shorts compete to provide Q to the seller of crypto-Q at the highest price possible (lowest profit possible).  

The key to making the above relationship between shorts-and-longs track the actual market value is to ensure that all of the shorts maintain a cost basis as close to the current value of Q as possible.    This means we want a market force that causes someone who is short crytpo-Q at 10 BS to close out their position long before crypto-Q falls to 5 BS.  If the short didn't close out their cost basis at 5 BS then in theory the last seller of crytpo-Q would only get 50% of face value and that would be a failure of the system.    

Fortunately, there is another market force at work.  The individual short of Crypto-Q at 10 BS is paying is paying 2x as much interest to the holders of crypto-Q as the shorts with a 5 BS cost basis.  Therefore, the short who was short at 10 BS will cover his position and potentially go short again closer to 5BS and in the process make a profit by freeing 5 of his 10 BS from being mortgaged.

What drives price parity while the value of Q is rising and the value of BS is falling is the demand of depositors to receive equal value.
What drives price parity while the value of Q is falling and the value of BS is rising is shorts competing to buy BS (which is appreciating) at a discount.

I think I have given many, many examples of how the price stabilizes.  I would like to see one example that causes the price to not follow parity assuming:

1) No buyer accepts crypto-Q in exchange for Q unless it has equal value.
2) No one can short new crypto-Q so long as there is someone with existing crypto-Q willing to sell at that price.
3) There is a large market for crypto-Q with many competing shorts, buyers, and sellers.
4) All parties are profit seeking / maximizing.  

Lets construct one trade sequence at a time... and demonstrate the instability.  So far all I see is a lot of handwaving and straw-man that violate the rules of the blockchain.  Lets stick to the rules.  Propose a sequence of transactions that you would execute and if they conform to the rules of the blockchain and yet result in easy manipulation by a single party then it clearly will not work.  

The worst-conceivable-outcome that I can contrive involves the complete disappearance of all depositors seeking crypto-Q paying interest.   In this event, the value of crypto-Q gradually falls until the very last crypto-Q reaches parity with the short-issuer with the lowest basis.   But this outcome is so out-there that I cannot conceive of a single reason why all of a sudden no one would care about getting a high-yield crypto-Q revenue?   I need to see what would drive all depositors away when they could get crypto-Q at a discount?    If you told me I could get $100 of bank credit for a $90 deposit, I would jump on it.  As the price fell further more and more people would jump on it until the price reached parity.  
 



  

Points:
1. The strategy detailed above to wipe the value of the crypto-USD is ironclad.  
2. And so it the point that nobody would trust a currency where if the bottom bid got filled it would zero the currency value.
3. And the observation that paper USD price information never enters the system so therefore no currency can track it has been simply ignored by you.

You know I originally started looking into this to possibly help implement it.  But continued posting instead of just disappearing to save you your $20000 and earn the bounty just like you said here:


This is exactly what I have done.  I have shown how the system as originally formulated would fail.   I have spent significant time thinking and writing about BitShares and seemingly convinced everybody but you about the faults with the BitShares system.  You are clearly no longer ready to pay 20k for an implementation so you changed the title of your thread from: "Help Wanted $20,000+ Job creating Distributed Blockchain-based Exchange" to "BitShare Economic Theory and 10 BTC Bounty to prove it wrong."  So I guess I have essentially "convinced you to drop your current idea".  Now you have started simulations to see the failure modes I described occur and figure out some fix.  But even if you find a fix, I still fulfilled the bounty.

You are no spring chicken here.  You should know that posting a 10BTC bounty for one thing and then transmuting that offer (after I posted a winning strategy) into paying the winner of a simulation is going to get you a scammer tag.  And double-shame on you since as a 2010 registrant those 10 BTC may have cost you almost nothing.

But at the same time I think your heart is in the right place.  Like Mathew Wright, you just don't understand the value of other people's time.  That is why, as you yourself admitted, nobody was paying attention to your idea until you offered the bounty/ implementation.

So prove you value other's time and that we can take you at your word.



EDIT: thread readers help me out here... did I convince you the crypto-USD could not track USD and that it is in the a person's interest to devalue it by minting millions of coins?

This is extremely interesting. Somebody is going to figure out the best way to get a stabilized decentralized currency which tracks external values like USD and gold, and that person and their investors will be very, very rich.

bytemaster, I also wrote a whitepaper with another way to do this (see my signature). I think you might find it interesting. If you see anything there which you like, feel free to steal it.

You are the first person I have seen offering to collect bitcoins now to fund future development of a cool new protocol like a kickstarter project, with backers getting a cut of the new protocol. I wish more people would do that - its a great way to invest in innovation!

