Proposal for Having Alternate Smartcoin Designs

[yvv]

First, we need to understand the roll of the short seller.  The short seller is a lender that provides a service to the network by lending bts to borrowers in the form of smartcoins.  Lenders do not lend money for free so they will either charge interest or charge a premium.  Otherwise, the risk/reward doesn't work.  For example, if I lend 1 usd into existence at a rate of 300 bts/usd with 2x collateral of 600bts, then my risk of loss is 100% of outlay (with a 50% drop in price) while my potential gains are only 50% of outlay.  It just does not make sense to enter this agreement unless I charge a fee to compensate for the added risk.

The short seller is a borrower, not a lender. You borrow bitUSD to provide them to the market. Those who buy these bitUSD are the lenders.

If that is correct, then why do short sellers not PAY premium to smartcoin buyers?  Borrowers almost always pay either premium or interest to lenders.  By your logic, the short seller should pay premium to the buyer.

Short sellers should not pay any premiums, they are already fucked up enough for borrowing bitAssets. Why do you want them to be fucked up even more?

[Helikopterben]

I don't think we will be able to get smartcoins to trade at parity with the current design, which is perfectly fine because I think the current design is the best design, possibly with a few tweaks and feature adds. 

  I think removing forced settlement should enable prices to range around parity, just like Nubits. (Not making any comments on the rest of Nubits design, only that at least they are trying for 1:1)

Nubits trades 1:1 because market makers maintain a peg around parity with buy and sell walls.  Their sells have the same risk/reward profile as their buys.  This is possible when your assets aren't securely collateralized. 

First, we need to understand the roll of the short seller.  The short seller is a lender that provides a service to the network by lending bts to borrowers in the form of smartcoins.  Lenders do not lend money for free so they will either charge interest or charge a premium.  Otherwise, the risk/reward doesn't work.  For example, if I lend 1 usd into existence at a rate of 300 bts/usd with 2x collateral of 600bts, then my risk of loss is 100% of outlay (with a 50% drop in price) while my potential gains are only 50% of outlay.  It just does not make sense to enter this agreement unless I charge a fee to compensate for the added risk.

Not sure I follow you here.   The short gets leverage by shorting bitUSD.  I'm not sure why you are explaining it in terms of the short as lender.  The short actually creates a debt into the system so the short should be a borrower of 1 bitUSD in your example.  The short's BTS collateral is necessary to make sure the 1 bitUSD debt is paid back in full.  I assume the system adds the bitUSD buyer's BTS into the collateral pool such that both long and short have collateral for each bitUSD.

As I asked above, If the short seller is the borrower, then why is the short seller not paying the premium?  If I decide to buy a house with a loan, I go to the bank and ask them to tie up some of their money so I can buy the house.  I will then pay interest over time for this service.  Smartcoin buyers are asking short sellers to tie up their money to create the smartcoins.  The short seller will charge a premium for this service, hence the short seller is a lender.  Maybe lender doesn't describe the concept very well.  Its more like an options seller.  Its kind of difficult to find a traditional example that would describe the concept because smartcoins are designed differently because of the desire to preserve the currency function.  This is why I think lending smartcoins vs selling smartcoins will have much lower premiums.  Perhaps we should create a parallel lending market if possible and see if that helps bootstrap liquidity.  Im sure that would be costly though.

In the current design shorts put twice the collateral, but I don't think that's necessary and I assume can easily be changed.  If the short only had to put 300bts in the scenario and the long had to put 300bts you would have 600bts collateral when a trade is executed (200% collateral overall).  A loss of 50% of BTS against USD would trigger a black swan so to be safe the system can just trigger margin-calls at 75% of initial collateral.  That's very safe esp since price feeds update every hour.

That was part of the original design, but it is more efficient IMO for the short seller to create usd with his own collateral and then sell it, which is how it is done now.  The short seller can then take the proceeds from the sale and adjust collateral if he wants.  Perhaps an order type could be created where the creation of usd is contingent on the sale and the proceeds are automatically added to collateral.  I'm not sure this would make much of a difference though.

I just want to run one hypothetical example of the current system if a bitUSD trade is executed at the price feed:   
Feed price is: .00345606 USD/BTS  (289.3468 BTS/USD)
Collateral is 675 BTS (at 2.33x collateral against price feed) or $2.33.  A bitUSD buyer would add 289.34 more BTS of collateral to purchase the bitUSD.  Hence the total collateral in the system is 3.33x)
Call price is: .00259249 USD/BTS (385.7295 BTS/USD)

The shorter will be called at a 25% loss of BTS vs USD.  The shorter's loss in BTS will be 96.4 BTS (385.7 BTS - 289.3 BTS) or about 14.3% of the initial collateral.  The shorter will end up with 578.6 BTS (675BTS - 96.4BTS) worth $1.50. 

