As I am person who have long-term interests to have BitCoin alive and safe (mainly because I have invested into massive mining operations, and plan to deploy further projects using BitCoin) I propose following to happen:
As you see system with codes - like Mt.Gox codes, CryptoXChange codes, BTC-E codes works rather well, and these actually represents something like promissory note (debt note, i.e. that finally Mt.Gox would pay to owners of these codes) of Mt. Gox for some USD for example. This already works, but it has multiple dangers - first - you ABSOLUTELY HAVE to trust exchange, as in cases when they get in trouble, all of their obligations to pay against holding of USD / BTC will be troubled. This would result in massive panic on market, exchange rate drop, etc. So this approach would definitely produce problems.
So what is proposed:
1) either embed into current block-chain or add supplementary USDchain block-chain.
2) add ability (or check if that is already available with scripts) to bitcoin code that would allow to lock some BTC for SECURED LOAN
of USDcoins - you see - unlike BTC "mining" production, USDcoins produced in the same conventional way - as debt
----- How it works:
Alice has 50 BTC and would like to sell them for some 5.0 USD, Jane has 30 BTC and would like to sell them for 5.2 USD.
So Alice takes 250 USDcoins (please note that is not US dollar note!) for her 50 BTC being LOCKED
and available to anyone to PULL
at specified exchange rate by her. Jane takes 156 USDcoins and 30 BTC is LOCKED
as well. That way 406 USDcoins created from thin air as secured loan against BTC.
Then Alice finds some exchanger that would exchange to paypal for 2%, and sells to that exchanger her USD coins and gets 245 USD on her paypal account (please note that it is again not US dollars, but paypal obligation to have enough reserves...).
Jane in other way finds Bob who want to buy BitCoins and offer him her 156 USDCoins and gets paid with cash.
Bob goes to market and buys 31.2 BitCoins from Alice, liquidating 156 USDCoins in circulation. Market is p2p as well - cross-chain trading!.
So now we have on market 250 USDcoins and 48.8 BTC available.
Then John buys 250 USDcoins from exchanger for his 255 USD available on dwolla and buys the rest 48.8 BTC available on market.
Main problem is inflation of USDcoins, as if such basic schema as written above implemented. Nothing stops Rob to go into market and put 50 BTC for 1'000'000 USDcoins . This should be LIMITED and it is the most tricky part of this play,
as this limit must be well understood math and research.
Solution could be to implement limitation for such operations - say first - Alice sells 50 BTC for 5.0, but actually get not 250 USDcoins, but only 125 USDcoins, and could get another 125 USDcoins, when her previous 125 USDcoins were destroyed in market. This actually requires research and financial experiments.
Second part - 50 BTC for 1'000'000 USDcoins - actually secure loan algorithm should prevent this kind of deal totally.
As you see - USDcoins would be as secure, as would be smarter algorithm that controls emission of them. If emission is too high, then USDcoins worth would drop against US dollar. If emission is too low, then it would be pain to trade on this market, and USDcoins would worth more than true US dollar. So multiple market parameters should be taken into account, and USDcoins could stick then to actual USD supply-demand for BitCoin transaction.
Long term problem - it may happen that transaction rates in BTC would drop, while in USDcoin increase, as some people would prefer to stick more with USD equivalent of worth. Then - overall worth of BTCs that were put to secure USDcoin emission would decrease, and USDcoin would become unsecured bubble.
Solution to this problem could be seen for example as adding demurrage on USDcoin, so if you HOLD USDcoin on your account, value of it decreases 0.1% - 0.2% per single day. This could be also payment to miners.http://en.wikipedia.org/wiki/Demurrage_%28currency%29
Adding demurrage rule would prevent of using such USDcoin for asset accumulation purposes. But again - building correct model, taking in account actual economic behavior of people is not easy task here. Looking forward for some help.
Mining. If BitCoin could be mined, USDcoin cannot be mined in the same way. So probably only fees like 0.2% per transaction could fund mining, and that would be pennies today.
------ Solutions (roadmap):
1) Find out algorithm that can calculate correctly secured loan value;
2) Implement necessary additions to bitcoin protocol, and deploy some prototype on servers in non-p2p way, but more like carthel with some exchanges;
3) Invite users into this experiment;
4) If everything goes good - deploy it for wider usage as p2p software, possibly integrating into existing BitCoin software;