With the bitAssets 2.0 proposal, forced settlement at the feed price means the blockchain essentially becomes a market maker, because it will need to be able to exchange USD for BTS when the existing book is insufficient to meet demand.

This simple exchange process exposes the blockchain to inventory risk. Inventory risk occurs when a market maker holds more of one currency than it does of the other - the risk is that the currency it holds the greater quantity of could fall in price leading to further risk and potential cascade of losses.

A market maker is risk neutral if they hold equal value on both sides of the market. Then, they are essentially in a hedged position. This is exactly equivalent to an equal match of bitUSD longs and shorts.

Whenever the blockchain is forced to settle at the feed, it must be able to respond to the change in demand by altering the price it is able to settle at. However, in the current design, this will not happen instantly, it relies on the feed price changing. This will only happen after arbitragers have taken advantage of the profit available between the price on an external exchange and the price on the internal exchange. Once this has happened, only then will the feed price update to compensate for this discrepancy.

IMO the gap between any large settlement and the blockchain's ability to adjust to it will mean the blockchain will lose every time. Thoughts?

The high memory usage is probably a huge factor in the bad performance of the bitshares client. Why does it need so much RAM compared to bitcoind?

Votes for delegates are built into transactions - the stake transferred becomes the vote, but how do you arrive at the totals?

It isn't initially obvious where the source of complexity in the implementation of market pegged bitAssets (MPAs from here on) comes from. It might be informative to discuss and reason through the key points.

As I see it, the key source of complexity in the model is the need for MPAs to remain in existance after the contract which was formed between a matching short/long pair is ended. Intuitively, an MPA should cease to exist after the contract ends, but this isn't a very useful property for a currency to have - once you buy it, it should remain yours.

If the market was just pure CFD, the problem wouldn't exist, but then MPAs also wouldn't exist, since a long and a short would both be forced to close their orders and end up back in the same currency they started (BTS).

Can the two approaches be linked in a way that results in less complexity than the current or proposed bitAsset models?

Hi guys,

I wanted to brainstorm refund addresses in metaexchange. As it stands bitcoin transactions are refunded if necessary and sent back to where they came from. This is fine mostly, except when users withdraw from an exchange or hosted wallet of some kind where they don't own the addresses they send from.

We could add a 'refund address' parameter to metaexchange to fix this problem, but it can be gamed because we store and associate data with the bitshares account name (deposit address, and soon price and expiry time). An attacker could simply run through all bitshares account names on the site adding their own bitcoin address as the refund address for each account.

You might suggest that we don't store refund address in the DB, making them single use. The problem with that is people often store their metaexchange deposit addresses in their wallet and don't even use the site at all when they want to convert funds (which is entirely within our design spec).

I'd love to hear if anyone has an alternative to address this problem?

Cheers, Paul.

wallet_get_transaction behaves differently from wallet_account_transaction_history in that the ledger data returned from wallet_get_transaction contains only public keys (in from_account, to_account), even when the account names are registered. wallet_account_transaction_history however, returns account names in from_account, to_account.

Are there any legitimate uses for an inverse USD pegged asset? Buying bitUSDINV would be like shorting bitUSD.