It's true that some people might be tempted to make themselves look "rich" by awarding themselves a whole pile of fictitious IOUs. The fact it that there is nothing in the current OWE'M system that would prevent someone creating a whole series of different identities using different email addresses, and then instructing those different characters to send IOUs to one particular person, making them look incredibly rich.
For example, lets suppose that, in addition to my main account under the name of Simon Thorpe, I create 10 other accounts using gmail accounts that I created myself
- Bob the Builder (firstname.lastname@example.org)
- Barbara the Baker
- Chris the Carpenter
- Colin the Cleaner
- Eric the Electrician
- Fanny the Farmer
- Glenda the Gardener
- Harry the Hairdresser
- Naomi the Nurse
- Terry the Teacher
On the face of this, this could look like a fatal flaw in the system. What is to stop someone generating arbitrarilly large amounts of "wealth" in this way?
However, with a bit of thinking, I can see a way out of the problem. The fact is that it would be fairly trivial to examine the structure of the underlying database to determine that none of the 10 fictitious people in my system has ever themselves received an IOU from anyone else. That's hardly surprising because Bob the Builder can't actually build anything for anyone.
Thus, it would be relatively trivial to flag Simon Thorpe as a dubious bet, because all the "people" who "owe" him money, have never exchanged with anyone except Simon Thorpe. Under those conditions, you would be foolish to accept an IOU from Simon Thorpe, essentially because there would be next to no chance of every generating a loop that would allow the debt to be cancelled. Only if Bob the Builder (or one of the other characters in the plot) actually did some things for other people would it be a reasonable bet to take an IOU from Simon Thorpe.
It would thus simply be a question of using the transaction database to rate members by the number of bilateral transactions in which they have been involved to weed out the real players from risky bets. It's not been implemented yet within OWE'M, but I see no reason why we couldn't do that.
This brings out an interesting feature of an IOU based system relative to the money system that we are all used to. Confidence in members comes not from how positive they are, but rather from whether or not they have been actively participating in the past, and whether they are worthy of trust. That's a major change from our current system where power and influence just follows the people that have accumulated the largest amounts of "money" - irrespective of how they got their wealth. A crook who cheated people to get €10,000 has just the same "wealth" as someone who worked hard to earn the same sum.
Let me know if you think that there is still a major risk that the system could be easy to cheat. I'm trying to make sure that we can make such a system reliable, so your feedback will be very useful.