The shorter's total loss is $0.83 (or 36%) because of the leverage instead of $0.58 (or 25%) without leverage.

If we only required the shorter to hold 1x collateral of 289.3468 BTS (or $1.00), then in the same scenario the shorter would end up with 192.9 BTS (289.3BTS - 96.4BTS) or $0.50 cents. 

The total loss in that scenario is $0.50 (or 50%) because of the leverage instead of $0.25 (or 25%) without leverage.

Anyways I'm doing this late so hopefully my numbers are ok.  I assume the design & collateral system works as described above.  If so we should be able to reduce collateral for the shorts.

Looks good to me.

[Helikopterben]

First, we need to understand the roll of the short seller.  The short seller is a lender that provides a service to the network by lending bts to borrowers in the form of smartcoins.  Lenders do not lend money for free so they will either charge interest or charge a premium.  Otherwise, the risk/reward doesn't work.  For example, if I lend 1 usd into existence at a rate of 300 bts/usd with 2x collateral of 600bts, then my risk of loss is 100% of outlay (with a 50% drop in price) while my potential gains are only 50% of outlay.  It just does not make sense to enter this agreement unless I charge a fee to compensate for the added risk.

The short seller is a borrower, not a lender. You borrow bitUSD to provide them to the market. Those who buy these bitUSD are the lenders.

If that is correct, then why do short sellers not PAY premium to smartcoin buyers?  Borrowers almost always pay either premium or interest to lenders.  By your logic, the short seller should pay premium to the buyer.

[yvv]

First, we need to understand the roll of the short seller.  The short seller is a lender that provides a service to the network by lending bts to borrowers in the form of smartcoins.  Lenders do not lend money for free so they will either charge interest or charge a premium.  Otherwise, the risk/reward doesn't work.  For example, if I lend 1 usd into existence at a rate of 300 bts/usd with 2x collateral of 600bts, then my risk of loss is 100% of outlay (with a 50% drop in price) while my potential gains are only 50% of outlay.  It just does not make sense to enter this agreement unless I charge a fee to compensate for the added risk.

The short seller is a borrower, not a lender. You borrow bitUSD to provide them to the market. Those who buy these bitUSD are the lenders.

[merivercap]

I don't think we will be able to get smartcoins to trade at parity with the current design, which is perfectly fine because I think the current design is the best design, possibly with a few tweaks and feature adds. 

  I think removing forced settlement should enable prices to range around parity, just like Nubits. (Not making any comments on the rest of Nubits design, only that at least they are trying for 1:1)

First, we need to understand the roll of the short seller.  The short seller is a lender that provides a service to the network by lending bts to borrowers in the form of smartcoins.  Lenders do not lend money for free so they will either charge interest or charge a premium.  Otherwise, the risk/reward doesn't work.  For example, if I lend 1 usd into existence at a rate of 300 bts/usd with 2x collateral of 600bts, then my risk of loss is 100% of outlay (with a 50% drop in price) while my potential gains are only 50% of outlay.  It just does not make sense to enter this agreement unless I charge a fee to compensate for the added risk.

Not sure I follow you here.   The short gets leverage by shorting bitUSD.  I'm not sure why you are explaining it in terms of the short as lender.  The short actually creates a debt into the system so the short should be a borrower of 1 bitUSD in your example.  The short's BTS collateral is necessary to make sure the 1 bitUSD debt is paid back in full.  I assume the system adds the bitUSD buyer's BTS into the collateral pool such that both long and short have collateral for each bitUSD.

In the current design shorts put twice the collateral, but I don't think that's necessary and I assume can easily be changed.  If the short only had to put 300bts in the scenario and the long had to put 300bts you would have 600bts collateral when a trade is executed (200% collateral overall).  A loss of 50% of BTS against USD would trigger a black swan so to be safe the system can just trigger margin-calls at 75% of initial collateral.  That's very safe esp since price feeds update every hour.

I just want to run one hypothetical example of the current system if a bitUSD trade is executed at the price feed:   
Feed price is: .00345606 USD/BTS  (289.3468 BTS/USD)
Collateral is 675 BTS (at 2.33x collateral against price feed) or $2.33.  A bitUSD buyer would add 289.34 more BTS of collateral to purchase the bitUSD.  Hence the total collateral in the system is 3.33x)
Call price is: .00259249 USD/BTS (385.7295 BTS/USD)

The shorter will be called at a 25% loss of BTS vs USD.  The shorter's loss in BTS will be 96.4 BTS (385.7 BTS - 289.3 BTS) or about 14.3% of the initial collateral.  The shorter will end up with 578.6 BTS (675BTS - 96.4BTS) worth $1.50. 

