BitAsset 2.0 Requirements & Implied Design

[merivercap]

I'll buy all bitUSD sold for less than $1 and settle them for $1.

Just for clarification I'm talking about a design without forced settlement. 

But without forced settlement people can buy bitUSD at a discount and then wait for bitUSD to go back to parity or even a premium.  People will hopefully do that to keep the market pegged so if it's a discount there will be more demand to buy and if it sells at a premium there will be more demand to sell.  People should trade towards parity. 

[xeroc]

I'll buy all bitUSD sold for less than $1 and settle them for $1.

[merivercap]

So in your design, will the BitUSD buyer ALWAYS pay a little more than $1, and ALWAYS have to sell for a little less than a $1?  In other words, will the bids always be under $1 and the offers always above $1 by design?

No there will be no ALWAYS.  Bids/asks and trades will just fluctuate around $1.  Sometimes it will be .90 and other times 1.1 .. Early on without liquidity we can have a range of .85 to 1.15.. as we get more liquidity spreads may tighten to keep the range between .9 to 1.1 and with even more liquidity between .95 and 1.05.     

Eventually the street price will dominate and people would probably trade 1 bitUSD for $1 locally or purchase bitUSD at $1.01 via ACH.  Local exchange and gateway purchases will dictate the pricing in the internal market such that the average prices for bitUSD might trade at a discount... maybe around .98/.99...

[merivercap]

I am not apposed to a policy of adjusting the force settlement price such that the average of all trades in a given day occur at around $1.00.   This would have to be a very slowly adjusted parameter (long term moving average of the premium).   

It would produce a currency that has the guarantee that on average it can be sold for $1.00.   Perhaps this is an opportunity to create a new asset called BUX that implements this policy.  Changing the policy on USD without sufficient warning could be detrimental.

I think this would be much better and we can always give people plenty of warning...  The only kind of forced settlement I think is somewhat acceptable is if it's at 90% such that we don't have too steep a discount, but most trades would be generally free to swing at a premium or discount.

I still think it's better to experiment more strongly and have one with and without force settlement.  We will learn less with incremental changes.  It's like when you do an A/B test you might find a local maxima, but unless you test bigger changes you won't find larger maxima. 

[tbone]

@merivercap: I've certainly appreciated having the chance to read all of your excellent input on this matter lately.  However,  I'm having trouble understanding why you think consumers would care so much whether BitUSD fluctuates around $1 or just at/above $1.

Thanks @tbone!   

Yeah every minor issue with most consumers can add significant friction.  I'm ok with not going after peer-to-peer transactions as a business because I know 4-20 cent fees are a significant enough psychological friction to prevent mass p2p transactions.  Hence we can go after merchants who can  cover the cost of transactions and the consumers won't see the small transaction fees. 

This leads me to the other friction for consumers of obtaining and using bitUSD.  Even a minor premium is psychological friction.  Every fee to get into bitUSD is friction.   We are already facing an uphill battle.   We don't need to give additional benefits to merchants.   They are willing to pay up to 1-2% more so giving them guarantees is not necessary. 

1.  Consumers naturally will think of 1 bitUSD as equivalent to 1 dollar.   Maybe around 20% of consumers might exchange it 1:1  despite a permanent premium out of laziness.  Many will start taking out their change purse and actually charge 1.05 or 1.10 or 1.15  for each bitUSD so people are going to be transacting with change.  That's extremely inconvenient.  I don't even want to do that.  We need most general exchanges to happen 1:1 or adoption will grind to a halt.  We're trying to create a new money and so even the tiniest issue will psychologically deter users.

2.  On top of the premium there will be fees to get in and out so the prices will be even higher especially in the beginning.  We're going to see bitUSD for 1.25 - 1.35.   (Note: BitQuick.co was selling bitUSD at 1.35 - 1.45 in Bitshares 1.0 near the tail end of the transition and during the bear market).  Those kind of premiums are non-starters for consumers.   So it's already hard for people that will pay 1-5% fees to get in, but a premium will make the market even more out of whack.  Adoption will grind further to a halt.

Minor friction in this space is a big deal and there's already too much friction already.  If we hope to have a shot at getting the mainstream or even early adopters to use this system as money we have to be careful at every little thing we do.   This premium is not a minor issue either. 

