Author Topic: Neolithically simple alternative bitasset peg  (Read 4474 times)

0 Members and 1 Guest are viewing this topic.

Offline monsterer

If what you are imagining is a continual decline in the BTS price that forces increasingly heavy dilution on BTS, yes that would be an issue. The way to avoid infinite dilution I think is to pass some risk to the bitAsset holders through lower percentage redemptions as collateralisation falls below 3x. Or have you given up on this concept altogether - when I read it I thought it was an interesting variation on bitShares that could be tried.

I think the better way would have been to have the purchase/sale price depend on the volume in some continuous linear fashion - so during the course of one, say buy order for 1M bitUSD, the sale price the blockchain provides would start at the tightest possible spread and decrease linearly as these 1M bitUSD are dealt out to the buyer, so the price gets more expensive as each part of the buy order is consumed.

The problem with this is that you need to define the 'steepness' of this linear increase somehow, and that depends on a 'liquidity factor' which I don't really know how to estimate.

This is basically the market maker model that Kyle et al talk about in their 1985 paper on the subject http://people.stern.nyu.edu/lpederse/courses/LAP/papers/Information,Fundamental/Kyle85.pdf

Bytemaster's point was that if you have a free market, this liquidity factor appears automatically in the shape of the order book, so you end up trying to replicate what we already have now.
My opinions do not represent those of metaexchange unless explicitly stated.
https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline bitmeat

  • Hero Member
  • *****
  • Posts: 1116
    • View Profile
If there is $1M held in BitUSD and other assets, then the BTS necessary to cover is about 2500 per BitUSD.

We have 50 BTS per BitUSD valuation at the moment, that's 50 times less than the price where it really couldn't move further based on utility alone. So the real demand for BTS is 1 part BitUSD and 50 parts pure speculation.

Offline starspirit

  • Hero Member
  • *****
  • Posts: 948
  • Financial markets pro over 20 years
    • View Profile
  • BitShares: starspirit
...potentially causing excessive dilution as the pool becomes under-collateralised.
Could you clarify this bit? Not sure of the situation you are envisaging.

It's basically adverse selection cost - it is possible to be selected against to such a degree that BTS becomes excessively devalued. The problem lies in the fixed pricing and 'infinite' supply available at those prices.
If what you are imagining is a continual decline in the BTS price that forces increasingly heavy dilution on BTS, yes that would be an issue. The way to avoid infinite dilution I think is to pass some risk to the bitAsset holders through lower percentage redemptions as collateralisation falls below 3x. Or have you given up on this concept altogether - when I read it I thought it was an interesting variation on bitShares that could be tried.

Offline monsterer

...potentially causing excessive dilution as the pool becomes under-collateralised.
Could you clarify this bit? Not sure of the situation you are envisaging.

It's basically adverse selection cost - it is possible to be selected against to such a degree that BTS becomes excessively devalued. The problem lies in the fixed pricing and 'infinite' supply available at those prices.
My opinions do not represent those of metaexchange unless explicitly stated.
https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline starspirit

  • Hero Member
  • *****
  • Posts: 948
  • Financial markets pro over 20 years
    • View Profile
  • BitShares: starspirit
...potentially causing excessive dilution as the pool becomes under-collateralised.
Could you clarify this bit? Not sure of the situation you are envisaging.

Offline monsterer

There would still be a free market in the trading of bitAssets against fiat, cryptos and even BTS on external exchanges. Market prices and spreads would reflect the timing risks of going through BTS to purchase and redeem units, as well as the market's perception of collateralisation and partial redemption risks. To some extent this may self-regulate the demand for bitAssets where there is a growing chance of black-swans or other risks.

I'm not sure if that was the OPs thinking or not. I was just trying to understand how it might feasibly work.

Your interpretation of the OP is actually much nicer than what I was proposing, but along the same lines. I think the problem still lies with the fixed-price-for-any-volume that this system provides.

If a bitAsset has fixed buy/sell prices and the the price moves more than the spread, it will become possible for arbitragers to take advantage of this on an external exchange, potentially causing excessive dilution as the pool becomes under-collateralised.
My opinions do not represent those of metaexchange unless explicitly stated.
https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline starspirit

  • Hero Member
  • *****
  • Posts: 948
  • Financial markets pro over 20 years
    • View Profile
  • BitShares: starspirit
The current bitasset pegging system is so fantastically complex because the assets are traded on a free market, which can be manipulated by any other market participants.

