0.5 BTC Bounty - Creating a Fiat/Bitcoin Exchange without Fiat Deposits

[bytemaster]

Note:  most of what I post on this forum is based upon me being very open and honest about the development process.  I have nothing to hide and often 'brainstorm in public' so that we can generate new ideas.   I am posting bugs, airing incomplete ideas, and seeking investment all at the same time because I want to encourage everyone to join in the process of creating an amazing new system and *not* because I want to get rich off of investors money.  I would be happy to be an early miner on this new chain.

    Any investor who wants concrete terms can negotiate it and the result would be a signed contract between people.   Clearly before any *serious* money poors in there will be a new business started which will handle all of the invested funds.  I have been talking with some legal experts on how to handle this in the safest manner possible.   In the mean time, this is more of a kick-starter idea.

   Join have fun, and earn some pre-mined BitShares.   If you want to accelerate the process donations are welcome!

[aigeezer]

In the mean time, this is more of a kick-starter idea.

I'm not sure if you mean that in the literal kickstarter (dot com) sense, but when I read this thread that is the first thing I thought of - I would guess it could work there.

Good luck with the project, regardless.

[CurbsideProphet]

Here is how it works, you can send money via the banks, dwolla, or any other system.  You could use gold/silver if you like.  The point is the exchange does not depend upon it.  Keeping money in the exchange will actually yield a higher return and because businesses no longer have to worry about 'exchange risk' they will be far more likely to accept it.   The end result is that the existing banking system will become increasingly irrelevant.

I understand how you would send money, I'm just skeptical people would use these methods on any significant scale.  I know I can wire anyone on this forum money, that doesn't mean I'm going to do it.  The issue here is trust, you need to trust someone that the exchange will happen and they won't walk away with your money.  I get that there's no parameters on how the money will be exchanged, but on the flip side, that also means there are no parameters (or guarantees) that you will not get scammed.

As for businesses, wouldn't exchange risk be substituted with counterparty risk?  In this essence it sounds very much like Ripple.  Essentially what you have is an exchange built upon leverage (margin accounts) and trust (off exchange, USD <--> cUSD transfers).

Here is the deal, I can pay someone else to do the work and keep my current job.  I can give the idea away and let the VC pay someone else to do it.  Unfortunately, I am not independently wealthy and have to eat.   So the question is am I paying myself above market rates for the work I am doing?   The answer is no, I would have been my self 50% less than I am currently making.   That said, I believe I have a workable solution with my current employer that will allow me to go part time and thus VCs will no longer have to pay for my living expenses.  I will work 2 days per week at my current job and then the rest of my time dedicated to this.   Hopefully that is better for VCs.

I understand where you're coming from but this is the way all entrepreneurs think.  As someone who is familiar with VC firms, it has been my experience that they tend to stop listening to you once you bring this up because even if you are taking a pay cut, it is your decision to do so, they couldn't care less if it's only a fraction of your old salary.  I'm glad you found a workaround, it was more or less just a bit of advice based on experience.  I commend you for thinking outside the box and being transparent with the community.  IMHO, I wouldn't even mention anything about paying yourself from investors as they're just going to counter with the "sweat equity" adage.

[greBit]

Say two people want to use your exchange: Dave - the cUSD dealer, and Bob the cUSD buyer.

• Dave deposits $150 worth of Bitshares to act as collateral and gets $100 worth of cUSD
• Dave sells the cUSD to Bob for cash and then ceases trading & withdraws the excess $50 worth of Bitshare collateral.
• At this moment Bob's cUSD is backed by $100 worth of Bitshares, since the USD/Bitshare exchange rate has been stable.

But two days later the market crashes and the collateral behind the cUSD is now only worth $10.

Who is responsible for pumping in more collateral when the dollar price of a Bitshare falls?

What have I missed?

[bytemaster]

This system does not replace escrow for exchanges of fiat.  You even need escrow for exchanges of Bitcoin.  

This system replaces the need to use a 'bank' to actually store your money and gives you freedom of movement without the need to worry about exchange rate changes.

[bytemaster]

Say two people want to use your exchange: Dave - the cUSD dealer, and Bob the cUSD buyer.

