Author Topic: FDIC for BitUSD  (Read 26229 times)

0 Members and 1 Guest are viewing this topic.

Offline luckybit

  • Hero Member
  • *****
  • Posts: 2921
    • View Profile
  • BitShares: Luckybit
The above makes sense but what happens if we need to create more than $2 bil XTS? Although it might be necessary, this will be a dilution isn't it? Not sure I understand how we can avoid this but I believe we should avoid to create more than 2 bil at any time as this will create uncertainty about the future value and we can't afford that now.

Then you fall back to option 1 because in my opinion if you have to create more XTS than ever existed then the solution becomes worse than the problem.

It's like asking "well what if you need more credit than the total value that exists on the earth?"

If you create more than 2 billion you're still working with the same amount of value so basically we all lose value if you do that. The more XTS exists the more value we lose (yes I know you can hedge by using BitUSD but long term holders are important).

If you keep it at 2 billion then long term holders and true believers will maintain support. They will have bought in at 2 billion and aren't exactly losing anything if temporarily it goes back up to 2 billion. But once you start diluting it beyond the point they bought in at now you're operating backwards and for a business which is supposed to steer toward profitability it is catastrophic.

Keep the 2 billion cap as a constant and fall back to option 1 if option 3 reaches the 2 billion dollar cap. If the business is profitable like it should be then there will be plenty of XTS which could be recreated out of burned XTS. If you think you'd need more than 2 billion where will it end? Someone could make the argument that a volatility event could happen where we'd need 20 billion XTS to be created but that would destroy the business because it's too much debt to be operating in, not unlike the state of the United States deficit.
« Last Edit: August 12, 2014, 02:24:01 pm by luckybit »
https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline bytemaster

What if the BTSX gets created, but the creators account gets tagged in the event their collateral can't cover the shorts, and they have to pay an increased transaction fee for every transaction until the difference gets made up?

There are no ties to identity that could enforce this.  They just "walk away".
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 bytemaster

Quote
How likely is it to happen?

The event we are looking at here is called a "flash crash" and must result in a 50% fall in *real value* in less than an hour.   I would wager not even BTC has seen this particular event happen.    Lets assume that there are $1 million BitUSD created when the market cap was $100 Million and share supply was 2B.   That means there are $2 million in XTS backing it.   A 65% "flash crash" happens which means the collateral is only worth $700,000 and the market cap is now $35 Million.   An additional $300K of XTS would be created to raise capital to cover the loan.  This would represent a 0.85% dilution event.   

I would hope that transaction fees which will soon include market fees will be more than 0.8% per year. 

Now the question is what is the ratio of BitUSD (and other BitAssets) to market cap.   In this case it was 1%.  The MAXIMUM ratio is ~33% based upon current rules which means almost all XTS would be locked up as collateral and the remaining XTS held in reserve to buy back USD to cover the collateral.    Under this situation a 65% flash crash would result in a $8.5 million short fall which would mean a 25% dilution event. 

So the question remains is this.... are the shareholders entitled to "make profits no matter what"?   I think that shareholders are playing a risk-reward game where if the market peg holds and volatility is kept in check then they earn a very nice dividend.   If the delegates are reliable, avoid down time, etc shareholders profit.  However, for every profit opportunity comes a potential loss and in this case the primary cause of losses is a fall in XTS value by over 50%

There is one other type of event that is not addressed... if Gold and Silver go to the moon relative to XTS.... I would be very surprised to see their price double in just one hour, but it could happen. 

One last thing I would like to point out is that a rapid fall in price would not execute immediately because the trading range would hit.  The delegates would have to "confirm the fall" by lowering their feeds *or* the moving average would have to catch up.   This "time delay" protects against most short-term false-alarms.

As a shareholder, if you think XTS is over-valued you should buy BitUSD.  Holding BitUSD is supporting system as much as holding XTS.  You can then minimize your risk of dilution and/or benefit from the dilution if it were to happen.   I suspect you could even do this:

If you have 1% of the XTS and you own 1% of the BitUSD then you are fully hedged against dilution.   To keep numbers simple... lets assume the following:

100 XTS supply
You own 1 XTS
10% dilution event occurs...
10 new XTS are created, devaluing your 1 XTS to .9 XTS (original)
You get 1% of these 10 new XTS or .1 bringing your total to 1.1 XTS

Congrats, you just avoided all dilution with a slight hedge.  But it is better than that!
Your USD at least doubled in value relative to XTS.

