Author Topic: Announcement on BSIP42 relevant actions  (Read 3814 times)

0 Members and 1 Guest are viewing this topic.

Offline JohnR

  • Committee member
  • Full Member
  • *
  • Posts: 64
    • View Profile
Re: Announcement on BSIP42 relevant actions
« Reply #75 on: October 13, 2018, 02:46:47 pm »
I read the entire thread and of course it is possible I misunderstood you.

* A trader who really wants to convert his bitUSD to BTS doesn't care where he will get the BTS,
  1) the BTS can come from other traders who are selling BTS for bitUSD
  2) the BTS can come from bitUSD debt positions who has less than required collateral ratio (margin calls)
  3) the BTS can come from bitUSD debt positions who has least collateral ratio in the system (force settlements)

You very clearly laid out the three ways for bitUSD holders to get BTS today correct?  And disabling force settlement would eliminate option 3 correct?  What I am trying to point out is that both option 1 and 2 require something from another party (posting a limit order in 1 & holding an undercollateralized position in 2).  In this way it chips away at the trust-less nature of bitUSD.  Not entirely, but recognize it does injects a level of trust in another party.  The community can elect to go this route of course, I would be interested to see how a vote ended up on this front.  You changed your mind on this issue, so maybe my thinking is less evolved.  I'm definitely open to it but rather than disabling maybe it would be better to make it an advanced feature so newbies don't stumble into it.  This would allow us to still claim that bitUSD is 'trust-less'.


Your comment "margin call (which only happens in periods of stress/when the collateral is at its lowest value" is wrong, because, with BSIP42 or another MCR-based approach, margin calls can happen at any time as long as there is more demand of selling bitUSD for BTS than buying bitUSD with BTS.


You're right, this was too strong of a statement.  I did not mean to imply it is or would be a rare event.  But surely you agree that the situation you described in response is one where the value of BTS is lower and/or decreasing.  One concern is if the platform only allows bitUSD holders to get their BTS when price of BTS is decreasing this may incentivize a run on bitUSD.


Offline spark

Re: Announcement on BSIP42 relevant actions
« Reply #76 on: October 13, 2018, 06:53:49 pm »
A `margin call` has a very specific meaning for a very specific investing concept.

So by changing what you are doing like with BSIP42, this is no longer a `margin call` in my opinion.

When the value of the collateral falls, the investor must add more collateral value or else liquidate enough or all of the position in order to maintain collateral for the loan.

Losers must pay up or else there is no incentive to win. That is what will serve everybody best.




Offline abit

  • Committee member
  • Hero Member
  • *
  • Posts: 3482
    • View Profile
    • Steemit Blog
  • BitShares: abit
  • GitHub: abitmore
Re: Announcement on BSIP42 relevant actions
« Reply #77 on: October 14, 2018, 12:10:43 am »
I read the entire thread and of course it is possible I misunderstood you.

* A trader who really wants to convert his bitUSD to BTS doesn't care where he will get the BTS,
  1) the BTS can come from other traders who are selling BTS for bitUSD
  2) the BTS can come from bitUSD debt positions who has less than required collateral ratio (margin calls)
  3) the BTS can come from bitUSD debt positions who has least collateral ratio in the system (force settlements)

You very clearly laid out the three ways for bitUSD holders to get BTS today correct?  And disabling force settlement would eliminate option 3 correct?  What I am trying to point out is that both option 1 and 2 require something from another party (posting a limit order in 1 & holding an undercollateralized position in 2).  In this way it chips away at the trust-less nature of bitUSD.  Not entirely, but recognize it does injects a level of trust in another party.  The community can elect to go this route of course, I would be interested to see how a vote ended up on this front.  You changed your mind on this issue, so maybe my thinking is less evolved.  I'm definitely open to it but rather than disabling maybe it would be better to make it an advanced feature so newbies don't stumble into it.  This would allow us to still claim that bitUSD is 'trust-less'.


No matter if bitUSD holders convert their bitUSD to BTS via margin calls or via forced settlements, the BTS are (forced) paid by debt positions with least collateral ratio, aka by bitUSD borrowers/creators, so what's the difference? IMHO there is no difference.

When someone wants to sell bitUSD but nobody wants to buy thus caused devaluation, in order to maintain the peg, the bitUSD borrowers/creators should buy back the oversupplied bitUSD (either via force-settlements or via margin calls). This is the risk that a bitUSD borrower/creator have to bear when entering the position.