As it stands, I would probably be a small investor, just to see where this goes. Like many others here, I am skeptical that your price discovery model will work, but I don't see many opportunities to invest in this space, so I am hopeful that if your idea doesn't work, you will evolve it into something which does work.

If you wanted me (and perhaps others) to be a major investor, your proposal would have to be amended as follows:

• Build on top of bitcoin as a new protocol layer rather than starting your own alt-chain
• Do price discovery by having multiple data sources who publish a "ticker" into the block-chain. Crypto-USD holders then vote on which ticker (or group of tickers) they are using

I look forward to seeing what comes of this.

The Zerg,  I truly appreciate your efforts here, but I am not going to pay the bounty AND continue to invest time and money into this idea.   If and when you can prove that I have abandoned this idea and I have not explained my reason for abandoning the idea nor awarded the bounty to anyone, then you can call me a scammer.

You have not understood my idea because what you keep saying back is not what I am actually suggesting.  This is why I am offering the game.   However, I do not want this discussion to turn into a back and forth debate about my integrity.  If you think I am a scammer, then just sit back and watch the game play out and don't invest any more time.  Otherwise, lets have some fun and discover something great! 

I am putting a lot of time and effort (and money) into teaching the details of my idea.  It is very hard to do in writing, so if you would like to talk over skype then I would welcome that.

Yes, but now we are discussing whether I can destroy the currency.  I think I have proven in my previous postings that it will not track USD which is the original issue.

Additionally, to repeat myself, nobody will trust a system that if ever there are no reasonable bids, my unreasonable bid that renders the currency worthless gets minted.  Every 5 minutes 24x7 I have a chance to "win". 

If the currency got any interest at all, what would happen is a type of ponzi scheme -- not a rising one, but an exponentially increasing death.  This would happen because people might bid higher then my "infinite" bid, but lower then the original minter's backing.  The highest of those would get minted if the original minter does not sell at a loss.  Or if the original minter sells at a loss, the value of the currency is reduced.  Rinse and repeat and you have a slowly descending currency value.  But at some point all holders would realize what's going on and sell at any price, wiping all the bids from the books and then my "infinite" bid would get filled.

This is why Bitcoin has a limited number of coins.

If you limit the number of coins, OR limit the velocity of coin creation, you would possibly have a viable currency, but then you cannot track USD because you cannot print enough to match helicopter Ben!!! :-)  And that STILL assumes you have an "oracle" that knows how much should be printed...


I am sorry if you now regret placing the bounty & I do think that your idea of a dividend is quite interesting and creative.  But when I examined it in depth and for many hours it turns out there are unresolvable issues.   

You offered to pay for a critical examination and now you have gotten one.  Its time to be the man and pay up or duck and take your scammer tag.   I have worked hard to understand your system.

But life is not entirely over.  It may be possible to add dividend payments to help "debounce" a USD tracking currency.  But it is fundamentally (at a minimum) going to require a trusted source -- an "oracle" as people call it that reports USD to BitShare exchange prices, and minting/price limits based this oracle.  If done properly, this would be "better" than a crypto-currency backed by $ in a vault because if the oracle is shut off, at least the currency retains its last value and simply diverges from USD.  Whereas a vault backed currency means you could open the vault and see nothing inside.

Let me clarify one more time what the purpose of this bounty was:  I wanted to discover the small little detail that I missed that is 'unresolvable'.   I placed the bounty because I believed that any 'details' that remained could easily be 'resolvable' and result in a system that still produced about the same result. 

This is supposed to be a fun, exploratory process for everyone involved.  The reward for figuring out how to build a system with the properties I described are huge for all of us.   If you want "all of the details now" the best way I know to explore this is the game.

I truly apologize for sounding that way, but I started this thread with an inspiration that I felt could work and I had the major parts figured out.  Minor details are being filled in as we go.   

While minting can only occur once per-currency every block, trading among existing crypto-USD holders can occur with as many transactions as will fit in a block.  Furthermore, I have stated repeatedly that the blockchain trading is designed for low-frequency trading and high-frequency trading would have to occur 'off-chain'.    If I need to tweak the block-chain rules to use 5 minute blocks instead of 10 minute blocks to improve this a bit, then I do not think that would change the fundamentals of the system. 

When matching a accept transaction to a bid transaction, the transaction that pays the highest transaction fee wins the 'race'. 

I will create a market thread for the game, that is a good idea.

I am open to suggestions on how we can establish the changes in the exchange rate.  After each transaction is 'committed' you get to pick a direction +5% or -5%.    However, you only get to pick the direction again after 5 other players have had an opportunity to pick the direction.   Therefore, you all can conspire to work together or defect and attempt to move it in different directions.     My only concern is that if we pick something that doesn't accurately reflect real market behavior then we may need to refine this aspect of the experiment.