So it is still possible for someone to "buy" protection against dilution.  However, I think looking at a single number "% of the system" you own as having tunnel vision and ignoring the bigger picture.   You are part of an economic system where all players (BitUSD and XTS holders) have a stake in the systems success.   Both parties are equal players and the risk/reward is always balanced between them.  Anything we do to help XTS holders hurts BitUSD holders.  Both parties benefit more from a solid BitUSD peg than from single-minded chase for deflation.

So I encourage everyone to look at both sides of the market as team players in a new economic system and every team player is responsible for allocating resources to minimize their own losses and thus the system grows in value for everyone.

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.

merockstar

  • Guest
What if the BTSX gets created, but the creators account gets tagged in the event their collateral can't cover the shorts, and they have to pay an increased transaction fee for every transaction until the difference gets made up?

Offline mf-tzo

  • Hero Member
  • *****
  • Posts: 1725
    • View Profile
Quote
Of course it's a loan. The profits in a DAC are the burned XTS. The burned XTS is what functions as our dividend but if you recreate new XTS you're taking it from the profits which were distributed via the burned XTS.

So what this means is that we should continue to promote a high enough burn rate to create a sort of buffer so that we have the profits stored up to bail ourselves out with a loan. When you create new XTS you're borrowing value from the previously burned XTS to pay off the short collateral. So you're using long term profits to pay off a short term debt is how I see it.

But once you took out the loan to create new XTS then if you increase the burn rate it's just as if you're buying back the XTS again. So everyone would be happy as far as I could tell and no one loses any value except temporarily. The only people who might be unhappy are short term speculator types but even they would be able to make some profit if they timed it to buy it during the event knowing the burn rate will be high.

This would be like buying stock knowing a massive stock buy back is about to happen. The stock buy back is how you repay the debt to shareholders for the dilution. Burning in our case is what functions as our stock buy back.

The above makes sense but what happens if we need to create more than $2 bil XTS? Although it might be necessary, this will be a dilution isn't it? Not sure I understand how we can avoid this but I believe we should avoid to create more than 2 bil at any time as this will create uncertainty about the future value and we can't afford that now.

Offline luckybit

  • Hero Member
  • *****
  • Posts: 2921
    • View Profile
  • BitShares: Luckybit
Makes sense. I agree on the fdic nomenclature. Fun but likely troublesome.
Would the fees diverted to fund the short collateral really be a loan? I thought it was just gone to wherever the rest of the collateral went.

Of course it's a loan. The profits in a DAC are the burned XTS. The burned XTS is what functions as our dividend but if you recreate new XTS you're taking it from the profits which were distributed via the burned XTS.

So what this means is that we should continue to promote a high enough burn rate to create a sort of buffer so that we have the profits stored up to bail ourselves out with a loan. When you create new XTS you're borrowing value from the previously burned XTS to pay off the short collateral. So you're using long term profits to pay off a short term debt is how I see it.

But once you took out the loan to create new XTS then if you increase the burn rate it's just as if you're buying back the XTS again. So everyone would be happy as far as I could tell and no one loses any value except temporarily. The only people who might be unhappy are short term speculator types but even they would be able to make some profit if they timed it to buy it during the event knowing the burn rate will be high.

This would be like buying stock knowing a massive stock buy back is about to happen. The stock buy back is how you repay the debt to shareholders for the dilution. Burning in our case is what functions as our stock buy back.

The auto-correction algorithm should look like this:

  • The cap on XTS should be a constant. It should not be possible to have more than 2 billion XTS.
  • If the volatility event occurs then recreate enough XTS to cover it but stop at the 2 billion cap so that the maximum XTS in circulation should always be equal to or less than 2 billion.
  • Create an network IOU from which the delegates must pay by burning 100% of their fees and temporarily raise fees immediately after the event.
  • As the loan gets paid down the fees and pay should return back to normal.

The key points are:
1. There should never be more than 2 billion XTS for any reason or Bitshares will be thought of as broken.
2. The loan should be an IOU which is just enough to cover the shorts or the volatility event and bring it all back into equilibrium.
3. Delegates pay a price for this in that none of them get paid as 100% of fees get burned while users also pay a price because fees are temporarily higher than usual.