• Dave deposits $150 worth of Bitshares to act as collateral and gets $100 worth of cUSD
• Dave sells the cUSD to Bob for cash and then ceases trading & withdraws the excess $50 worth of Bitshare collateral.
• At this moment Bob's cUSD is backed by $100 worth of Bitshares, since the USD/Bitshare exchange rate has been stable.

But two days later the market crashes and the collateral behind the cUSD is now only worth $10.

Who is responsible for pumping in more collateral when the dollar price of a Bitshare falls?

What have I missed?

Dave has $150 worth of BS and then short-sells cUSD worth $75 (100% initial margin requirement).  The network will not allow him to spend his $150 worth of BS.   When the exchange rate changes and now his BS is only worth $100 and thus no longer has sufficient backing for $75 cUSD, the exchange will purchase cUSD on his behalf converting his cUSD short position and leaving Dave with BS worth $100.

[bytemaster]

Say two people want to use your exchange: Dave - the cUSD dealer, and Bob the cUSD buyer.

• Dave deposits $150 worth of Bitshares to act as collateral and gets $100 worth of cUSD
• Dave sells the cUSD to Bob for cash and then ceases trading & withdraws the excess $50 worth of Bitshare collateral.
• At this moment Bob's cUSD is backed by $100 worth of Bitshares, since the USD/Bitshare exchange rate has been stable.

But two days later the market crashes and the collateral behind the cUSD is now only worth $10.

Who is responsible for pumping in more collateral when the dollar price of a Bitshare falls?

What have I missed?

Dave has $150 worth of BS and then short-sells cUSD worth $75 (100% initial margin requirement).  The network will not allow him to spend his $150 worth of BS.   When the exchange rate changes and now his BS is only worth $100 and thus no longer has sufficient backing for $75 cUSD, the exchange will purchase cUSD on his behalf converting his cUSD short position and leaving Dave with BS worth $100.

If the market crashes by more than the available margin, then the 'longs' will get to split the collateral.   There is *nothing* I could ever code that would insulate people from market crashes beyond a certain point.   

[bytemaster]

Say two people want to use your exchange: Dave - the cUSD dealer, and Bob the cUSD buyer.

• Dave deposits $150 worth of Bitshares to act as collateral and gets $100 worth of cUSD
• Dave sells the cUSD to Bob for cash and then ceases trading & withdraws the excess $50 worth of Bitshare collateral.
• At this moment Bob's cUSD is backed by $100 worth of Bitshares, since the USD/Bitshare exchange rate has been stable.

But two days later the market crashes and the collateral behind the cUSD is now only worth $10.

Who is responsible for pumping in more collateral when the dollar price of a Bitshare falls?

What have I missed?

Dave has $150 worth of BS and then short-sells cUSD worth $75 (100% initial margin requirement).  The network will not allow him to spend his $150 worth of BS.   When the exchange rate changes and now his BS is only worth $100 and thus no longer has sufficient backing for $75 cUSD, the exchange will purchase cUSD on his behalf converting his cUSD short position and leaving Dave with BS worth $100.

If the market crashes by more than the available margin, then the 'longs' will get to split the collateral.   There is *nothing* I could ever code that would insulate people from market crashes beyond a certain point.   

Lets get one thing straight, what would have to crash in value would be BitShares and not $USD.   Ultimately Long's are still trusting in BitShares they just don't have to worry about the volatility.  If BitShares appreciates like Bitcoin does then there is no problem.   The one way to mitigate this risk of 'crash' is to have early adopters post more collateral.  You could actually have  cUSDx2  cUSDx3 and cUSDx4 depending upon how much collateral the various cUSD had backing them.   The cUSDx4 would pay much higher interest rate and thus command a premium price in exchange for USD compared to cUSDx2 which might trade closer to actual USD if not at a slight 'discount' depending upon whether the interest rate justified the risk or not.   

[bytemaster]

sounds to me like you are now thinking: http://en.wikipedia.org/wiki/Foreign_currency_denominated_account
where the "foreign currency" is dollars, gold, etc. and the actual backing account is BitShares (or BTC if you do it centralized).  

The only way in or out it through the crypto-currency.

Pretty interesting...

I wonder how well the dollar denominated crypto-currency, i.e. crypto-USD will track the USD.  Search these forums for "Goxmoney" -- at times during the bubble people seemed to be willing to pay a premium for "USD" located on Mt Gox as opposed to located somewhere else.  Not so much of a premium that is became an issue though.

