RE: How do we fix the peg without popping the SBD bubble? I may have an idea.
The problem with this, or really any similar proposal is that there is no way to tell what can trigger a "shock" crash. Markets can exhibit chaotic behavior, like snowflakes falling on a mountainside. All is quiet as the snow accumulates until, at some point, the snowbank suddenly gives way and there is a huge avalanche.
Even though this proposal doesn't by itself force a crash, it may still trigger one, as market participants realize the inevitability of soon adjusting back to $1. Or doing nothing at all we may still have a crash (as I write this SBD is down 16% in the last 24 hours, which isn't quite a crash, but shows that big down moves for no particular reason are very possible).
The only reliable way to avoid a crash is to set the peg high and lower it slowly over a period of time. This may, somewhat perversely, require the peg to function to increase the SBD price even though it is grossly overvalued, just to prevent it from falling more quickly (i.e. crash). That doesn't require an explicit mechanism at all, only a consensus of witnesses to adjust the price feed accordingly, and then remove the adjustment over a defined schedule. (To have a reliable peg regardless of the peg price still requires adjustments to the mechanism however.)
That's exactly what I am proposing: a slow lowering of the peg to the desired level. A 1.7% drop per day would give us 94 days from $5 to $1. If the market moves quicker we could lower the peg sooner, or enforce the slower reduction rate via the SBD feed. That way the price of SBD drops slowly in a controlled manner, avoiding a crash.
The mechanism must change. We could find ourselves faced with a SBD pump to well over $10 again very soon, but if we have a plan to implement a slowly moving downward peg, that may prevent a massive pump at the very least. The first step is to cap SBD.
Yes you are right. When you wrote
I misread that as implying that it you would allow SBD to float between $1 and the ceiling. But in fact you meant there would be a strict peg that could not move above the ceiling, more or less the same as what I suggested.
In any case, this does not require a SBD price feed, which is fortunate because a code change like that would be significantly more substantial.
What it requires is:
Bingo! That's a lot easier than I explained.. but I don't know how the cogs turn all that well. It that can be done, then it just takes agreement, and a minor change to enable bidirectional conversion.
Could this work in practice?