A86, this does not explain how Alice gets $0.90, 10 seconds after she purchased the dollar.

There's nothing wrong with large demand for short-selling per se. That just reflects optimism about the near-term value of BTSX. (I don't believe that the demand is in any way due to the belief that BitUSD could be going to zero, not with I3's open commitment to helping the peg to hold.) In a liquid market, huge short selling should simply translate into faster price appreciation--just like large options or futures trades lead to price discovery in equities. So why did the price of BTSX go down despite large short sales? I think there were confounding events--wallet bugs, natural correction after the run-up to launch--that made some people believe that short-selling would not lead to BTSX price appreciation. Also, up to now the market rules may have been a bit too favorable and not punitive enough to short-selling. Why not make a conservative change, like increasing collateral requirements or increasing the margin call penalty, and see then if the market can achieve a good peg on its own?

It's clear that achieving a stable peg would help immensely in increasing confidence and growing the network. We all agree that getting there sooner is better than getting there later. But, to really condition traders' expectations and build confidence for long-term success, we need to prove that the peg can work without a heavy-handed restriction based on price feeds. Market participants need a chance to condition their beliefs and learn. Disallowing short sales below the median feed price would not give them this opportunity. Furthermore, such a stop-gap measure is not really a long-term solution. Even if such a measure enforced the peg for now, it would really be a crutch that would become harder and harder to remove over time. If the crutch seemed to work for now, it would prevent true price discovery, much like the blunt short-sale prohibitions for traditional equities that proved to be ill-advised. When the network size is much bigger in the future, it could be very difficult to remove the crutch without causing widespread concern about the robustness of the peg, simply because the market would not have learned on its own to achieve a viable peg. What if the crutch were dispensed at some point and the peg began to fail again? Confidence could be shattered. I believe it would be a mistake to try to bypass the steady learning and conditioning process that needs to take place for long-term robustness of the peg.

So, I say let the market work mostly on its own for a little while longer, when the costs of failure are relatively low. See if things get better. Fix bugs in the client, make it easy for more people to get on the network and participate in the internal market. Make one or two simple, conservative changes to the market rules that don't rely on price feeds. E.g., just increasing collateral requirements from 2x to 3x, decreasing the margin call price, and/or increasing the margin call penalty to 10% may be enough to restore balance between shorts and longs. If the peg can start to hold better with minimal changes, that would be a home run and far superior than using a heavy-handed approach such as restricting short sales below a median price feed.

Quote from: Agent86 on August 29, 2014, 04:46:12 pm

please just make the change that prevents any short that is below the median price feed from being matched...

Any reason not to make the change?

Because then the bitasset ecosystem is not truly decentralized. Isn't that what bitassets are all about in the first place? Accurately determining the price of an asset in a completely decentralized manner?

Give the system time to mature. Many people cannot even access their funds and thus the market because of software bugs. It is a broken experiment thus far that was setup for failure. Give time for the bugs to be fixed, market depth and liquidity to increase, THEN judge whether the experiment was successful or not.

1) BitUSD vs USD will work with no market restrictions provided it is a mature market with a proven track record.

Given those facts here are some additional things I have realized:
1) There is never any reason for new BitUSD to be created below the market value

Here is what I have implemented as an intermediate step:

1) If delegates publish a feed, then the feed will be used rather than the average.  In this way if there is a service interruption on the feed, then the network can still function like it does today.

2) This puts the "shareholders" in control over whether to use feed or average.

3) I have restricted the creation/execution of new shorts to the median / average price.

Result... those who have USD have priority in selling over those who would like to short.   This adds liquidity for those with USD by removing competition from those looking to short. 

Next we are going to open up a second bitasset so that "short demand" can spread between two BitAssets and the bitassets can trade against one another. 

I believe this should implement your proposal without the arb. bot.   

With the whole market price depending on a feed but not on depth this might allow a few delegates to slightly manipulate the price which will allow someone with large capital to abuse the market. I also think this is not a significant problem for now but it could be exploited.

Quote from: bytemaster on August 28, 2014, 08:03:21 pm

Quote from: MolonLabe on August 28, 2014, 08:01:06 pm

From libertarianism to democracy to oligarchy to tyranny, even faster than I expected.

  that is an interesting point of view... and I see your parallels.

I am for the feed and against the network owned arb bot. Its a free market and this would be a bad step. The whole, is it ok to do bad things for the greater good kind of thing, their has to be another way. Why not just use the feed and let the market decide the value of bitUSD