You sound like you are making this up as you go along.  The bandwidth limitation of 1 "minting" per block will render the currency useless for nontrivial economic activity.  That would be like only one mortgage can be closed per day (or every 10 minutes, whatever it does not scale) across the entire USA.

Regardless, I keep my bid there until I can fill it.  It costs me nothing.  So people who want to keep the price of crypto-USD high have to keep issuing "fake" bids.  "Fake" meaning they don't want the crypto-USD but have to bid for it to lock me out.  If they forget to do so once, my bid is filled with minted coins so the currency becomes worthless.  Nobody would hold money in a system that is one filled bid away from destruction like that.

And even if they never forget to issue a fake bid, these actions will actually slowly mint new crypto-USD currency, thus devaluing it.  Ergo, it STILL is not tracking USD.

Correct, you can open the bid and issue from the same account *but* there is a 1 block delay between you placing your bid and it being filled.

Secondly, only the highest bid can be filled.  You can *not* arbitrary decide to fill a lower bid.  The highest bid can be derived from the block-chain history.

I am offering the game as a means of proving to everyone definitively whether it works AND to make wining this bounty less 'subjective' and instead based upon a somewhat experimental result.    Clearly, you all are having a hard time convincing me and I am having a hard time convincing you... therefore debating is actually not helping either of us reach the desired goal (winning the bounty or winning support for the idea).   So in all fairness, I am not changing the rules (the rule is you have to convince me), but instead offering you a fool-proof way of convincing me.  Thus entirely in the spirit of this bounty.

I open the bid of course (I've said this several times).  If you don't allow me to open the bid and mint the coin from the same account,  I simply do it from 2 separate accounts that I control.  So if nobody fills my bid, I do so from my other account.  If someone else fills my bid even better.  Either way crypto-USD is devalued to 0.

This violates the rules of the chain.  While you can choose to issue shares at what ever exchange rate /quantity you like, you can only do so in response to an open bid confirmed in the prior block.

Ok.  The primary problem we have here is me properly communicating what the rules of the new blockchain are.  The purpose of the game is to teach you all the rules of the chain.  As we play the game and walk through each transaction you will discover what you can and cannot do.   This isn't me changing the rules, it is me revealing them to you. 

This should be acceptable to everyone because I am the one risking real money, while you are the student attempting to understand. 

The only requirement of my 'rules' is that they can be defined / implemented in a block chain and do not depend on arbitrary 'outside' information.

If in the process of playing the game the rules are revealed to work, then we all win and we can get on with building Bitshares.   If in the process of playing the game I am unable to define a set of 'instructions' that the blockchain/computer could follow and maintain near parity then you will win.

If you are concerned about 'design-by-question' and want 'paid' for helping to design a working system then I cannot help you out.  If I am wasting your time discussing an unworkable system then you will win the bounty.  If by working together we can refine my ideas into something that is workable then we all win and you can get paid for your time by the existence of a revolutionary new system that you can be an early adopter of.  Win/Win for you and me.   

 

Exactly. QED.  bytemaster, please send the bounty to: 1CKeoT8vBDQDEpMHz5VAswV39pZJ2GTGYd.  Remember, I'm saving you 20000 bucks.

Edit: But if I have to sign me up for the game.  I'll take every currency anyone creates, instantly mint 100000000000000 units (or whatever the maximum currency units are) for 1 bitshare (or whatever the minimum is), therefore reducing the currency's value to essentially 0.  These currencies will never track USD.   And as a bonus, if I'm the only one doing the above, I'll get all the dividends.

I'd like to play this game. Shouldn't we start an 'Market' thread to keep this 'Theory' thread and the 'Blockchain' neat?


Agree with this; the 10BTC bounty should stay here. Its still ultimately up to OP if he isn't convinced.

Bytemaster: RE: Simulation :

1. "minting" currency is not defined in your game rules.

2. In fact again the actual rules of the BitShare system is not defined.  Your "rules" just define how we issue buy sell orders.

3. Choosing Mt. Gox exchange rate is just going to delay and obfuscate the situation.  Its theoretically (and practically when you look at crypto currencies) possible for the exchange rate to vary by 90% in weeks.  But I don't want to play this game for an entire year just to prove a point.  And just to remind you, the point was NOT to make money, but to prove that crypto-USD does not track USD relative to BitShares.

Please respond to my prior posting.  You have offered the 10BTC bounty.  Now you are changing the rules AGAIN, with this endless game.