Over time you can easily do highly leveraged trading of bitasset against bitasset. It will have exactly the same risk normal leveraged forex trading, minus the counterparty risk and plus the systemic risk of BTS crashing. 20 years into the future and with enough traction bitshares could eat the entire forex market.
I think we should make it cheap for anyone to create a new BitAsset where a BitAsset is defined by the tuple (unique ID, collateral asset type, margin call limit). It would also have a description, authority slate, and price feed definition associated with it. Majority approval of the authority slate could change the description, authority slate, or price feed definition. The price feed definition could be defined as either an arithmetic expression (actually only multiplication and division operations allowed) of other existing price feeds in the system, or the median price of the feeds published by the members of a particular price feed group (typically would be the same group of people as the members in the authority slate but don't necessarily need to be). By the default the authority slate could be set to the current 101 active delegates, or it can be assigned to any particular group of up to 101 BTS accounts.
So, for example I should be able to create a new BitAsset instrument where the collateral asset is defined as BitUSD, the margin call limit is set to 105%, I set the authority slate to some group of individuals I trust to maintain this asset class properly, I set the description to "1 Euro", and set the price feed definition to P1 / P2 where P1 = the existing BTS/BitEUR price feed published by the delegates and P2 = the existing BTS/BitUSD price feed published by the delegates. This creates a new fungible BitAsset instrument that allows for speculating on the price of the Euro relative to USD with up to 20x leverage (realistically less than that to avoid easily triggering margin calls). There is no additional price feed burden that doesn't already exist. People holding this asset understand it is highly leveraged and it might more easily become undercollateralized, and they accept that risk. Furthermore, they understand that the authority could potentially come to a consensus to change the price definition at any time. I also have ideas of how a quorum of the shorts and longs can come to a consensus to replace the authority slate.
When one of the parties involved in this leveraged trade goes under as a result of a sharp move, and has insufficient collateral to back their position, how will this situation get handled? The extra loss has to go somewhere.
I also have ideas on how undercollateralization should be handled for BitAssets:
The worst case scenario for BitAssets right now is undercollateralization followed by a "bank run" (converting BitUSD to BTS) by everyone other than you, leaving you with the only remaining BitUSD with 0 BTS in collateral to back it and no one else willing to short BitUSD anymore. To fix that problem we need to make some tweaks to the system such as converting just enough of a fraction of everyone's BitUSD into a BitUSD-IOU at the moment undercollateralization would occur to prevent undercollateralization from happening. This additional uncollateralizated BitUSD-IOU liabilities may or may not be later automatically reconverted back into BitUSD as the blockchain earns further revenue from future short interest on BitUSD. But even if that never happens, all BitUSD holders take the same fair percentage cut rather than having the early movers get their full BitUSD value while the slow movers lose all of their BitUSD value.