This idea has been discussed in other threads, but the more I think about it the more critical I think it will be. To describe what I am talking about lets start out with a simple example.
Alice and Bob decide to take opposite bets and one ends up with BitUSD the other is Short BitUSD.
Immediately after this trade occurs Alice decides she wants out of her position, but Bob doesn't.
Alice has to shop around looking for someone else to take her BitUSD position... no takers.
Alice drops the price to 95%, 90%, 50% of the dollar .... still no takers.
The value of the dollar hasn't changed during this time, there is just no BitUSD liquidity. Bob hasn't actually made any money, he is just refusing to give up his position.
So we have a situation where people are looking to exit their BitUSD position and they are willing to pay a fee to do so. If the network knows the price then it is easy to implement this. We simply change the terms of the short "contract".
Bob agrees that Alice has the option to exit her position at $.90 per BitUSD at any time. Bob makes money even though the dollar did not fall against BTSX and Alice is assured some liquidity should she need it.
If we are going to rely on a price feed we can force covering any time the highest offer to buy BitUSD is less than 90% of the feed price.
Does this punish shorts? I don't think it does. I think it supports the peg by adding liquidity without adding any risk to the shorts.
I think this added liquidity should come form which ever shorts are least collateralized. This way the shorts which don't want to be forced into providing liquidity pre-maturely can avoid it by having a large surplus of collateral and thus making the entire network more secure.
Under this system BitUSD is always worth at least $0.90 and the market makers / market will likely drive that to near $0.99- $1.01.