Agent86 will be happy.
That is about the only positive thing in today's changes/announcements.
Let's trigger Black Swan event with 125% average collateralization (or more... choose your number) + screw the shorts even more + let's add another not so useful order type to our buggy as hell product.
How does this screw the shorts even more? Removed 5% fee and allowing them to cover without maintaining an extra 100% BTS outside their collateral position. This move is very PRO-SHORT because it actually reduces the amount they must keep on hand.
Lets assume a "WORST CASE" pre-mature calling of a Black Swan. Prior to the swan there was 1 short with 201% collateralization and all other shorts were just created with 300% collateralization. The price falls by 51% while no orders are on the book. The 1 short hits 100% collateralization, all of the other shorts are at 150% collateralization (below the maintenance position). So we could say that "worst case" a black swan is called with 150% *average* collateralization, but that would be a black swan among black swans. The far more likely case is that short positions are evenly spread prior to the swan. The swan hits and the least collateralized are covered in the market by consuming the order book. As the price continues to fall eventually the order book runs out. Lets assume that 10% of all open shorts cover into the open order book, this makes the new "least collateralized" position 210% which falls to 100% to trigger liquidation. In this case the average collateralization at time of liquidation would be just 121%
So I would wager that the probability of having exactly a 53% fall to trigger the minimal possible black swan is very low, it could easily fall 60% if it just fell 53%. This means that while the swan may have been called "early" these rules once again FAVOR the pre-swan shorts by minimizing their losses.
I didn't add a new order type for this, I merely removed a fee and allowed the user to increase the call price.
1.So you made my point on the black bird event -not 125%, 121% in your average scenario and you agree it could be much more - 150% in your example. And while 120% might not seem enough, but that is because of your arbitrary 200% margin call level. In my view 133% collateral is about the right margin call level and because of that I consider 125% average collateral way too fast triggered….
2. On the shorts being screwed issues:
a. I continue to insist that shorts are paying interest to lend money.Due to the simple fact that they provide 2x collateral! Discussed in detail elsewhere.
b. Going from 150% collateral to 200% to trigger a margin call is not screwing them up more? I get the consistency needed to match your writings and the actual code but I suggest changing your blog not the code.
3. No new order type? That’s nothing short of strange...
What is the current order type that does NOT follow the 'Get what you asked for' market mechanics?
Assuming margin order for 100bitUSD (call price of 100 BTS/bitUSD). Feed price of 100 BTS/bitUSD
And sell bitUSD order book:
20 bitUSD @ 100 BTS/bitUSD
40 bitUSD @ 105 BTS/bitUSD
40 bitUSD @ 110 BTS/bitUSD
How you gonna match those with our margin order?
-First scenario (aka screw the short even more), no new orders required probably
, pay for 100 bitUSD @ 110 BTS/bitUSD
-Make a new order type that matches to the price of each ask order, do not screw the shorts, but allow for potential front running?