Author Topic: BitUSD Market Maker - Proposal for Discussion  (Read 34123 times)

0 Members and 1 Guest are viewing this topic.

Offline bytemaster

Given price feeds, I think mandatory covering at $0.90 on the BitUSD for which ever short position is least collateralized adds liquidity and protects the network from under-collateralized shorts.

$0.9, why not 0.89, why not 0.99? when you pull a number out of thin air, you are bound to fail.

I agree all "fixed" numbers are a form of "price fixing" and the market has a way of bending elsewhere.
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 yiminh

  • Full Member
  • ***
  • Posts: 66
    • View Profile
Given price feeds, I think mandatory covering at $0.90 on the BitUSD for which ever short position is least collateralized adds liquidity and protects the network from under-collateralized shorts.

$0.9, why not 0.89, why not 0.99? when you pull a number out of thin air, you are bound to fail.

Offline starspirit

  • Hero Member
  • *****
  • Posts: 948
  • Financial markets pro over 20 years
    • View Profile
  • BitShares: starspirit
The basic problem appears to be that without a mechanism for directly redeeming or exchanging BitUSD for real USD, I can't see how real arbitrage is possible, and there is no reason for tracking to occur in the absence of a crowd consensus that they should. Tell me if I'm wrong, but my understanding is that every bitAsset is essentially a replica, just with a different label on it. Change the label on BitUSD to just Bit, and whose to say why it should track realUSD, pretzels or anything at all? Yet the desire is to convince everyone that it should track its namesake, which relies on this idea of a specifically trained social consensus forming around each of these different labels. In light of this, its already a tremendous result that BitUSD tracks even as closely as it has. I think liquidity and usability is more critical at this stage than how well it tracks realUSD - just let it run for now without heavy handed intervention.



Offline bytemaster

I am against the market maker concept on the principle that it exposes the network to risks if a BitAsset crashes (ie: dollar hyperinflation).   
I think the feeds are not an issue given 100 sources and a median.  I think that eventually the market will make the feeds less and less necessary.
Given price feeds, I think mandatory covering at $0.90 on the BitUSD for which ever short position is least collateralized adds liquidity and protects the network from under-collateralized shorts.

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 Agent86

  • Sr. Member
  • ****
  • Posts: 471
  • BTSX: agent86
    • View Profile
I do not totally get what they think the price feeds will get them (positive thing that is) and personally prefer the average.
Average is the only way to go, price feed continuously is complicated, messy and will fail eventually
2) I also don't like the market relying on people to feed data into it all the time, I'd like a more natural price discovery.

I think you guys are thinking the price feed is a critical point of weakness when it isn't.  Everyone is responsible for placing their own orders at the price they want.  As long as everyone expects the feed to tend toward tracking the dollar in the long run, that's all that really matters.  If you had a market maker, than control over the feed could be a more critical problem but I oppose a market maker.

You can't use an average in place of the price feed because it just won't work and wishing it doesn't make it so.  The moving average basically just slows down market movement; it's a totally different thing.

Offline tonyk

  • Hero Member
  • *****
  • Posts: 3308
    • View Profile
4) we need more liquidity on https://bter.com/trade/btc_bitusd and https://bter.com/trade/bitusd_usd, that is the bottleneck to many arb bots at the moment. (disclosure: i've been working on a market making bot for that. I'm thinking of starting a "market making" delegate where all proceeds get feed into the bots to aid price discovery and peg)

Actually that is an idea that I really like, just the current delegate fees will make too small of a difference. But hey it is a start.
Lack of arbitrage is the problem, isn't it. And this 'should' solves it.

Offline maqifrnswa

  • Hero Member
  • *****
  • Posts: 661
    • View Profile
There are so many innovative concepts in bitsharesx, a built in market maker would be another one.

However, there are several points I want to make:

1) I like the idea of a built-in market maker as a way to improve liquidity, but I'm worried about the market getting too complex with too many "external" rules. External rules always add a "cost" and an inefficiency. There is a reason why the offset exists, and removing it might not be the best for the market. If the market maker exists within current rules (e.g., buying all shorts listed below spot), that would be ok.

2) I also don't like the market relying on people to feed data into it all the time, I'd like a more natural price discovery.