This way the incentives for everyone is for this not to happen. Also include the ability to execute option 1 as a fallback mechanism.
« Last Edit: August 12, 2014, 02:09:30 pm by luckybit »
https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline Riverhead

Makes sense. I agree on the fdic nomenclature. Fun but likely troublesome.
Would the fees diverted to fund the short collateral really be a loan? I thought it was just gone to wherever the rest of the collateral went.

Offline luckybit

  • Hero Member
  • *****
  • Posts: 2921
    • View Profile
  • BitShares: Luckybit
1) How likely is this to happen on a large scale? Only in times of huge volatility?

2) Can a certain percentage of fees be set aside as an FDIC fund? The larger this fund grows the more robust the market and over time supply may not have to be messed with.

3) The delegates can adjust fees to replenish the fund or they can increase burn rate (which is at their own expense) to compensate if minting is needed. If this happens we need a mechanism for delegates to either lower their burn rate or only raise the burn rate for a short time. Currently it's a one way street.

How about calling it a volatility fund if you're going to treat it as a fund? Otherwise just treat it as a loan. I prefer treating it as a loan because you're borrowing value from burned XTS to lend to temporarily create new XTS. Once the event is resolved then the XTS has to be reburned to where it was before or even at a slight profit so that shareholders have nothing to fear from this process. It could be treated as a buying opportunity because you know the burn rate is going to be very high after these events.

We should avoid official language like FDIC or language which seems overtly statist being used the source code, the design, or the framing of language documenting how it works.

Basically when people hear about it you do not want them to get the frame of FDIC even if the FDIC serves a critical function. Volatility fund actually makes sense because it would specifically exist only for this purpose and that is if you need to have it exist at all. So in the code you would want to label it as a volatility protection algorithm or similar.

I agree that 1 and 2 are bad choices.

I would like a variant of 3 similar to what dxtr and luckybit propose.
When there are more than 2e9 BTSX around a certain part (some hardwired formula in the range of up to 10 % maybe?) of the delegates' pay should be channeled off and destroyed - until total outstanding is 2 billion again.
Internally we may think about it as a market correction tax or FDIC fund but we should refrain from using words like tax and FDIC. We should call it a market correction algorithm and volatility fund both in the official language and in the source code. On our forum discussions we can speak freely as the goal is to come up with the best possible algorithms we can.

But we do not want the negative associations or connotations which can come from bad communications especially when it's unnecessary to express the idea in that way.

The automatic correction algorithm still requires delegates to meet their obligation to the network. If there is a problem then delegates must increase the burn rate of the network as a way to pay for the losses incurred. If the cap is 2 billion XTS then there can never exist more than 2 billion XTS. That still doesn't mean that we want to see it trending toward 2 billion when it can be trending way from 2 billion.

We want to see it trending toward 1 billion XTS and then it is out of those profits that you could afford the fund. Those profits are basically contained as the burned XTS which could be recreated up to 2 billion XTS but if somehow that happens you have to increase the burn rate to bring it back to the baseline level. Ideally you want the result to equalize to being in a place better than where it was prior to the loan.

This way shareholders and the network itself all feel like over the long term no matter what happens the burn rate is going in their favor. The treadmill is going in a direction which favors long term holders so they'll tolerate the occasional creation of new XTS because it's not coming out of their pockets. The burn rate is increased for a while until after some weeks the total number of XTS is back where it was prior to the event or even a slight profit.





« Last Edit: August 12, 2014, 01:35:26 pm by luckybit »
https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline Riverhead

I agree that 1 and 2 are bad choices.

I would like a variant of 3 similar to what dxtr and luckybit propose.
When there are more than 2e9 BTSX around a certain part (some hardwired formula in the range of up to 10 % maybe?) of the delegates' pay should be channeled off and destroyed - until total outstanding is 2 billion again.


An automatic market correction tax? I like it.  That solves the need for delegates to volunteer to temporarily increase their burn rate. Really if this is needed the delegates aren't doing their job as well as they could be (fees weren't adjusted to prevent the FDIC from emptying out) so it makes sense that the correction should come out of their pay.

