The idea of an insurance or MAS is to insure against devastating, low probability events that happen to people largely statistically independent.

I saw some catastrophic events take place during that selloff. I'm sure some were unavoidable, but I'm willing to bet some should have been easily preventable.

. In this case the danger of margin calls moves identically for all MAS members.

At a glance, this appears to be the case, but I do not believe the notion holds when you have large sums of data (such as that which is already stored on our blockchain) that you can analyze. One example: the attentiveness with which debt positions are maintained (adding collateral s needed) on average is something we can measure with relative ease. So if a situation were to arise where a member who normally is on top of his game when it comes to adding collateral, who additionally already had the necessary liquidity to increase their collateral on a transparent public ledger (easily verifiable, given the right prerequisite information having been provided by members), the safety net would trigger for them.

If the MAS only activated safety nets for members under specific conditions, the details of which all members must be obviously be consensus about by default, I see no reason that the creation of such a MAS would not be a worthwhile endeavor . I'm not talking about creating an insurance policy to keep people from getting called because they don't have the collateral - obviously that would implode in not time at all. I'm talking about a safety net that can protect them when they have all they need to avoid disaster, but due to circumstances for whatever reason are unable to sign & broadcast the necessary transaction.

It can be a MAS, it can be something else, the specifics aren't important.

I'm suggesting a MAS, because it seems appropriate to me for this use case. What I'm really getting at though, is in order for BitShares ecosystem to be robust, we need to create infrastructure that minimizes the human element as a risk factor disasters that cause capital to evaporate. We need to identify weak points, such as the ones that (potentially) are revealed every time there is a massive selloff like that, and address them. If you think my proposal is no good for the problem, please suggest an alternative. If you don't believe that there is a problem that need addressing, please try and convince the rest of us.