Ask your lawyer if there's some loophole where "dollar-denominated" does not count as actual dollars, therefore (in Canada at least) its all "play-dough" crypto-currencies and so not actually transmitting money...

Welcome to the new thread!   I am not sure what that wiki link has to do with with this design.     

Like I stated in my other thread, premiums above / below parity are to be expected based upon the relative demand for deposits and withdraws.  I am fairly certain that the nature of interest-bearing USD deposits will cause there to be a slight premium over actual USD, but perhaps the 'market' will factor that into the exchange rate between cUSD and BS. 

I am looking forward to your attacks on this new design!     

[greBit]

I see that there are collateral requirements for opening a position and that the position needs to be well maintained with Bitshare collateral, otherwise it will be automatically closed.

But since Dave is a dealer of cUSD, and cares little for speculation. He just needs a way to transfer $75 of cUSD to Bob upon receipt of $75 in paper money.

So I don't quite get how transfers of fiat balances will work. Who takes on responsibility for the ongoing maintenance of the cUSD collateral?

If the price of Bitshares halves then we will need double the amount of Bitshare collateral backing up all the cUSD.

I think im not really understanding things!

[bytemaster]

sounds to me like you are now thinking: http://en.wikipedia.org/wiki/Foreign_currency_denominated_account
where the "foreign currency" is dollars, gold, etc. and the actual backing account is BitShares (or BTC if you do it centralized).  

The only way in or out it through the crypto-currency.

Pretty interesting...

Ok, I dug deeper into wikipedia and have concluded that your analogy is very apt and much appreciated.    

There are many ways in or out.   You could only ever deal with $USD and cUSD provided you could trade with people who have cUSD balances.  

[bytemaster]

I see that there are collateral requirements for opening a position and that the position needs to be well maintained with Bitshare collateral, otherwise it will be automatically closed.

But since Dave is a dealer of cUSD, and cares little for speculation. He just needs a way to transfer $75 of cUSD to Bob upon receipt of $75 in paper money.

So I don't quite get how transfers of fiat balances will work. Who takes on responsibility for the ongoing maintenance of the cUSD collateral?

If the price of Bitshares halves then we will need double the amount of Bitshare collateral backing up all the cUSD.

I think im not really understanding things!

If the value of Bitshares falls in half, then all cUSD will bid up in a short-squeeze.   This will causes holders of cUSD to sell for BS and take a profit.  New players will come on the market and short at the short-squeeze high and provide the new collateral.  They will then profit when the price falls back to parity where they may cover their position in order to free up their collateral.

[bytemaster]

If the value of Bitshares falls in half, then all cUSD will bid up in a short-squeeze.   This will causes holders of cUSD to sell for BS and take a profit.  New players will come on the market and short at the short-squeeze high and provide the new collateral.  They will then profit when the price falls back to parity where they may cover their position in order to free up their collateral.

Yes, I am responding to myself... I couldn't help but notice how beautiful and elegant these market forces are!  Because shorts are forced to automatically cover once their margin reaches 50% this short squeeze could result in the price of cUSD being bid UP by up to 50% creating a huge selling opportunity.   Shorts would then anticipate this squeeze and thus provide additional collateral to avoid taking a loss during the squeeze!   Elegant!

[thezerg]

Given that bytemaster has offered a bounty for attacks on his exchange system, I'd like to mention briefly here that he has met his commitments on the 10 BTC bounty that he offered for the prior system.

I haven't had time to really digest this new system, but in glancing over them I exceeded myself and had what I think is actually a constructive idea so I thought I'd share it:

Maybe the system should comprise 2 independent block-chains that none-the-less work together.  First,  a bid/ask (offer) chain with much lower difficulty (blocks found every minute, say), and a slower chain that handles filled transactions and "normal" block-chain transfers.

The advantages are:

1. The complete history of unfilled bid/asks is unimportant.  So the bid/ask chain could be checkpointed every few days and old data deleted.
2. The system would be more useful if bids/asks changed faster then the transactions.

[bytemaster]

I think that any block chain will be too slow for high-speed trades and the two chains would have to be synchronized.   The marginal improvement of having a separate bid/ask chain would still not yield a system with enough performance to justify the cost.

