On whether we should have have a bitUSD centred on parity or at a premium, assuming both are possible:
I do not understand all the subtleties of merchant, payment processor and other third party operations. However, it seems to me that everything that can be done with a bitUSD trading at a premium can also be accomplished with a bitUSD trading at parity, only with different labels on the currency flows. For example the equivalent financial result for all parties can be achieved by having a bitUSD centred on parity, with consumers paying the fee directly to the payment processor. However, maybe bytemaster is right, based on his advice, that the way in which fees are labelled or packaged in the marketplace makes a big difference to actual behaviour. I can't argue with the way other people think.
I do have a concern however that a bitUSD at a premium reduces its utility for a number of other application areas that I touched on in my previous post. A bitUSD trading around parity would serve these applications much better IMO. Regardless of the direction taken in the near term, I would like to continue pursing the method for a bitUSD based around parity, because I personally think it would give bitShares the best long term result.
On whether it is possible to have a bitUSD that trades close to parity: I reject the notion that we can establish a peg with 0 premium and 0 volatility in the premium and those seeking this perfection are seeking the impossible.
This is knocking down a straw man, and perhaps you have misunderstood my goal. What I wish to see is bitUSD that centres around parity rather than a premium. Clearly there will be deviation either side of that with movement in supply and demand, at least to the point that arbitragers are incentivised to jump in and prevent larger deviations. This is the model I am working toward. There is no plan for 0 volatility.
Do you think this notion is impossible? I don't have enough evidence of that yet.
Lets take a "perfect" and "balanced" system of a simple prediction market that settles once per year at a "perfectly fair" price. In this prediction market both sides have the exact same liquidity and expectation of profit. Any deviations from the value of $1.00 is a profit opportunity for both parties.
Under such a system what is the "volatility" of the premium? It is proportional to the holding cost until settlement combined with the demand for leverage on BTS. I would expect a range around $1.00 that could be off by over 10% at times. So even with guaranteed settlement at a fair price on a specific date in the future you do not get an asset that is always worth $1.00 with a stable premium. The best we can do is get an asset that is approximately worth $1.00 and the volatility in the premium is beyond our control.
This is not the most useful comparison for bitUSD. What you describe here is a futures contract, that expires and settles at a fixed date in the future. The fair price for a futures contract is not equal to the spot value of the underlying asset, because of the cost of carry. That is, the difference between the funding cost related to the currency and that of the BTS in which it is settled. The market will price this in when it compares the cost of establishing a position with similar exposure in the external market. The fair value for the future will also fluctuate as the funding costs in these markets fluctuates. None of this means that the future is trading at wide variations from its fair value, its just that its fair value is
not parity. In a different context, we would not argue for example that a discount bond is trading at large deviation from its fair value because it is trading at a large discount to its par value.
In these cases however, large deviations from the "fair value" are contained by the action of arbitragers, and that is really what is important.
A bitUSD is a tracking instrument, more like an exchange-traded fund (ETF), in that it does not expire. The fair value of an instrument like this
is its underlying asset value. Instruments like ETFs do trade close to the spot value because of the action of arbitragers in those markets. Authorised Parties are granted the ability to exchange units in the fund for units in the underlying asset, making this arbitrage possible. I think a similar mechanism is possible for a bitUSD.
On whether it is possible to have a stable bitUSD premium:
If the premium is persistently biased by a certain minimum amount then the market will end up adjusting prices to account for that bias. This "constant bias" can be relatively stable and can be removed by adjusting the feed.
Would you mind expanding on this mechanism as I still don't understand it? If the idea is to adjust the feed price downward (equivalent to a higher settlement fee) for larger bitUSD premiums, I've raised my concern about that in the link below. However, I could also be at risk of knocking down a straw man if I misunderstand your intended mechanism.
link:
https://bitsharestalk.org/index.php/topic,16143.msg210901.html#msg210901