3a) We can learn from other similar BTC futures markets: If you played the futures BTC market on https://icbit.se/WebTrade/Trading.aspx, you'd know that BTC futures are almost always trading above spot. People are bullish on BTC, and believe that BTC should generally be in "contango." (they call it contango, but technically it's not really contango... but that's another story...). That offset is actually very similar to the current BitUSD market, and is due to perceived future strength of BTC relative to USD, the same way the discount of USD on bitsharesx is probably due to the perceived strength of BTSX to USD. When there is a crash, those shorts would be killed. Once that happened the first time, everyone learned their lesson and the market has been significantly closer to spot ever since. We may need a painful correction for everyone to learn to be careful.

3b) We can learn from other similar BTC futures markets: icbit.se uses variable fees as a way to make the market (adds liquidity to both asks and bids) https://icbit.se/discount. What if the fee to sell short  bitUSD is made much larger than sell/buy? People wouldn't be as willing to sell short far below spot and instead find other ways to buy bitUSD (exchanges), which brings me to...

4) we need more liquidity on https://bter.com/trade/btc_bitusd and https://bter.com/trade/bitusd_usd, that is the bottleneck to many arb bots at the moment. (disclosure: i've been working on a market making bot for that. I'm thinking of starting a "market making" delegate where all proceeds get feed into the bots to aid price discovery and peg)
« Last Edit: August 31, 2014, 05:25:43 pm by maqifrnswa »
maintains an Ubuntu PPA: https://launchpad.net/~showard314/+archive/ubuntu/bitshares [15% delegate] wallet_account_set_approval maqifrnswa true [50% delegate] wallet_account_set_approval delegate1.maqifrnswa true

Offline tonyk

  • Hero Member
  • *****
  • Posts: 3308
    • View Profile
I don't think the first proposal is bad and I will support it if we can not create a better solution, though I would prefer no market maker at first because again that introduces another layer which may not be necessary.

Ohh wait, the part about the market bot buying and printing money is nothing short of TERRIBLE, I already expressed my utter and complete disagreement with it in the first thread dedicated to the subject.
Lack of arbitrage is the problem, isn't it. And this 'should' solves it.

Offline Empirical1

  • Hero Member
  • *****
  • Posts: 884
    • View Profile
I don't think the first proposal is bad and I will support it if we can not create a better solution, though I would prefer no market maker at first because again that introduces another layer which may not be necessary.


Quote
price feeds  - Yes the proposal does require/includes those. I do not totally get what they think the price feeds will get them (positive thing that is) and personally prefer the average.

Price feeds require a greater degree of constant trust than a collateral level which can be changed with advanced warning.
(51% of the delegates agreeing to change the collateral level has a limited short term effect, whereas if some collude to alter the feeds, they can potentially create an 'event'? I don't know.)

Quote
'price fixing BitUSD creation above 1-1' - In my view 'No', it merely restricts new short position to prices >= 1:1...  ;)

I think you are trying to make the unelegant sound elegant :) Though I am personally happy with it if it is the best solution.

Quote
But you suggest market rule changes every other month, and the proposal suggests elegant market based alternative.

Yes I think you are correct here. This could require a change as frequently as every month for the first three months but then once every 2/3 months after as it slowly scales down to the 2X collateral level.

This is still only monitoring a total of 7 decisions with advance notice from the delegates over 18 months vs. trusting them with price feeds all the time for that period.

 

Offline yiminh

  • Full Member
  • ***
  • Posts: 66
    • View Profile

I.e. We find the collateral level (which may be very high) that makes the BitAsset mean/average within 2% of 1-1 in the next month, over time though as BitAsset demand increases the mean/average will rise and if it gets above 1-1, the collateral level could be lowered with advance warning by delegates ultimately settling on the current minimum 2x collateral.

..................

That proposal requires price feeds and price fixing BitUSD creation above 1-1.

price feeds  - Yes the proposal does require/includes those. I do not totally get what they think the price feeds will get them (positive thing that is) and personally prefer the average.


'price fixing BitUSD creation above 1-1' - In my view 'No', it merely restricts new short position to prices >= 1:1...  ;)


But you suggest market rule changes every other month, and the proposal suggests elegant market based alternative.

Average is the only way to go, price feed continuously is complicated, messy and will fail eventually

Offline tonyk

  • Hero Member
  • *****
  • Posts: 3308
    • View Profile

I.e. We find the collateral level (which may be very high) that makes the BitAsset mean/average within 2% of 1-1 in the next month, over time though as BitAsset demand increases the mean/average will rise and if it gets above 1-1, the collateral level could be lowered with advance warning by delegates ultimately settling on the current minimum 2x collateral.