[bytemaster]

So here is a challenge, how can a group of people decide in a decentralized manner what currencies are represented by what crypto-Currency?

Lets start with a simplifying assumption:  the first version of the client only enables support for a dozen well known currencies.  These currencies are
hard-coded into the client and therefore everyone knows what they should represent.

The first users of the system would have to establish precedent by buying and selling USD, Gold, and Silver.  Theses early adopters would be taking
some risks, but these risks could be mitigated by observing the bid/ask spread prior to placing their bid.   For example, the first 'bid' might be extremely low, and
the first ask would be very high.  Eventually the market will negotiate a price because it is in everyones interest to reach a consensus.  Every party
to the exchange "wants" consensus that crypto-USD represents an agreement to pay USD.   No one wants to 'go first', but then again someone, like me,
might just go first because I know the market will eventually converge on an understanding.  I also know that there are out-standing bid/asks that would
allow me to reverse my position at a "known" loss.   

The real question is this:  how would the participants of the system sabotage the consensus building effort?  Do they place a bid out-of-line with the
USD or Gold label?   Do they bid too low and no one accepts?   Do they bid too high only to risk losing money?  They would have to 'bet' the market would
do something other than come to an agreement.   There are only two safe ways to 'play' this market:  assume it will reach consensus *or* assume no one will
try to reach consensus.   

Fortunately, this consensus building effort would probably settle out within the first day or two of operation, if not almost immediately.  Once it has momentum,
every party will be investing according to what they thing everyone else will do.  There is only one *safe* assumption and that is to assume everyone else will bid
according to the actual exchange rate.

[bytemaster]

Here is a basic video that introduces the concept:  http://www.youtube.com/watch?v=-8ZJ3xTDwbI 

The target audience is people who do not even know bitcoin so it includes a minute or introduction to Bitcoin that you may just skip.

More to come.

[phillipsjk]

Why did you abandon your original DS Share thread?

Why are you trying to peg the value of shares to specific currencies? For that, CV Tokens would work much better, assuming they are possible to implement. (Croesus says he was just anthropomorphising the network to avoid confusing the economic side of things.)

Your DS shares have value in themselves if they allow you to "short" the value of bitcoin.

BTW, I don't think any decentralized coloured coin proposals will work: they essentially try to get people to value bit-dust.

To answer your question why nobody can explain to you in simple terms why your ideas won't work: You don't appear to clearly explain your ideas. In an effort to make sure I was understanding your DS share concept correctly, I walked through a contrived example. When the results were not what you expected, you claimed I was mis-understanding the concept and opened an new thread that appeared to tack on CV-Tokens (poorly).

What concerns me is that this latest proposal appears more complex than the last. I hope the extra rules follow logically from a sound design, rather than being bugfixes that may introduced new unexpected behavior.

[bytemaster]

The DS proposal was a bad first step forward and was complex and flawed.  I have suffered from having a hard time explaining my original ideas.

This new idea is simple to explain and is not 'bug fixes' on the old idea... if it were merely 'bug fixes' I would not have paid out 10 BTC to TheZerg.   The new design is based entirely upon the model used by traditional brokerages and simply uses a crypto-currency as backing. 

I am moving forward with a team of people who have come to understand the idea and will attempt to explain it to anyone who will listen. 

I have flushed out many new 'rules' for the block-chain based upon vectors of attack identified by TheZerg in the prior chain.  The economics behind the new system is essentially the same as a brokerage that allows people to 'naked-short' with collateral.    Naked-shorting is generally considered bad in traditional markets because those engaged in the naked shorting have a license to print the USD backing that position and therefore can manipulate the price of gold/silver without limit.   Under BitShares there is significant opportunity cost for holding a short position AND the backing of that position cannot be printed out of thin air.   

[bytemaster]

Investors:

      I have decided to no-longer pre-mine the shares because I believe it will undermine trust.  I have also decided to stop taking direct investment and only seek to work with those who value the idea on its own merits.   Everyone will have equal opportunity to profit from this when it goes live.   

     So, instead I would encourage everyone who wants this to succeed to help find bugs and/or fund bounties and development for its own sake and not for pre-mined shares. 

     Those who have already invested I will personally mine the required shares for you.

      Bounty still exists for bugs in algorithm.