Option 3 is not trustless because it depends on witnesses to provide fair price feeds.

Option 2 requires trust of witnesses to provide fair price feeds and fair MCR, but not trust of another party holding a position. There will always be debt positions if someone wants to settle or sell bitUSD, so this is not a problem. Since MCR is a subjective parameter, whether "undercollateralized" is subjective as well. In other words, witnesses are able to feed parameters to change fully-collateralized debt positions to undercollateralized when needed. So the debate is still about whether to maintain a tight peg.

Newbies are not the problem wrt removing force-settlements, advanced users who take advantages when parameters mismatch are the problem.

Quote


Your comment "margin call (which only happens in periods of stress/when the collateral is at its lowest value" is wrong, because, with BSIP42 or another MCR-based approach, margin calls can happen at any time as long as there is more demand of selling bitUSD for BTS than buying bitUSD with BTS.


You're right, this was too strong of a statement.  I did not mean to imply it is or would be a rare event.  But surely you agree that the situation you described in response is one where the value of BTS is lower and/or decreasing.  One concern is if the platform only allows bitUSD holders to get their BTS when price of BTS is decreasing this may incentivize a run on bitUSD.

IMHO BSIP42 or a MCR-based solution will work not only when trading price (I won't say value) of BTS is decreasing, but will also work when price is increasing. Increasing of MCR can cause margin calls even when BTS price is increasing.

Sorry I don't understand what "incentivize a run" means.
BTS account: abit
BTS committee member: abit
BTS witness: in.abit

Offline abit

  • Committee member
  • Hero Member
  • *
  • Posts: 3482
    • View Profile
    • Steemit Blog
  • BitShares: abit
  • GitHub: abitmore
Re: Announcement on BSIP42 relevant actions
« Reply #78 on: October 14, 2018, 12:15:40 am »
A `margin call` has a very specific meaning for a very specific investing concept.

So by changing what you are doing like with BSIP42, this is no longer a `margin call` in my opinion.

When the value of the collateral falls, the investor must add more collateral value or else liquidate enough or all of the position in order to maintain collateral for the loan.

Losers must pay up or else there is no incentive to win. That is what will serve everybody best.
There can be win-win situation, which would be better to serve everybody.

No matter whether it is traditional margin call, it is the term we're using in the system. Of course we can change it to another more appropriate word if possible.
BTS account: abit
BTS committee member: abit
BTS witness: in.abit

Offline JohnR

  • Committee member
  • Full Member
  • *
  • Posts: 64
    • View Profile
Re: Announcement on BSIP42 relevant actions
« Reply #79 on: October 14, 2018, 02:30:25 am »
Margin call is initiated by debt holders maintaining poor ratios; forced settlement is initiated by bitUSD equity holders.

The difference may be small to some but as I see it, if the ecosystem is well collateralized and there is no force settlement then it's possible that no user can get BTS autonomously.  The only option is to trade it for BTS.  Maybe that is the way of the future, but there will be a lot of old hodlers who will be suspicious that we are limiting choice.

Maybe this is just an unnecessary rabbit hole.. By "incentivize a run on bitUSD" I mean encourage a bank run where bitUSD is the bank.  In crisis deposit holders want their money out of banks.  When this happens at one time banks go bankrupt.  So by a run on bitUSD I mean a scenario when users trade en masse to BTS and turn around and sell to another crypto like BTC.  If BTS were to experience a chronic low price maybe it would not happen quite like this.  Still thinking about it.  To be honest I don't know exactly how it would play out.

Offline abit

  • Committee member
  • Hero Member
  • *
  • Posts: 3482
    • View Profile
    • Steemit Blog
  • BitShares: abit
  • GitHub: abitmore
Re: Announcement on BSIP42 relevant actions
« Reply #80 on: October 14, 2018, 11:33:51 am »
Margin call is initiated by debt holders maintaining poor ratios; forced settlement is initiated by bitUSD equity holders.

The difference may be small to some but as I see it, if the ecosystem is well collateralized and there is no force settlement then it's possible that no user can get BTS autonomously.  The only option is to trade it for BTS.  Maybe that is the way of the future, but there will be a lot of old hodlers who will be suspicious that we are limiting choice.