Imo, a peg fluctuating above 1:1 is as good as a peg fluctuating around 1:1 if deviation from the peg is small. If a premium was less than a penny, nobody would care.

Premiums in the beginning are going to be huge.  All the p2p exchanges are going to be selling bitUSD at 1.2+
Furthermore you will get far less liquidity if you don't know what the accepted value of bitUSD is.  The best way is to declare it to be 1:1 among consumers and traders.  That's it.  Simple.

Otherwise  if you only set a floor at 1 you will have an ever changing premium around 1.1, but the accepted value of bitUSD will be different for everyone .. some people will say 1.08 .. others will say 1.12..some will say they don't know.. and it will always fluctuate... So with the confusion in trading you'll get far less liquidity (and I'm not even including forced settlement in this conversation which will make liquidity even worse.)

So if I can guess what the outcome of two designs would be you will have one trading between .95 and 1.05 (spread of 10 cents) or another trading at 1.00 to 1.25 (spread of 25 cents) all things being equal and in non-bull markets.   

The floor could be lower... for example, if the floor is set to $0.97, then the worst case scenario is that it's still competitive w/ PayPal. If the average premium is around 3%, it means USD will float right around $1.00.

So don't guarantee $1.00 .. instead, offer $1.00 on average, but guarantee that one can ALWAYS get 97% of face value.

It's the best of both worlds, no?

So in your design, will the BitUSD buyer ALWAYS pay a little more than $1, and ALWAYS have to sell for a little less than a $1?  In other words, will the bids always be under $1 and the offers always above $1 by design?

[bytemaster]

I am not apposed to a policy of adjusting the force settlement price such that the average of all trades in a given day occur at around $1.00.   This would have to be a very slowly adjusted parameter (long term moving average of the premium).   

It would produce a currency that has the guarantee that on average it can be sold for $1.00.   Perhaps this is an opportunity to create a new asset called BUX that implements this policy.  Changing the policy on USD without sufficient warning could be detrimental.

[merivercap]

That's a little better, but we don't need a guarantee in the first place.  Markets fluctuate so if you want to get out on average you should be able to get out at $1.00 anyways...some trades you'll get $1.05 and others you'll get .95 and yet others you'll get $1... your average exit will still average $1 without needing a guarantee.

As USD is a "premium" product, such a guarantee sets us apart from the competition & we should take advantage of it, IMO.

I think it will confuse people and make things more inconvenient.   It will set us apart in a negative way and will most likely be a non-starter for consumers.  I personally don't believe in a bitUSD premium and as a company that depends on bitUSD it puts us in a dilemma.   I can't sell to people something I don't believe in.  Do we proceed hoping it will eventually change in the future or wait and advocate for an alternative.

There are those that want to experiment with the current design and I think that's fine.  However we should have an alternative version of bitUSD for those that strongly believe in an alternative design. 

[roadscape]

That's a little better, but we don't need a guarantee in the first place.  Markets fluctuate so if you want to get out on average you should be able to get out at $1.00 anyways...some trades you'll get $1.05 and others you'll get .95 and yet others you'll get $1... your average exit will still average $1 without needing a guarantee.

As USD is a "premium" product, such a guarantee sets us apart from the competition & we should take advantage of it, IMO.

[Xeldal]

The floor could be lower... for example, if the floor is set to $0.97, then the worst case scenario is that it's still competitive w/ PayPal. If the average premium is around 3%, it means USD will float right around $1.00.

So don't guarantee $1.00 .. instead, offer $1.00 on average, but guarantee that one can ALWAYS get AT LEAST 97% of face value.

It's the best of both worlds, no?

^^ This!

And @Xeldal figured out that we (read: the committee) can do this by tuning the force_settlement_offset_percent parameter of a bitasset

Thought it would "change the deal" of bitassets ..

I'm not sure if that was me , but I have brought up the general idea a few times.  Pulling from older quotes from this thread you find the same thing.

We can lower the feed price such that the average trading price is $1.00 and that has always been my stance.  Adjusting the feed will adjust the average trade price and the premium is always relative to the feed. 

It all comes down to a choice on how you want the BitUSD holder to *pay* for their liquidity.   Do they pay up front when they buy it at a premium, or do they pay on the back end when they sell it at a discount? 

There is NO WAY AROUND IT.   Liquidity has a price proportional to market depth and volatility. 