The shorter's total loss is $0.83 (or 36%) because of the leverage instead of $0.58 (or 25%) without leverage.

If we only required the shorter to hold 1x collateral of 289.3468 BTS (or $1.00), then in the same scenario the shorter would end up with 192.9 BTS (289.3BTS - 96.4BTS) or $0.50 cents. 

The total loss in that scenario is $0.50 (or 50%) because of the leverage instead of $0.25 (or 25%) without leverage.

Anyways I'm doing this late so hopefully my numbers are ok.  I assume the design & collateral system works as described above.  If so we should be able to reduce collateral for the shorts. 

[tbone]

I think the real reason why I prefer to create another smart coin instead of CNY is not because of some unreasonable parameter
the importan reason is committee will never  response as fast as the market's change, they can't and shouldn't take care of all business
an alternate smartcoin can  have a committee which can focus on it's business,  and can get more resouses to suport it's business

Am I the only one who thinks this is an INCREDIBLY bad idea?

[Helikopterben]

I don't think we will be able to get smartcoins to trade at parity with the current design, which is perfectly fine because I think the current design is the best design, possibly with a few tweaks and feature adds.   

You are bullish on parity if the right tweaks are made or it's okay to not be at parity?

It's ok to not be at parity with the current design of premium being paid up front.  The only way to trade at parity is for the smartcoin buyer to pay interest payments to the smartcoin creator instead of paying a premium.  So you can either have premium or interest payments.  From what I understand, interest payments would be more difficult with a blockchain.

[Helikopterben]

I don't think we will be able to get smartcoins to trade at parity with the current design, which is perfectly fine because I think the current design is the best design, possibly with a few tweaks and feature adds. 

First, we need to understand the roll of the short seller.  The short seller is a lender that provides a service to the network by lending bts to borrowers in the form of smartcoins.  Lenders do not lend money for free so they will either charge interest or charge a premium.  Otherwise, the risk/reward doesn't work.  For example, if I lend 1 usd into existence at a rate of 300 bts/usd with 2x collateral of 600bts, then my risk of loss is 100% of outlay (with a 50% drop in price) while my potential gains are only 50% of outlay.  It just does not make sense to enter this agreement unless I charge a fee to compensate for the added risk.

So the question becomes:  Why would anyone pay $1.10 for $1.00?  First of all, I outline why this may be the case when trading physical commodities vs digital assets such as bitcoin, usd, ect.  Second of all, this is why I think we need some sort of leveraged trading, if possible.  Traditional options and futures markets are highly leveraged.  If I ask someone to deposit $1,100 and he gets credited in the system for $1,000, then he will be reluctant to do it and say that he would rather use a centralized exchange where he can deposit $1,000 and get credited for $1,000.  However, if I tell someone they can deposit $1,100 and get credited in the system for $2000, then he may be more willing to do it because leveraged traders are comfortable with paying a premium to borrow money.  In this situation, the trader would pay $100 to the lender to borrow $1,000 and put in $1,000 of his own money at 2x leverage.  Obviously he will have to be aware of margin call requirements and the loan may have to have a contract end date.  This would also work out well for the lender as he could earn a return on his bts and not be exposed to exchange rate risk as long as he maintains proper collateral.  For instance, if the smartcoin creator creates 1,000 usd and lends it out vs selling it, then all that is required is for the borrower to repay the loan in full, at which point the lender can adjust collateral back to 0 without having to buy any usd on the open market.  Because of this, I imagine premiums would be much lower to borrow smartcoins than to buy smartcoins. 

I agree that forced settlement is questionable as to whether it is needed or not, but I don't think removing it will affect premium that much.  As someone else suggested, maybe parameters can be changed for cny vs usd vs euro to determine which design works best.  I think it would get confusing to have multiple designs of usd and I'm not sure how beneficial it would be, but it is definitely worth a shot at this stage. 

[merivercap]

What I hope to see is more detailed explanation how and why your system would work. Saying just "should be enough" is not very convincing for people like me who haven't fully grasped these things.

But I somewhat like the idea of MPA that floats around the peg and not over. Could we test it first on a different asset than fiat-MPA? Maybe gold or oil? For now I support fiat-MPAs that are pegged at least for the face value (although I might change my mind later if I see that's not working) because it guarantees the value for users. This is propably a good thing for commerce. But maybe assets like gold do not need such guarantees and we could let them to float around face value?

Ok.  Yeah I might find time to write something up and model it like a white paper.  Most of the current design seems to be fine so I can explain the current system as well.   I'll leave out the forced-settlement/guarantees/'at-least' face-value part.  I'll just explain how and why the system without forced settlement should work.

[merivercap]

I agree. I'd like to see a design for parity where risk is equal on each side of the trade. This is made difficult because of the need for fungible longs.

To keep this simple, let's start with a design without fungable longs and go from there.