Maybe this is just an unnecessary rabbit hole.. By "incentivize a run on bitUSD" I mean encourage a bank run where bitUSD is the bank.  In crisis deposit holders want their money out of banks.  When this happens at one time banks go bankrupt.  So by a run on bitUSD I mean a scenario when users trade en masse to BTS and turn around and sell to another crypto like BTC.  If BTS were to experience a chronic low price maybe it would not happen quite like this.  Still thinking about it.  To be honest I don't know exactly how it would play out.
> "Margin call is initiated by debt holders"

No, margin call is initialized by witnesses via price feed. As I've said, whether a debt position is "well collateralized" is subjective.

> When someone wants to sell bitUSD but nobody wants to buy thus caused devaluation of bitUSD, in order to maintain the peg, the bitUSD borrowers/creators should buy back the oversupplied bitUSD (either via force-settlements or via margin calls). This is the risk that a bitUSD borrower/creator have to bear when entering the position.

> witnesses are able to feed parameters to change fully-collateralized debt positions to undercollateralized when needed.
BTS account: abit
BTS committee member: abit
BTS witness: in.abit

Offline Thom

Re: Announcement on BSIP42 relevant actions
« Reply #81 on: October 14, 2018, 04:52:16 pm »
No, margin call is initialized by witnesses via price feed. As I've said, whether a debt position is "well collateralized" is subjective.

> witnesses are able to feed parameters to change fully-collateralized debt positions to undercollateralized when needed.

Definitions are crucially important, so if you don't define terms we will keep going in circles. The bare minimum amount of collateral required is 100% of all short positions. Anything less and you risk a house of cards scenario like mainstream financial institutions are about to experience. We have set a standard of 175% to provide more safety to reduce margin calls, and I've seen comments we could reduce that to 130%, but for some that's not enough. 

The last statement appears to be yours abit. Putting it upon witnesses to define what they believe is too little collateral creates a conflict of power between witnesses and the Committee who sets various blockchain parameters. The minimum collateral ratio should not be determined by witnesses, it should be uniform and globally defined, and that is the role of the committee. It should not be dynamically set by each witness.

Witnesses should be explicitly "reporters of facts they observe" and block producers. Giving them the responsibility to set asset prices (and acceptable collateral levels) that differ from what can be observed in the market is manipulation and potentially opens witnesses up to liabilities and law suits.

When the MCR core bug is resolved we can experiment to achieve a tighter peg by preserving the market price feed and changing MCR or MSSR. Although some may argue that is also price manipulation b/c it achieves an almost identical effect (yet to be demonstrated, but that is general consensus), I don't see it that way.

As I've said and will continue to say, under-collateralization is my main concern regarding BSIP42. It is also Xeroc's and others who have PM'd me. Liquidity and a tight peg are obviously important, but IMHO and others not more than preserving the collateral.

I believe it would also be disastrous to the ecosystem's reputation to renig on the promise of redeemability as some are so willing to do. Yes I recognize redemption to 100% is not always possible but that is not by design as much as it is by the supply and demand of the market.
Injustice anywhere is a threat to justice everywhere - MLK |  Verbaltech2 Witness Reports: https://bitsharestalk.org/index.php/topic,23902.0.html

Offline abit

  • Committee member
  • Hero Member
  • *
  • Posts: 3482
    • View Profile
    • Steemit Blog
  • BitShares: abit
  • GitHub: abitmore
Re: Announcement on BSIP42 relevant actions
« Reply #82 on: October 14, 2018, 06:18:21 pm »
No, margin call is initialized by witnesses via price feed. As I've said, whether a debt position is "well collateralized" is subjective.

> witnesses are able to feed parameters to change fully-collateralized debt positions to undercollateralized when needed.

Definitions are crucially important, so if you don't define terms we will keep going in circles. The bare minimum amount of collateral required is 100% of all short positions. Anything less and you risk a house of cards scenario like mainstream financial institutions are about to experience. We have set a standard of 175% to provide more safety to reduce margin calls, and I've seen comments we could reduce that to 130%, but for some that's not enough. 

We had those numbers as parameters before doesn't mean we have to stick with those numbers forever. "for some"? for who?

Quote
The last statement appears to be yours abit. Putting it upon witnesses to define what they believe is too little collateral creates a conflict of power between witnesses and the Committee who sets various blockchain parameters. The minimum collateral ratio should not be determined by witnesses, it should be uniform and globally defined, and that is the role of the committee. It should not be dynamically set by each witness.

