There's obviously a lot of detail here. I've read some of it but I'm still trying to understand it. The FAQ is missing some obvious high-level questions for newcomers :
- How do I transfer my real USD for digital USD? Send it to an issuer? How do I trust the issuer?
Yes. For any contract-based currency, including commodity-based, you must trust the issuer that he is actually upholding his end of the contract, i.e. holding the appropriate assets in a legally-acceptable fashion.
Therefore you must trust his governance: his auditing procedures, his standing in his local jurisdiction, his use of separation of powers, his bonding (eCache uses bonds)...
Those are all governance
issues (meaning there is no technical
solution, since these are issues of governance.)This is one area where I think Bitcoin is actually superior to physical commodities (such as gold), since technical solutions are possible with Bitcoin.
- If an issuer's assets are stolen, or the issuer is bad actor, the reserves will less than stated, or zero. Presumably everyone who owns digital versions of those assets will see their value wiped out. In effect, owners of a digital USD currency take counterparty risk to the USD issuer that they use?
This is absolutely true. If the issuer himself is a bad actor, then his digital currencies are suspect. My goal with OT is to eliminate this risk at a technical
level (the transaction server), but obviously a piece of software can never fix governance
issues such as an issuer's need to audit and insure his warehouse full of gold.
The issuer is in the business of trust; my hope with OT is that jurisdictional arbitrage
will enable a global trust market, with all of the liabilities and risks of the transaction server itself removed from the equation (by OT).
- If a Bad Event happens (eg. MtGoxIssuer is compromised), is there a way to alter the transaction chain similar to "a 51% of Bitcoin Miners" vote?
1) The OT transaction server itself cannot forge transactions or change balances. So the "bad event" that happened, will not happen on OT. (OT cannot forge receipts.)
2) The OT transaction server itself technically could store all of its receipts, and thus roll back the system to any point based on the information in those receipts. (In practice the OT transaction server doesn't need
to store any receipts except for the last one
for each user.)
3) While the OT transaction server cannot forge receipts, it COULD use an illicit account to counterfeit (inflate) funds, and could do so without having to forge any issuer receipts. This is prevented using an audit protocol between the issuer and the transaction server. (Which could be performed daily.) OT normally allows parties to destroy all receipts except for the last one. However with an audit protocol in place, I recommend that parties store all receipts between audits.
4) This way, if a transaction server ever counterfeits, the issuer will catch it out, and yes, would have the ability to roll everyone back to the last audit, and re-issue their funds on a new transaction server.
This is described above in my long post. The transaction server has little incentive to pull this, since he wouldn't get away with it, he doesn't hold any of the gold, and he would immediately lose his daily revenues from transaction fees.
An even better solution is possible using Bitcoin (coming soon) which I have hinted at recently. Using Bitcoin, unlike any other commodity, it's possible to accomplish these things without even having to trust an issuer. (No one is doing it yet, but they will soon.)
Assuming I understand this so far :
- Would you intend to counteract the requirement to an issuer by diversification? ie. Have 1000 digital USD issuers, and trade a basket of those issued currencies?
This is one intention of mine, yes.
It is possible in OT, using baskets, to distribute the risk of a single currency across multiple issuers.
If one issuer ever disappears, the loss could then be made up through insurance, paid for through transaction fees from using the basket.
I'm sure there are other creative solutions but that's one I had in mind.
- If I set myself up as a USD issuer, my reliability would be low. It seem logical that my Isosceles USD would be worth less than 1 USD issued by MegaBank Issuer, right? So I wouldn't bother, I'd just send my funds to MegaBank. This reduces the pool for diversification. What incentive is there to be a large number of digital USD issuers?
I agree that funds are likely to flow more to trusted parties than to untrusted parties.
Due to this, it creates a profit incentive to innovate in trust solutions, which I believe entities will do (and some will succeed, some will fail.) That is why there are markets.
There are also other incentives for users to choose one issuer over another, such as bailment policies, jurisdiction, etc.
- CDOs are extremely complicated things, not suitable for retail investors. Why do you have them on your To Do list?
Because OT is, first and foremost, a financial crypto library. The OT server and OT API are just examples of what can be built using that library.
It makes sense to me that such a library should be technically capable of implementing various instruments, even if they are "complicated" or "not suitable for retail investors".
The thing with a library is, you don't know all the ways that people might end up using it.
OT's infrastructure is such that, while CDOs may seem complicated, they are rather easy to implement using the same mechanisms used in basket currencies.
I don't know when I will actually get around to CDOs, and perhaps they will be implementable using script clauses, I don't know. I tend to prioritize so I doubt you will see CDOS before stocks, for example.