BTS/USD market:

* Longs and shorts trade on margin in a CFD style market
* The only currency in use is BTS (different to the current design where bitUSD is also used)
* Long BTS orders post BTS collateral
* Short BTS orders also post BTS collateral
* A matching long BTS order with a short BTS order begins the contract
* Both sides need to maintain their positions in order to avoid getting margin called
* Both sides start in BTS and end up in BTS
* Either side can close their position at any time, or get margin called in which case that contract is closed
* The other party which didn't initiate the close (directly or indirectly) gets rematched with next eligible order on the market

Thoughts?

Your design seems fine in general and will be good to try out if the current design doesn't work, but I'm actually optimistic about 95% of the current design.   One huge benefit to the current design is that you don't need to have people close positions as often or ever if you're long the bitAsset and this is especially important if you want to maintain consistent circulation of bitAssets to be used as money.   I think having a stable balance of long term bitAsset holders and long-term bitAsset shorts is incredibly important. 

I am proposing adjusting parameters first before doing a more extensive re-design.  Really with the recent fork update, the only remaining major issue I have is with forced settlement.  It forced settlement was set at 90 cents and used as a emergency safety valve instead of the way it is currently I wouldn't be as concerned. 

BTW I remember you brought up something about the difference between the long's and short's collateral and I agree.  I think the short is currently putting 200% collateral and when the bitAsset long purchases the bitAsset together there is 300% collateral at the outset.  Hence I don't think you need to have the short put up 200% collateral, only 100%.  Hence the long and short would put up 200% BTS collateral when the trade happens.  Liquidity is directly proportional to leverage and we should loosen up the requirements of the short.  Furthermore the price feed updates hourly so we should be safe because a 50% draw down is unlikely to happen within one hour to cause a black swan.  Let me know if you're seeing the same thing.   That may be the other adjustment I would propose. 

[fuzzy]

I think the real reason why I prefer to create another smart coin instead of CNY is not because of some unreasonable parameter
the importan reason is committee will never  response as fast as the market's change, they can't and shouldn't take care of all business
an alternate smartcoin can  have a committee which can focus on it's business,  and can get more resouses to suport it's business

Multiple committees for various business functions makes sense.  It is actually kind of what it sounds like identabit is doing.

[monsterer]

I would have a set daily allowance where BTS provides liquidity by selling at 1.01 and buy back at 0.99. (Then once that's breached a further daily allowance at 0.98, 0.97) Ideally the blockchain would be break-even to profitable long term providing this service. These limits and ranges can be tweaked by the committee based on results.

Fairly sure this cannot work, because you cannot enforce the peg inside the dex.

edit: excepting through IOU markets - BTS / IOU USD on the DEX would enforce the peg

[xeroc]

What customers want is liquidity within a tight range around the peg.

Nubits is able to offer this because they are operating on fractional reserve & subsidizing liquidity providers.

Uphold (formerly BitReserve) is able to offer this and do it for free because they are investing the reserve in yield-bearing instruments.
https://support.uphold.com/hc/en-us/articles/205290895-How-does-Uphold-make-money-

The name of the game is providing liquidity in a tight range around the peg. 

Personally, considering BTS already has dilution, I would use the blockchain as a market maker for three key market CNY, USD and GOLD.
I would have a set daily allowance where BTS provides liquidity by selling at 1.01 and buy back at 0.99. (Then once that's breached a further daily allowance at 0.98, 0.97) Ideally the blockchain would be break-even to profitable long term providing this service. These limits and ranges can be tweaked by the committee based on results.

Committee could use accumulated fees for market making as well. Atm therd are $6900 worth in USD alone.

I'll try to talk to the committe members and see if we can figure something out

[Empirical1.2]

What customers want is liquidity within a tight range around the peg.

Nubits is able to offer this because they are operating on fractional reserve & subsidizing liquidity providers.

Uphold (formerly BitReserve) is able to offer this and do it for free because they are investing the reserve in yield-bearing instruments.
https://support.uphold.com/hc/en-us/articles/205290895-How-does-Uphold-make-money-

The name of the game is providing liquidity in a tight range around the peg. 

Personally, considering BTS already has dilution, I would use the blockchain as a market maker for three key market CNY, USD and GOLD.
I would have a set daily allowance where BTS provides liquidity by selling at 1.01 and buy back at 0.99. (Then once that's breached a further daily allowance at 0.98, 0.97) Ideally the blockchain would be break-even to profitable long term providing this service. These limits and ranges can be tweaked by the committee based on results.

[yvv]

The idea to play with different designs of smart assets is absolutely reasonable. You never know which concept is the best unless you test all of them. Just don't screw up current bitAssets, because they sort of work. Introduce new types of assets, and let's see how they perform.