So you're expressing your own opinions as well. Witnesses act according to rules in consensus, rules are defined by BSIPs, consensus is made by stake based voting. Committee doesn't have the final say about parameters, stake holders have.
BTS account: abit
BTS committee member: abit
BTS witness: in.abit

Offline armin

  • Full Member
  • ***
  • Posts: 91
    • View Profile
Re: Announcement on BSIP42 relevant actions
« Reply #83 on: October 16, 2018, 10:26:54 pm »
No matter if bitUSD holders convert their bitUSD to BTS via margin calls or via forced settlements, the BTS are (forced) paid by debt positions with least collateral ratio, aka by bitUSD borrowers/creators, so what's the difference? IMHO there is no difference.

bitasset value = max(market price, settlement price)
If you take out settlement, this means bitasset value = market price.
The former formula gives a relative minimum price to the feed, but the latter formula looks exactly like UIAs.

So effectively, taking out settlement will make bitassets the same as UIAs in terms of value.

Thoughts?

Offline abit

  • Committee member
  • Hero Member
  • *
  • Posts: 3482
    • View Profile
    • Steemit Blog
  • BitShares: abit
  • GitHub: abitmore
Re: Announcement on BSIP42 relevant actions
« Reply #84 on: October 16, 2018, 11:40:00 pm »
No matter if bitUSD holders convert their bitUSD to BTS via margin calls or via forced settlements, the BTS are (forced) paid by debt positions with least collateral ratio, aka by bitUSD borrowers/creators, so what's the difference? IMHO there is no difference.

bitasset value = max(market price, settlement price)
If you take out settlement, this means bitasset value = market price.
The former formula gives a relative minimum price to the feed, but the latter formula looks exactly like UIAs.

So effectively, taking out settlement will make bitassets the same as UIAs in terms of value.

Thoughts?

Essentially you're saying that if you sell some amount of PMA and nobody will buy, its trading price will be pushed lower than its value.

Actually, you ignored how MPA supply is created and who are on the market. Not like UIAs, PMAs are created with debt positions, this means you won't have unlimited amount to sell. In addition, debt positions are always on the market, with BSIP42 or a dynamic MCR-based mechanism, if you place an order to sell some PMA below its value, there will be margin calls appear to fill your order, thus the trading price would be guaranteed. Force-settlements aren't in this process at all.
BTS account: abit
BTS committee member: abit
BTS witness: in.abit

Offline Thom

Re: Announcement on BSIP42 relevant actions
« Reply #85 on: October 17, 2018, 12:17:51 am »
No, margin call is initialized by witnesses via price feed. As I've said, whether a debt position is "well collateralized" is subjective.

> witnesses are able to feed parameters to change fully-collateralized debt positions to undercollateralized when needed.

Definitions are crucially important, so if you don't define terms we will keep going in circles. The bare minimum amount of collateral required is 100% of all short positions. Anything less and you risk a house of cards scenario like mainstream financial institutions are about to experience. We have set a standard of 175% to provide more safety to reduce margin calls, and I've seen comments we could reduce that to 130%, but for some that's not enough. 

We had those numbers as parameters before doesn't mean we have to stick with those numbers forever. "for some"? for who?
Bitcrab for one, and I thought you as well wanted to reduce required collateral. So I gather from various comments you both have made recently.

Quote
The last statement appears to be yours abit. Putting it upon witnesses to define what they believe is too little collateral creates a conflict of power between witnesses and the Committee who sets various blockchain parameters. The minimum collateral ratio should not be determined by witnesses, it should be uniform and globally defined, and that is the role of the committee. It should not be dynamically set by each witness.

So you're expressing your own opinions as well. Witnesses act according to rules in consensus, rules are defined by BSIPs, consensus is made by stake based voting. Committee doesn't have the final say about parameters, stake holders have.

Yes. You also stated governance model well. However, I didn't say committee had final say on parameters, but they are the only ones authorized by consensus in current graphene design to change them. Sure, if they do that badly they'll risk being voted out, nevertheless the definition of roles is clear, and I maintain far better and more inline with the current design of graphene we have consensus on, than alter the design and put a new responsibility on witnesses that would be more logical with committee.

