The current proposal for 2.0 is that when the least collateralised short has insufficient collateral remaining to cover 100% of their debt, then a liquidation event is triggered. The liquidation event settles all shorts, and leaves the pool of remaining collateral for longs to settle against at a time of their choosing. In effect, the shortfall loss from any under-collateralised shorts is shared by all the longs, and thereafter the token no longer reflects price movements of the original bitAsset, but only the collateral.
I've been thinking about a couple of issues with this. First, even an immaterial under-collateralisation loss would result in complete liquidation of the market. That could be detrimental for popular Smartcoins (e.g. bitUSD or bitCNY) and affect BitShares' brand even if it occurred once or twice (especially if there is a rush to sell the collateral, leaving big losses). Second, the full collateral backing of the token is not available to longs in such an event - they bear a loss as soon as the least collateralised short becomes under-collateralised, equal to the shortfall. This is even though total collateral in the pool may still be much greater than 100% of total debts. (As an aside, the collateralisation levels we generally quote probably give a false sense of security, because adequately collateralised shorts also have a claim on that collateral).
I would like issuers of privatised Smartcoins to have more flexibility in the rules governing black swan events, so that they can manage these issues in different ways they think is appropriate - without making judgements in this post on what ought to be done, because people will have different preferences here.
Here are some key options I would like to see:
i) Ability to let the pool of longs absorb the loss, while leaving the market open, and only allowing business-as-usual settlements at the reduced value (equivalent to what the liquidation value would be at any time under the 2.0 approach). This maintains equity amongst tokens that settle and those that remain trading. However, it also means that the shortfall amount (under-collateralisation loss) that has not been diluted through settlements, continues to live on as a deduction against fair settlement value for remaining longs, unless there is some way to plug this gap (either through (iii) below, or if the market cap grows significantly and the percentage discount gets low enough, through (ii)).
ii) Ability to have the pool of shorts take the first loss, and leave the market open. That is, spread losses from under-collateralised positions across remaining shorts, via a pro-rata increase in the obligation of each short. This protects longs and allows the market to keep trading at parity as long as the total collateral in the pool is sufficient to cover all debts. In this case shorts share a collective interest in the insurance of the Smartcoin.
iii) Initiate a transparent reputation system for shorts, that incentivizes shorts to make good on margin called positions that were under-collateralised, or the broader community to voluntarily rectify under-collateralisation events. So for example, a short that is margin called and under-collateralised, has a debit against their name for the shortfall amount. This might prevent them from taking short positions on other assets, assuming issuers have the ability to grant or reject authority to individual shorts. If these shorts make good on this amount, within an acceptable time, that debit is removed, and their reputation maintained. Anybody else in the community could also voluntarily contribute to plugging an under-collateralised position, and receive recognition via a credit. This process would require a mechanism for users to contribute to the general collateral pool.
I don't have all the answers here, just trying to see what options are available other than a forced liquidation event. Feel free to express your views or other ideas.
Topic: Alternatives for dealing with Black Swans? (Read 1488 times)