..................

That proposal requires price feeds and price fixing BitUSD creation above 1-1.

price feeds  - Yes the proposal does require/includes those. I do not totally get what they think the price feeds will get them (positive thing that is) and personally prefer the average.


'price fixing BitUSD creation above 1-1' - In my view 'No', it merely restricts new short position to prices >= 1:1...  ;)


But you suggest market rule changes every other month, and the proposal suggests elegant market based alternative.

Lack of arbitrage is the problem, isn't it. And this 'should' solves it.

Offline Empirical1

  • Hero Member
  • *****
  • Posts: 884
    • View Profile
The percentage of people wanting to short on average is greater than the percentage of people willing to genuinely go long.

What about instead of price feeds first trying upping the collateral required?

A certain range of collateral required could make those two percentages exist (genuine shorts vs. longs) in a better equilibrium very close to the peg without price feeds?

The more collateral required the more secure the BitAsset system is too, which will also bring in more genuine longs.

I like this proposal... but because I am not an expert what are your thought's bytemaster regarding Empirical1 proposal?

I would add that it still requires some periodic action to be taken by the delegates.

I.e. We find the collateral level (which may be very high) that makes the BitAsset mean/average within 2% of 1-1 in the next month, over time though as BitAsset demand increases the mean/average will rise and if it gets above 1-1, the collateral level could be lowered with advance warning by delegates ultimately settling on the current minimum 2x collateral.

Did you guys read p7. of the proposal/OP?

That proposal requires price feeds and price fixing BitUSD creation above 1-1.


Offline tonyk

  • Hero Member
  • *****
  • Posts: 3308
    • View Profile
The percentage of people wanting to short on average is greater than the percentage of people willing to genuinely go long.

What about instead of price feeds first trying upping the collateral required?

A certain range of collateral required could make those two percentages exist (genuine shorts vs. longs) in a better equilibrium very close to the peg without price feeds?

The more collateral required the more secure the BitAsset system is too, which will also bring in more genuine longs.

I like this proposal... but because I am not an expert what are your thought's bytemaster regarding Empirical1 proposal?

I would add that it still requires some periodic action to be taken by the delegates.

I.e. We find the collateral level (which may be very high) that makes the BitAsset mean/average within 2% of 1-1 in the next month, over time though as BitAsset demand increases the mean/average will rise and if it gets above 1-1, the collateral level could be lowered with advance warning by delegates ultimately settling on the current minimum 2x collateral.

Did you guys read p7. of the proposal/OP?
Lack of arbitrage is the problem, isn't it. And this 'should' solves it.

Offline Empirical1

  • Hero Member
  • *****
  • Posts: 884
    • View Profile
The percentage of people wanting to short on average is greater than the percentage of people willing to genuinely go long.

What about instead of price feeds first trying upping the collateral required?

A certain range of collateral required could make those two percentages exist (genuine shorts vs. longs) in a better equilibrium very close to the peg without price feeds?

The more collateral required the more secure the BitAsset system is too, which will also bring in more genuine longs.

I like this proposal... but because I am not an expert what are your thought's bytemaster regarding Empirical1 proposal?

I would add that it still requires some periodic action to be taken by the delegates.

I.e. We find the collateral level (which may be very high) that makes the BitAsset mean/average within 2% of 1-1 in the next month, over time though as BitAsset demand increases the mean/average will rise and if it gets above 1-1, the collateral level could be lowered with advance warning by delegates ultimately settling on the current minimum 2x collateral.
« Last Edit: August 31, 2014, 03:23:04 pm by Empirical1 »

Offline liondani

  • Hero Member
  • *****
  • Posts: 3737
  • Inch by inch, play by play
    • View Profile
    • My detailed info
  • BitShares: liondani
  • GitHub: liondani
The percentage of people wanting to short on average is greater than the percentage of people willing to genuinely go long.

What about instead of price feeds first trying upping the collateral required?

A certain range of collateral required could make those two percentages exist (genuine shorts vs. longs) in a better equilibrium very close to the peg without price feeds?

The more collateral required the more secure the BitAsset system is too, which will also bring in more genuine longs.

I like this proposal... but because I am not an expert what are your thought's bytemaster regarding Empirical1 proposal?