The limit creates an existential/systemic vulnerability to market conditions already. I've posted about this in the past:
https://steemit.com/witness-category/@bacchist/witness-petition-do-not-abandon-the-peg
It has little benefit to offset the risk it represents. It is intended to mitigate the risk of a rising debt load that threatens to create destabilizing amounts of liquid STEEM. But in this case, the sickness isn't much worse than the cure.
You stated that your rationale for not using a positive bias to address the peg was that you wanted to preserve the "1 USD worth of STEEM" contract. Since the debt limit/haircut formally breaks that contract, you might want to consider whether that is something we can dispense of as well... or at least raise, to the point that it won't be an immediate concern as soon as SBD production halts at 5% debt load.
Moving it to even 15% or 20% will limit the vulnerability from bad actors if your proposal goes through, as well as allowing time for the throttling of SBD rewards to bring the debt load into a manageable range before the haircut.
I see your point. It is one of those things that has it's trade-offs. It just depends on which positives/negatives you weigh more heavily.
I do see the limit as different than a price feed premium, as it is coded in the blockchain - thus part of the 'contract'. The fact that it is not very well known/communicated is a different issue, but not one of violation IMO.
I am definitely open to discussion about the 'right' numbers. I suspect the ones we have today were largely arbitrary choices, and there may be better limits to have in place that would accomplish the same goal of mitigating risk without having as much of a negative side effect.
I do think though that wherever the limit is, there will be a natural tendency for the peg to get pushed down as the debt level reaches 50% of the limit (currently 5% debt) and definitely even more so when it reaches 100% of the limit (currently 10% debt).
To a large extent, regardless of where the limit is, I think the biggest risk in the whole thing is getting to 50% of that limit, and then seeing a 50% price drop in STEEM. Unfortunately wherever we place the limit, I think that risk will always be present so long as the limit is in place.
Likewise the ability of witnesses to feed any data they want as the alleged price is also coded into the blockchain. If stakeholders don't like it they can vote those witnesses out. So far we seem about evenly divided between voted witnesses using a premium (and doing so with at least some rational basis, not just feeding garbage or clearly-malicious values) and those not, so no clear stakeholder preference, but I'd still argue none of this violates the smart contract in the code.
Hi @smooth, glad you're back (heard you took a holiday)... I don't mean to chase you all round Steemit but I need you to check your Chat messages please. Thanks!!!
It is a valid point, and I know that you didn't mean it literally, but I think the idea that the witnesses can put in 'garbage' and cause the conversions to produce less than $1 USD worth of STEEM without getting voted out is a precedent I feel we should avoid. Not to say there aren't valid reasons for the 'garbage' but it does undermine trust in a part of the system that relies heavily on trust.
Maybe. When one considers "existential/systemic" vulnerabilities as @bacchist claims, there is a reasonable argument that should override other tradeoffs (at least others that do not also involve such systemic risks).
That being said, one can question whether his claim is correct. I have some similar concerns.
I see existential/systemic vulnerabilities if there is no debt limit :)
Potentially that argues to look for a way to wind down SBD altogether, and remove the threat. Or look for a third option. Dan's earlier proposal for forced liquidation has a lot going for it, but unfortunately creates serious ecosystem issues. So we need a fourth option I guess...