Author [EN] [ZH] [ES] [PT] [IT] [DE] [FR] [NL] [TR] [SR] [AR] [RU] [EN] [ZH] [ES] [PT] [IT] [DE] [FR] [NL] [TR] [SR] [AR] [RU] [EN] [ZH] [ES] [PT] [IT] [DE] [FR] [NL] [TR] [SR] [AR] [RU] Topic: drltc's interest rate and BitBond proposal  (Read 659 times)

0 Members and 1 Guest are viewing this topic.

Offline theoretical

drltc's interest rate and BitBond proposal
« on: September 06, 2014, 06:22:58 AM »


I've put my response to bytemaster's latest interest proposal in a whitepaper on Github.  Available at https://github.com/drltc/bitbond-proposal/blob/master/bitbond.md

Basically the conclusions are:

- bytemaster's proposal will sink BitUSD decentralization by making BitUSD non-fungible
- Fixed interest rate on BitUSD is important for marketing reasons
- A mechanism to reverse BitUSD destruction and return BitUSD-denominated fees to circulation is necessary to prevent a margin crisis if we get into a situation where too many BitUSD have been destroyed
- Market-determined variable interest rate (bond market) is necessary to absorb randomness of fees
- Short interest is necessary to guarantee fixed interest rate
- Implementing a bond market is not too hard in practice

To this I will add in a future revision:

- A highly USD-correlated, cryptocurrency-based bond market is a "killer app" that will drive (currently anemic) BitUSD demand through the roof, and BTSX valuation along with it.
BTS- theoretical / PTS- PZxpdC8RqWsdU3pVJeobZY7JFKVPfNpy5z / BTC- 1NfGejohzoVGffAD1CnCRgo9vApjCU2viY / the delegate formerly known as drltc / Nothing said on these forums is intended to be legally binding / All opinions are my own unless otherwise noted / Take action due to my posts at your own risk

Offline arhag

  • Hero Member
  • *****
  • Posts: 1213
    • View Profile
    • My posts on Steem
  • BTS: arhag
  • GitHub: arhag
Re: drltc's interest rate and BitBond proposal
« Reply #1 on: September 06, 2014, 07:17:25 AM »
There is a lot I agree with here. But I am not sure why the complicated BitBonds are necessary. Why not just give variable interest to the BitUSD holders (mostly coming from payments made by the shorts) and be done with it? I have to still read your BitBond sections in more detail. But if you are worried about fungibility issues with the bond market proposal that Agent86 was suggesting (I was) then I have to ask what you think about my idea here: https://bitsharestalk.org/index.php?topic=8396.msg109907#msg109907.

- bytemaster's proposal will sink BitUSD decentralization by making BitUSD non-fungible

I don't like bytemaster's proposal either because of the weird way the rewards are calculated. It encourages holders to hold BitUSD for longer lengths of time to get larger proportional rewards: see my analysis here. I think the rewards should be structured so that it doesn't matter how many times you transfer money from one of your accounts to another (ignoring transaction fees) you still get the same reward. Basically what you mention here:
Quote
The interest on BitUSD has no "lock-in" period; as soon as you receive a BitUSD transfer, it starts accruing interest, which is compounded every block (implemented by exponentiation of the per-block interest rate). This hopefully reduces people trying to "game" the system and gets rid of the incentive to form "banks."

I now realize that my proposal needs some tweaks to meet that goal. I shouldn't be calculated ACCUMULATED_INTEREST by a simple sum but rather by compounding the interest after every block; and the interest for each block should be calculated using a TOTAL_USD that includes the added interest allocated to BitUSD holders in the previous blocks even if they haven't been claimed yet. I think that should essentially implement a similar thing to "exponentiation of the per-block interest rate" as you mentioned.

- Fixed interest rate on BitUSD is important for marketing reasons
...
- Short interest is necessary to guarantee fixed interest rate

I think the interest rate needs to be variable. In my proposal the shorts pay the money that goes to the interest fund, but the payment levels are chosen by short sellers to respond to market forces (balancing short and long demand). We can't guarantee any fixed arbitrary interest rate forever. At best you can put a cap on it and redirect excess interest for other purposes which may be what you are proposing in your system. But I don't think BitBonds are necessary, so then the only other place the excess interest can go is to pad the insurance fund or converting it to BTSX and burning it as a dividend to shareholders.