A neolithically simple alternative way to peg a bit asset to an external price would be to remove the other participants from the equation.

Create a different class of asset, which can only be bought and sold by the issuer. The issuer maintains a buy and sell price with a spread which takes into account adverse selection costs.

Issues:

100% trust is required of the issuer. Trust that they don't sell you the asset and then raise their buy price to $1M USD. Trust that they have enough liquidity to cope with adverse selection costs, so they're able to buy back the assets they sold you after the price goes up.

Now, take that idea and make the blockchain the issuer.

You then gain 100% transparency on the algorithm controlling the prices. You'd be able to see the 'balance' of the blockchain in advance to be sure it has liquidity reserves.

In addition, bitshares is ideally placed to implement this idea, since the necessary price feeds are already a fundamental component.

The only real remaining issue is designing the algorithm to set the bid/ask prices - perhaps something like the kalman filter (http://www.r-bloggers.com/the-kalman-filter-for-financial-time-series/) could be used over the feed data.

Thoughts?
Not sure I understand the reserving. At the moment we have 3x collateral in BTS, 1x contributed by the buyer, and 2x contributed by the issuers (the shorts). What liquidity reserves are held in this proposal to back the USD, and from where are they contributed?

I was trying to understand whether such a system could work. I'm not necessarily recommending this, just thinking through. Is the following feasible from a collateral perspective, or did the OP have other ideas?

The block-chain could issue and redeem bitUSD against BTS at the price feed, normally no spread. The BTS collateral would be 1x (being what the purchaser pays) as there is no counterparty contributing further collateral at the point of purchase.

So as the BTS price moves around the pool will drift between being under-collateralised and over-collateralised at any point. When there are redemptions against the pool in an under-collateralised state, new BTS are created through dilution to make up the difference between that holder's share of the pool and full redemption value. When there are redemptions against the pool in an over-collateralised state, BTS is burned to the extent the holder's share of the pool exceeds their full redemption value. Should the BTS price rise over time, there should usually be more BTS burned than created.

So as an example, if bitUSD were $10m market cap and 80% collateralised, and BTS were $100m market cap, then if all holders redeemed there would be just over $2m of BTS created (allowing for the effect of dilution on the price per share) to ensure holders are paid out $10m in total.

In this way the pools of bitAssets are effectively backed by the entire market cap of BTS, and are therefore as secure as possible within the limits of the block-chain (i.e. no external reserving).  A black-swan event is still possible should the total market cap of BTS fall below the total market cap of the bitAssets combined, because at that point dilution can't help. And a bank-run type event is possible should the BTS price be only slightly over-collateralised, start redeeming en masse and selling BTS in a vicious cycle. Mitigation would be required in these events to ensure bitAsset holders are treated fairly and BTS holders not destroyed. For example, there might be ways for the block-chain coding to regulate the supply (e.g. through spreads) or even to redeem at lower prices (passing risk to bitAsset holder) should the BTS market cap start falling too close to this territory.

There would still be a free market in the trading of bitAssets against fiat, cryptos and even BTS on external exchanges. Market prices and spreads would reflect the timing risks of going through BTS to purchase and redeem units, as well as the market's perception of collateralisation and partial redemption risks. To some extent this may self-regulate the demand for bitAssets where there is a growing chance of black-swans or other risks.

I'm not sure if that was the OPs thinking or not. I was just trying to understand how it might feasibly work.

Offline bytemaster

Just want to make clear that I already suggested this as plan 'E' several months ago...the problem is we went from plan A to plan B without any need for doing so, other than delaying the whole project with a good 2 month and counting...

There was a need to move to the feed... you could short a BitAsset out of existence and market confidence in the peg would force it to 0 allowing shorts to cover cheaply.    Plan A only works in very large and liquid markets that are relatively equally balanced between bulls and bears.   

It was never tried!

I am fully aware that Vitalik and Agent86 think so...this does not make it a proven fact.

Speaking of things that work, we are all waiting for either of those 2 geniuses [no sarcasm, I really think they are both extremely smart] to actually produce something that actually works, instead of having their opinions on the BTS system and design...


And @ speedy the same is true for plan A... the bitUSD is discounted up  to 10% in plan B, it was discounted upto about 5-6% during plan A.