It is really quite simple, the BitUSD:USD price needs to be factored into the price feed, the feed should adjust in what ever way is necessary to maintain the BitUSD:USD price and otherwise ignore the USD:BTS price.     The USD:BTS price is volatile and will have a high spread at any reasonable volume.   

There is no need for the BitUSD:USD price to have a large spread and the difference in price between $1 and $100,000 of volume on the BitUSD:USD market can be much less than the same difference on the USD:BTS market.   

Therefore, forced-settlement doesn't matter so long as it exists it serves as a backdrop of liquidity that sets the floor on BitUSD : BTS.    The feed producers should simply move this floor until the BitUSD:USD market is trading at $1.00.

Like I have said 1000 times, with Privatized BitAssets there are enough variables to achieve almost everything except interest.   No settlement, small settlement, all settlement. High maintenance collateral, low collateral, etc.  Fixed parameters, Variable Parameters, and an infinite number of ways of calculating yield. 

The only reason not to buy BitUSD at more than 1.00 is the name.... call it a USD correlated asset and it works fine.

I honestly have no idea what the market needs or will prefer, I just know we have the tools in place to find it.

[xeroc]

The floor could be lower... for example, if the floor is set to $0.97, then the worst case scenario is that it's still competitive w/ PayPal. If the average premium is around 3%, it means USD will float right around $1.00.

So don't guarantee $1.00 .. instead, offer $1.00 on average, but guarantee that one can ALWAYS get AT LEAST 97% of face value.

It's the best of both worlds, no?

^^ This!

And @Xeldal figured out that we (read: the committee) can do this by tuning the force_settlement_offset_percent parameter of a bitasset

Thought it would "change the deal" of bitassets ..

[merivercap]

@merivercap: I've certainly appreciated having the chance to read all of your excellent input on this matter lately.  However,  I'm having trouble understanding why you think consumers would care so much whether BitUSD fluctuates around $1 or just at/above $1.

Thanks @tbone!   

Yeah every minor issue with most consumers can add significant friction.  I'm ok with not going after peer-to-peer transactions as a business because I know 4-20 cent fees are a significant enough psychological friction to prevent mass p2p transactions.  Hence we can go after merchants who can  cover the cost of transactions and the consumers won't see the small transaction fees. 

This leads me to the other friction for consumers of obtaining and using bitUSD.  Even a minor premium is psychological friction.  Every fee to get into bitUSD is friction.   We are already facing an uphill battle.   We don't need to give additional benefits to merchants.   They are willing to pay up to 1-2% more so giving them guarantees is not necessary. 

1.  Consumers naturally will think of 1 bitUSD as equivalent to 1 dollar.   Maybe around 20% of consumers might exchange it 1:1  despite a permanent premium out of laziness.  Many will start taking out their change purse and actually charge 1.05 or 1.10 or 1.15  for each bitUSD so people are going to be transacting with change.  That's extremely inconvenient.  I don't even want to do that.  We need most general exchanges to happen 1:1 or adoption will grind to a halt.  We're trying to create a new money and so even the tiniest issue will psychologically deter users.

2.  On top of the premium there will be fees to get in and out so the prices will be even higher especially in the beginning.  We're going to see bitUSD for 1.25 - 1.35.   (Note: BitQuick.co was selling bitUSD at 1.35 - 1.45 in Bitshares 1.0 near the tail end of the transition and during the bear market).  Those kind of premiums are non-starters for consumers.   So it's already hard for people that will pay 1-5% fees to get in, but a premium will make the market even more out of whack.  Adoption will grind further to a halt.

Minor friction in this space is a big deal and there's already too much friction already.  If we hope to have a shot at getting the mainstream or even early adopters to use this system as money we have to be careful at every little thing we do.   This premium is not a minor issue either. 

Imo, a peg fluctuating above 1:1 is as good as a peg fluctuating around 1:1 if deviation from the peg is small. If a premium was less than a penny, nobody would care.

Premiums in the beginning are going to be huge.  All the p2p exchanges are going to be selling bitUSD at 1.2+
Furthermore you will get far less liquidity if you don't know what the accepted value of bitUSD is.  The best way is to declare it to be 1:1 among consumers and traders.  That's it.  Simple.

