You believe because one side gets margin-called and has to actively maintain collateral that that creates a premium. I don't believe that.
With respect, you're wrong. I can prove it:
Lets call the risk of getting margin called Rm and the risk of a black swan Rb. Now, each of these risks has a probability associated with it (call them Pm and Pb)... Even if they are very difficult to quantify, we can say with certainty that Pm > Pb at all times.
Equilibrium price in the bitUSD market is the feed price. When I borrow bitUSD from the market I take on Rm. When I buy bitUSD from the market I take on Rb. If I borrow bitUSD and then sell it at the feed price, I'm saying Rm == Rb, which is obviously a false assumption. If we know that Pm>Pb at all times, then as a rational trader I must sell by borrowed bitUSD greater than the feed price, to make sure I satisfy Pm>Pb.
That's not really a proof, but it's just your hypothesis about what creates the premium. I don't agree that that causes the premium and if anything it would have a negligible effect. The risk of margin-calls should not be a significant risk to a trader, especially those with sufficient collateral. Price risk is the main concern for both long and short. When each side enters a trade they will be mainly concerned if the market is up or down when they want to exit. They should already take into consideration any other risk before entering a trade.
What premium are we talking about anyways? Just want to make sure we are taking the same perspective. I haven't been trading and have been speaking from a theoretical standpoint.. but currently I see ask for BTS below the settlement price/price feed on openledger. I see an ask of 387,587 BTS at .0030 USD/BTS and the settlement price at .0032 USD/BTS. This makes sense to me because the Margin Call price is .00291 USD/BTS when I want to open a short....so the SQP should be set to the price feed because buyers know there will be actual margin call settlement at .00291 USD/BTS instead of .0032 USD/BTS so there is no reason to bid anywhere near the higher 'Settlement Price' ... That's the biggest issue right now.
Note: Furthermore, we are trying to analyze all this in an illiquid internal market of maybe 20-30 traders? The price feed is just an estimate of prices on various exchanges. The pricing differences between exchanges can fluctuate 10+% so even trying to make a judgement about if there is a premium/discount between any two exchanges is questionable purely on volatility uncertainty. You also have different levels of fees (transaction/gateway), liquidity, custodial risk, ease-of-use etc., centralized vs decentralized, leverage/no leverage... so there should be a lot of factors in play. Beyond that we have bear market and bull market pressure that will effect premiums/discounts esp in leveraged markets...as well as the potential of forced settlement hovering over shorts in the internal exchange.... those are all factors.
I do agree very lowly-collateralized traders may have more margin-call/volatility/optionality risk and hence would not enter a trade if their own risk preference is not met. As collateral increases, the volatility risk should become negligible. Those that argue for random walk theory would argue volatility risk would not matter anyways because the probability that an asset goes up should be equal to the probability it goes down, but that's for another discussion. Either way these issues may effect liquidity and the price someone is willing to enter. As stated above premiums/discounts are based on comparing trading on one exchange to another and there are a variety of other factors at play there....
Bottomline: I say fix the SQP to equal the price fee in this illiquid market and you'll see trading happen around the price feed....