What do you think I do, take their opinion without thinking it through?   You imply that they are wrong and at the same time I am not able to evaluate their opinions and draw my own conclusions.

I find them too persuasive for my liking...

How they manage it...I have no clue...I am totally lacking this talent, so it is a mystery to me. Honestly it is.

What really really bothers me is the extremely suspicious timing of A86' discovery that the market engine #1 'will not work'

He has been an integral part for month and months of discussions about the market peg, yet he figured out that the peg won't work for sure , only when everything was already in place and coded.

Very nice timing if you want to slow the project the most or real unproductive thinking for months... Hmmm... Reading A86 posts, I can accuse him of many things, but I will never dare to say that he is not extremely smart guy.

Tony, I am a very patient guy and you are pushing the limits here.   
For the latest updates checkout my blog: http://bytemaster.bitshares.org
Anything said on these forums does not constitute an intent to create a legal obligation or contract between myself and anyone else.   These are merely my opinions and I reserve the right to change them at any time.

Offline liondani

  • Hero Member
  • *****
  • Posts: 3737
  • Inch by inch, play by play
    • View Profile
    • My detailed info
  • BitShares: liondani
  • GitHub: liondani
Multi Sig and Send to Address
Relative Bid / Ask Orders
Escrow Transactions

And that is just the tip.

I am confident that the peg will hold  MUCH better after the  Relative Bid / Ask Orders implementation...
I said that in the past too... In my opinion it will be the cut of the Guardian Knot...



zerosum

  • Guest
Just want to make clear that I already suggested this as plan 'E' several months ago...the problem is we went from plan A to plan B without any need for doing so, other than delaying the whole project with a good 2 month and counting...

There was a need to move to the feed... you could short a BitAsset out of existence and market confidence in the peg would force it to 0 allowing shorts to cover cheaply.    Plan A only works in very large and liquid markets that are relatively equally balanced between bulls and bears.   

It was never tried!

I am fully aware that Vitalik and Agent86 think so...this does not make it a proven fact.

Speaking of things that work, we are all waiting for either of those 2 geniuses [no sarcasm, I really think they are both extremely smart] to actually produce something that actually works, instead of having their opinions on the BTS system and design...


And @ speedy the same is true for plan A... the bitUSD is discounted up  to 10% in plan B, it was discounted upto about 5-6% during plan A.

What do you think I do, take their opinion without thinking it through?   You imply that they are wrong and at the same time I am not able to evaluate their opinions and draw my own conclusions.

I find them too persuasive for my liking...

How they manage it...I have no clue...I am totally lacking this talent, so it is a mystery to me. Honestly it is.

What really really bothers me is the extremely suspicious timing of A86' discovery that the market engine #1 'will not work'

He has been an integral part for month and months of discussions about the market peg, yet he figured out that the peg won't work for sure , only when everything was already in place and coded.

Very nice timing if you want to slow the project the most or real unproductive thinking for months... Hmmm... Reading A86 posts, I can accuse him of many things, but I will never dare to say that he is not extremely smart guy.


Offline monsterer

So this means a variable spread based upon redemption demand.   If only $1 BitUSD is demanded you can do it at the feed.  If $1 Million BitUSD is demanded it will have to be at a steep premium. 

So if you can identify the proper function for adjusting the price relative to liquidity demands then you can emulate what the market is already doing today....

These are excellent points. There is undoubtedly a proper function for this, I think that's what the famous Kyle market maker paper was written about http://people.stern.nyu.edu/lpederse/courses/LAP/papers/Information,Fundamental/Kyle85.pdf.

The thing is, this plan is scuppered if the price is permanently altered away from the feed after just such a massive order (just as it would be in a free market). I'm not sure if that would be a requirement, or whether the risk associated with filling the order is completely covered by the fact that the spread changes as the volume in the order is consumed by the system.

edit: I.e. does it elastically snap back to what it was pre-order.
« Last Edit: November 17, 2014, 10:13:27 pm by monsterer »
My opinions do not represent those of metaexchange unless explicitly stated.
https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline starspirit

  • Hero Member
  • *****
  • Posts: 948
  • Financial markets pro over 20 years
    • View Profile
  • BitShares: starspirit
The current bitasset pegging system is so fantastically complex because the assets are traded on a free market, which can be manipulated by any other market participants.

A neolithically simple alternative way to peg a bit asset to an external price would be to remove the other participants from the equation.