Offline Markus

  • Sr. Member
  • ****
  • Posts: 366
    • View Profile
I agree that 1 and 2 are bad choices.

I would like a variant of 3 similar to what dxtr and luckybit propose.
When there are more than 2e9 BTSX around a certain part (some hardwired formula in the range of up to 10 % maybe?) of the delegates' pay should be channeled off and destroyed - until total outstanding is 2 billion again.

Offline Rod

  • Newbie
  • *
  • Posts: 2
    • View Profile
3 is probably the best option, but why creating new tokens right away, when there are some unused "blockchain_accumulated_fees"?

If those fees are enough then no additional BTSX are created, so BTSX value stays untouched.
If there is more than fees can provide then option 3, and delegates work for free, but BTSX value is more stable.

I mean delegates are probably willing to sacrifice a days cut if this means everything the'yve earned so far doesn't loose value.

In case of new BTSX being created right away, BTSX price may fall faster, and value of everyone's (not only delegate's) stack might decrease compared to that before new BTSX creation.
+5% +5% +5% - Great idea

Offline Riverhead

1) How likely is this to happen on a large scale? Only in times of huge volatility?

2) Can a certain percentage of fees be set aside as an FDIC fund? The larger this fund grows the more robust the market and over time supply may not have to be messed with.

3) The delegates can adjust fees to replenish the fund or they can increase burn rate (which is at their own expense) to compensate if minting is needed. If this happens we need a mechanism for delegates to either lower their burn rate or only raise the burn rate for a short time. Currently it's a one way street.
« Last Edit: August 12, 2014, 09:38:45 am by Riverhead »

Offline xeroc

  • Board Moderator
  • Hero Member
  • *****
  • Posts: 12922
  • ChainSquad GmbH
    • View Profile
    • ChainSquad GmbH
  • BitShares: xeroc
  • GitHub: xeroc
3 is probably the best option, but why creating new tokens right away, when there are some unused "blockchain_accumulated_fees"?

If those fees are enough then no additional BTSX are created, so BTSX value stays untouched.
If there is more than fees can provide then option 3, and delegates work for free, but BTSX value is more stable.

I mean delegates are probably willing to sacrifice a days cut if this means everything the'yve earned so far doesn't loose value.

In case of new BTSX being created right away, BTSX price may fall faster, and value of everyone's (not only delegate's) stack might decrease compared to that before new BTSX creation.

 +5% +5% +5%

So far it seems delegates have flexibility.
1. They can take a pay cut and use that to provide the loan to secure the network.
2. They can increase the burn rate in the case of new XTS being created by option 3.
+5%

also this is MUCH better to market!!

Offline luckybit

  • Hero Member
  • *****
  • Posts: 2921
    • View Profile
  • BitShares: Luckybit
3 is probably the best option, but why creating new tokens right away, when there are some unused "blockchain_accumulated_fees"?

If those fees are enough then no additional BTSX are created, so BTSX value stays untouched.
If there is more than fees can provide then option 3, and delegates work for free, but BTSX value is more stable.

I mean delegates are probably willing to sacrifice a days cut if this means everything the'yve earned so far doesn't loose value.

In case of new BTSX being created right away, BTSX price may fall faster, and value of everyone's (not only delegate's) stack might decrease compared to that before new BTSX creation.

 +5% +5% +5%

So far it seems delegates have flexibility.
1. They can take a pay cut and use that to provide the loan to secure the network.
2. They can increase the burn rate in the case of new XTS being created by option 3.
« Last Edit: August 12, 2014, 08:22:04 am by luckybit »
https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline dxtr

  • Full Member
  • ***
  • Posts: 71
  • United We Stand, Dispersed We Are.
    • View Profile
3 is probably the best option, but why creating new tokens right away, when there are some unused "blockchain_accumulated_fees"?

If those fees are enough then no additional BTSX are created, so BTSX value stays untouched.
If there is more than fees can provide then option 3, and delegates work for free, but BTSX value is more stable.

I mean delegates are probably willing to sacrifice a days cut if this means everything the'yve earned so far doesn't loose value.

In case of new BTSX being created right away, BTSX price may fall faster, and value of everyone's (not only delegate's) stack might decrease compared to that before new BTSX creation.
Help me out:     wallet_approve_delegate mr.scroodge true