It has been discussed many times before that we should prioritize shorts based upon collateral rather than fee. Agent86 has been visiting me this week in Virginia and we have had some serious debates and have identified a refinement that should drive the peg *AND* liquidity much tighter than we have today.
Lets start with some fundamental principles:
1) the more collateral there is behind BitUSD the less leverage there is in the system (and less profit to the short)
2) price fixing the collateral at 2x is bad because 2x may not be what the market needs
3) charging the shorts a "fee" means they start out under 2x collateral and requires us to place an artificial limit on valid short prices.
4) charging the shorts a "fee" means that market makers cannot operate while maintaining a BTSX bullish stance. To be a market maker today requires that you sell BTSX and hold BitUSD which means that few players are willing to provide much liquidity due to the USD exposure.
Suppose we were to change things up as follows:
1) When you "short" you specify the following: Number of dollars you want to sell, BTSX per dollar you are willing to put in collateral, and the Maximum price ($ per BTSX) you are willing to short.
2) When the market executes, if there is a buyer of BitUSD at the feed price, then the short with the highest collateral per USD is chosen provided the feed is less than the maximum price ($ per BTSX) set by the short.
3) Delegates publish their feeds once per hour on average (but slightly randomized to get even distribution / steady updates)
What this would allow market makers to do is this:
1) Short USD with high collateral at the price feed.
2) Buy USD at .99 or more to cover their position.
3) Cover their shorts with lower-collateral speculators when the price falls (minimizing their downside risk)
Under this market the BTSX BULLS could safely play market maker and the market makers would be competing to offer the narrowest spread *AND* most collateral and thus increasing the backing/value of BitUSD. We need the BULLS to play market maker because BTSX is risky and the bears (those who would rather hold dollars than BTSX) are not likely to see BitUSD as a dollar equivalent. The result is that asking the BEARS to play market maker results is much wider spreads.
Lastly as a result of people competing to be the "market makers" so they could have both "leverage" and "buy/sell spread" BitUSD will end up with far more than 2x collateral on average.
There are two primary states the market can be in:
1) Real Price > Price Feed
2) Real Price < Price Feed
And in each of these cases the market makers / shorts have to make a decision on what to do:
1) Real Price > Price Feed
When this is the case then those with BitUSD will be willing to pay more for BTSX than the price feed will allow the SHORTS to execute, thus the market sets the price without restriction.
2) Real Price < Price Feed
- high collateral shorts and market makers will cancel/move their orders while covering (supporting the peg)
- market markers that are "short" will be looking to cover (buy USD) to minimize losses from their leveraged position supporting the peg.
- speculators who want to short "long-term" with higher leverage get an opportunity to be the most collateralized short. They start out at a "small loss" like they do today, but they help sell to the market makers looking to cover. Thus the speculators support the peg by providing market makers liquidity allowing market makers to operate with narrower spreads.
The result of this whole process is this:
1) market makers who want BTSX exposure with leverage risk and to earn money from the spread are able to operate safely with narrow spreads because speculators who want long-term leverage on BTSX price movements are there to provide market makers liquidity.
2) If market makers can be liquid and earn a profit with narrow spreads then BitUSD will be liquid and there will be high confidence.
The downside of this process:
1) A slightly higher reliance on the price feed. The more accurate/responsive the feed is at all times the better the market will function.
2) High reliance on the feed means that delegates will need to have many safe guards in place against feed manipulation by centralized exchanges and that delegates should not source their data from the same places:
a) price movements more than a certain percent should be *manually confirmed* by the delegates.
b) some delegates should operate at slower rates than other delegates
c) some delegates should be "completely manual" and set the price based upon the "buy walls" or "sell walls" on various exchanges rather than "last feed".
d) it is generally better for the price feed to be "low" than "high" because that will result in the BitUSD longs setting the price rather than the feed.
e) based upon this the feed should be "slow to rise" and "quick to fall"
f) with lagging price feed there will exist arbitrage opportunities for USD longs to maintain the peg by selling at a profit until the shorts (following the feed) can catch up.
g) all of these guidelines need not be coded into the blockchain, but delegates can compete, debate, and publish feeds that produce the most effective price peg.
h) shorts can minimize harm from feed manipulation by having tight limits on their order range.
3) Yield on BitUSD will be slightly lower because overlap from short orders will no longer go into the yield funds, but they still get overlap on Long/Long orders.
All things considered I think this will dramatically help the peg because all of us BULLS can make the market with minimal risk and thus make it a safe place for the bears (BitUSD holders).