Otherwise  if you only set a floor at 1 you will have an ever changing premium around 1.1, but the accepted value of bitUSD will be different for everyone .. some people will say 1.08 .. others will say 1.12..some will say they don't know.. and it will always fluctuate... So with the confusion in trading you'll get far less liquidity (and I'm not even including forced settlement in this conversation which will make liquidity even worse.)

So if I can guess what the outcome of two designs would be you will have one trading between .95 and 1.05 (spread of 10 cents) or another trading at 1.00 to 1.25 (spread of 25 cents) all things being equal and in non-bull markets.   

The floor could be lower... for example, if the floor is set to $0.97, then the worst case scenario is that it's still competitive w/ PayPal. If the average premium is around 3%, it means USD will float right around $1.00.

So don't guarantee $1.00 .. instead, offer $1.00 on average, but guarantee that one can ALWAYS get 97% of face value.

It's the best of both worlds, no?

That's a little better, but we don't need a guarantee in the first place.  Markets fluctuate so if you want to get out on average you should be able to get out at $1.00 anyways...some trades you'll get $1.05 and others you'll get .95 and yet others you'll get $1... your average exit will still average $1 without needing a guarantee. 

[roadscape]

The floor could be lower... for example, if the floor is set to $0.97, then the worst case scenario is that it's still competitive w/ PayPal. If the average premium is around 3%, it means USD will float right around $1.00.

So don't guarantee $1.00 .. instead, offer $1.00 on average, but guarantee that one can ALWAYS get 97% of face value.

It's the best of both worlds, no?

[yvv]

Imo, a peg fluctuating above 1:1 is as good as a peg fluctuating around 1:1 if deviation from the peg is small. If a premium was less than a penny, nobody would care.

[tbone]

Yeah I may have said before that a range of 1 - 1.2 was fine, but I always preferred a range of .9 to 1.1.

The local trade will be 1:1 so everything should fluctuate around 1 not 1.1.   I would call the current design bitUSD Plus and treat it like a separate coin and if merchants and others like the premium then that's fine there might be liquidity.   I have my doubts with consumers and think it will hinder liqudity and a regular bitUSD that fluctuates around 1 is by far my preference.  Hence I would like to see two Smartcoins if possible and I would like to promote a regular bitUSD that fluctuates around 1, not above 1.   

Note: I think the other motivation behind the floor is that in previous designs there were great discounts and premiums without liquidity so this seems to be a recompensation to enforce some kind of peg.  I would rather see a weak peg using the margin calls at the price feed as training wheels, not a permanent forced settlement feature.  Thanks.

@merivercap: I've certainly appreciated having the chance to read all of your excellent input on this matter lately.  However,  I'm having trouble understanding why you think consumers would care so much whether BitUSD fluctuates around $1 or just at/above $1. 

[alt]

if you have success design the bitUSD's price always more than 1.0
I'm happy to buy bitUSD at price 1.0 as much as posiblle
so all stupid person, please sell your bitUSD to me at price 1.0

1)  When I mean 'stable' I just meant generally pegged, but I'm using the word loosely so we may not really differ in opinion.  I don't really care that bitUSD fluctuates between $0.90 and $1.10 and I don't think bitUSD holders will care either, although adding the floor in the current design so prices fluctuate between $1.00 and $1.20 seems fine.

@Bytemаsteг
I remember from the beginning our target is peg at price 1.0 +/- 0.01
so everybody can be a market maker, sell at price more than 1.0, buy at price less than 1.0
and users can use bitUSD for payment just like 1.0 fiat USD

I want ask you if this is your new plan, peg at price from $1.0 to $1.2?
I want know how to use this bitUSD for payment, why I pay bitUSD to others when bitUSD's price always more than 1?
as a  gateway, what's the price to sell bitUSD, and what's the price to buy bitUSD?

It is always a good deal to buy BitUSD for $1.00
It is always a good idea to sell BitUSD for as much as possible.
As a gateway, I would have a buy wall at $1.00 and then sell whatever inventory I get on the market... this would be the safest bet, but will probably result in the Gateway not having much BitUSD.
As a merchant, you "buy" bitUSD with a dollar's worth of merchandise.

The market must discover the price premium, I obviously want it to be as close to $1.00 as possible, but I have no idea what kind of premium shorts will demand and it will change over time depending upon how many people are selling BitUSD, how many are shorting, and how many are buying.  Until the market discovers the equilibrium we will have to wait and see.  It would probably be helpful to have a graph of the premium relative to the feed over time.