Allowing witnesses to individually set minimum collateral level changes it from a global, uniform setting to one more akin to "pick your witness, pick your desired margin collateral level". I think it would create an unhealthy competition between witnesses.
Injustice anywhere is a threat to justice everywhere - MLK |  Verbaltech2 Witness Reports: https://bitsharestalk.org/index.php/topic,23902.0.html

Offline abit

  • Committee member
  • Hero Member
  • *
  • Posts: 3482
    • View Profile
    • Steemit Blog
  • BitShares: abit
  • GitHub: abitmore
Re: Announcement on BSIP42 relevant actions
« Reply #86 on: October 17, 2018, 01:15:43 am »
Allowing witnesses to individually set minimum collateral level changes it from a global, uniform setting to one more akin to "pick your witness, pick your desired margin collateral level". I think it would create an unhealthy competition between witnesses.
IMHO this is actually healthy. It means "vote for witnesses who are working well" when the consensus is to dynamically adjust MCR according to market conditions. Again, witnesses aren't predicting market price, they're acting according to perceived market conditions/data.
BTS account: abit
BTS committee member: abit
BTS witness: in.abit

Offline armin

  • Full Member
  • ***
  • Posts: 91
    • View Profile
Re: Announcement on BSIP42 relevant actions
« Reply #87 on: October 17, 2018, 04:22:59 pm »
No matter if bitUSD holders convert their bitUSD to BTS via margin calls or via forced settlements, the BTS are (forced) paid by debt positions with least collateral ratio, aka by bitUSD borrowers/creators, so what's the difference? IMHO there is no difference.

bitasset value = max(market price, settlement price)
If you take out settlement, this means bitasset value = market price.
The former formula gives a relative minimum price to the feed, but the latter formula looks exactly like UIAs.

So effectively, taking out settlement will make bitassets the same as UIAs in terms of value.

Thoughts?

Essentially you're saying that if you sell some amount of PMA and nobody will buy, its trading price will be pushed lower than its value.

Actually, you ignored how MPA supply is created and who are on the market. Not like UIAs, PMAs are created with debt positions, this means you won't have unlimited amount to sell. In addition, debt positions are always on the market, with BSIP42 or a dynamic MCR-based mechanism, if you place an order to sell some PMA below its value, there will be margin calls appear to fill your order, thus the trading price would be guaranteed. Force-settlements aren't in this process at all.

True, good point! Actually maybe I need to adjust the formula a little bit. Basically what you're saying is that the supply of bitassets is self-regulating with BSIP42 so that it maintains the peg. This means the supply will increase/decrease to fit a tight peg. Change in supply will lead to some change in market price. Interesting

EDIT: OOH that makes sense, so in an empty market there is almost always some form of BTS sell pressure, therefore it will only be "half empty"
« Last Edit: October 18, 2018, 04:05:23 am by armin »

Offline Thom

Re: Announcement on BSIP42 relevant actions
« Reply #88 on: October 18, 2018, 12:31:10 am »
Allowing witnesses to individually set minimum collateral level changes it from a global, uniform setting to one more akin to "pick your witness, pick your desired margin collateral level". I think it would create an unhealthy competition between witnesses.
IMHO this is actually healthy. It means "vote for witnesses who are working well" when the consensus is to dynamically adjust MCR according to market conditions. Again, witnesses aren't predicting market price, they're acting according to perceived market conditions/data.

Voters should already be voting "for witnesses who are working well", and they do so on the basis of their own subjective perspective of what that is with very few guidelines and understanding of what is important for an economic system that won't self-destruct like mainstream currencies eventually do. Although we all would like to see more people who are dedicated and attentive to the ongoing well being of BitShares, it's probably unrealistic to expect shareholders to care about issues unrelated to them, like trading if they aren't traders, or savers if they're traders as examples.

Unless all witnesses use the exact same sources how could shareholders know if they're working well? What is "working well"? For who is it working well, just the voter or the entire ecosystem? Some think it's only a matter of keeping tight pegs for MPAs, others think more broadly about the ecosystem and vote to balance out competing interests, which is a very difficult thing to do under our decentralized governance structure.

We'll just disagree on this point. I can live with that.
Injustice anywhere is a threat to justice everywhere - MLK |  Verbaltech2 Witness Reports: https://bitsharestalk.org/index.php/topic,23902.0.html