Create a different class of asset, which can only be bought and sold by the issuer. The issuer maintains a buy and sell price with a spread which takes into account adverse selection costs.

Issues:

100% trust is required of the issuer. Trust that they don't sell you the asset and then raise their buy price to $1M USD. Trust that they have enough liquidity to cope with adverse selection costs, so they're able to buy back the assets they sold you after the price goes up.

Now, take that idea and make the blockchain the issuer.

You then gain 100% transparency on the algorithm controlling the prices. You'd be able to see the 'balance' of the blockchain in advance to be sure it has liquidity reserves.

In addition, bitshares is ideally placed to implement this idea, since the necessary price feeds are already a fundamental component.

The only real remaining issue is designing the algorithm to set the bid/ask prices - perhaps something like the kalman filter (http://www.r-bloggers.com/the-kalman-filter-for-financial-time-series/) could be used over the feed data.

Thoughts?
Not sure I understand the reserving. At the moment we have 3x collateral in BTS, 1x contributed by the buyer, and 2x contributed by the issuers (the shorts). What liquidity reserves are held in this proposal to back the USD, and from where are they contributed?

zerosum

  • Guest
Just want to make clear that I already suggested this as plan 'E' several months ago...the problem is we went from plan A to plan B without any need for doing so, other than delaying the whole project with a good 2 month and counting...

There was a need to move to the feed... you could short a BitAsset out of existence and market confidence in the peg would force it to 0 allowing shorts to cover cheaply.    Plan A only works in very large and liquid markets that are relatively equally balanced between bulls and bears.   

It was never tried!

I am fully aware that Vitalik and Agent86 think so...this does not make it a proven fact.

Speaking of things that work, we are all waiting for either of those 2 geniuses [no sarcasm, I really think they are both extremely smart] to actually produce something that actually works, instead of having their opinions on the BTS system and design...


And @ speedy the same is true for plan A... the bitUSD is discounted up  to 10% in plan B, it was discounted upto about 5-6% during plan A.

What do you think I do, take their opinion without thinking it through?   You imply that they are wrong and at the same time I am not able to evaluate their opinions and draw my own conclusions.

I find them too persuasive for my liking...

How they manage it...I have no clue...I am totally lacking this talent, so it is a mystery to me. Honestly it is.


Offline bytemaster

Just want to make clear that I already suggested this as plan 'E' several months ago...the problem is we went from plan A to plan B without any need for doing so, other than delaying the whole project with a good 2 month and counting...

There was a need to move to the feed... you could short a BitAsset out of existence and market confidence in the peg would force it to 0 allowing shorts to cover cheaply.    Plan A only works in very large and liquid markets that are relatively equally balanced between bulls and bears.   

It was never tried!

I am fully aware that Vitalik and Agent86 think so...this does not make it a proven fact.

Speaking of things that work, we are all waiting for either of those 2 geniuses [no sarcasm, I really think they are both extremely smart] to actually produce something that actually works, instead of having their opinions on the BTS system and design...


And @ speedy the same is true for plan A... the bitUSD is discounted up  to 10% in plan B, it was discounted upto about 5-6% during plan A.

What do you think I do, take their opinion without thinking it through?   You imply that they are wrong and at the same time I am not able to evaluate their opinions and draw my own conclusions. 
For the latest updates checkout my blog: http://bytemaster.bitshares.org
Anything said on these forums does not constitute an intent to create a legal obligation or contract between myself and anyone else.   These are merely my opinions and I reserve the right to change them at any time.

zerosum

  • Guest
Just want to make clear that I already suggested this as plan 'E' several months ago...the problem is we went from plan A to plan B without any need for doing so, other than delaying the whole project with a good 2 month and counting...

There was a need to move to the feed... you could short a BitAsset out of existence and market confidence in the peg would force it to 0 allowing shorts to cover cheaply.    Plan A only works in very large and liquid markets that are relatively equally balanced between bulls and bears.   

It was never tried!

I am fully aware that Vitalik and Agent86 think so...this does not make it a proven fact.

Speaking of things that work, we are all waiting for either of those 2 geniuses [no sarcasm, I really think they are both extremely smart] to actually produce something that actually works, instead of having their opinions on the BTS system and design...


And @ speedy the same is true for plan A... the bitUSD is discounted up  to 10% in plan B, it was discounted upto about 5-6% during plan A.