- A mechanism to reverse BitUSD destruction and return BitUSD-denominated fees to circulation is necessary to prevent a margin crisis if we get into a situation where too many BitUSD have been destroyed

One other thing I didn't like in bytemaster's proposal was that he seemed to consider the interest to BitUSD as a substitute to the insurance fund. But someone who purchases some BitUSD right before a black swan event is still going to be damaged by it if they need to spend the money shortly afterward (the accumulated interest could only be considered a replacement of the lost value of BitUSD after some period of time of holding the BitUSD). So, I think it is important to keep a reasonable amount of BitUSD in an insurance fund (scaling with the amount of BitUSD issued).

But I don't like the way you say "reverse BitUSD destruction." I generally agree with the following idea:
Quote
A fund below its minimum capitalization level has first priority for income, and gradually takes a smaller fraction of the available income until it claims zero income at its maximum capitalization level. A fund above its maximum capitalization level will have negative income, actually paying its balance over time to one or both of the other funds until it is again at or below its maximum capitalization level.
But only applied to the insurance fund. I think that for each block a certain amount of BitUSD should be pulled out from that insurance fund to cover the interest added to all BitUSD balances during that block. That BitUSD is considered out of the possession of the DAC and cannot be reclaimed. If the fund levels in the insurance fund get too low, the interest rates should be lowered to build up the balance in the fund.
« Last Edit: September 06, 2014, 07:40:44 AM by arhag »

Offline theoretical

Re: drltc's interest rate and BitBond proposal
« Reply #2 on: September 06, 2014, 04:55:44 PM »
There is a lot I agree with here. But I am not sure why the complicated BitBonds are necessary. Why not just give variable interest to the BitUSD holders (mostly coming from payments made by the shorts) and be done with it?

To make the interest-bearing products simpler and more transparent to investors.  When the yield of an interest-bearing instrument is unpredictable, investors may be wary.

When marketing to BitUSD holders, saying "your BitUSD will always earn a return of 5% per year" is much more compelling than saying "your BitUSD will always earn a return, sometimes more, sometimes less, depending on the phase of the moon and the price of tea in China" (which is basically what financially un-sophisticated consumers will hear if we try to explain the factors that go into the variable interest rate).

Also, when new BitUSD are printed by shorting, the short sellers and buyers must come to an agreement about the valuation -- basically how much BitUSD should trade above the peg to account for the fact that it earns interest.  This is much easier for both sides to compute when the interest amount is known and fixed.

So basically we want to keep the market liquid and the pricing transparent by telling everyone up front exactly how much their interest will be.

BitBond interest in my scheme is *also* transparent in this sense -- a particular BitBond holder knows the exact cost and redemption schedule of a BitBond when they buy, thus they know their personal interest rate ahead of time.  We need to charge shorts interest to make sure BitUSD interest remains solvent, and raise capital for macro-economic purposes like insuring underwater shorts.  That short interest should be fixed for the same reasons of simplicity and transparency that the long interest should be fixed.

But sometimes the short interest will raise too much, so we need to get rid of the leftover capital somehow or risk a margin crisis (because the BitUSD needed to redeem shorts has been permanently removed from circulation).  The BitBond market basically lets people compete against each other on how many BitUSD they're willing to temporarily burn in exchange for eventually receiving a portion of that over-capitalization.
BTS- theoretical / PTS- PZxpdC8RqWsdU3pVJeobZY7JFKVPfNpy5z / BTC- 1NfGejohzoVGffAD1CnCRgo9vApjCU2viY / the delegate formerly known as drltc / Nothing said on these forums is intended to be legally binding / All opinions are my own unless otherwise noted / Take action due to my posts at your own risk

Offline theoretical

Re: drltc's interest rate and BitBond proposal
« Reply #3 on: September 06, 2014, 05:37:26 PM »
I think the interest rate needs to be variable...We can't guarantee any fixed arbitrary interest rate forever.

This is a great criticism, because it's intuitive that we can't create arbitrary amounts of real-USD value -- and ultimately wrong, because real USD itself can be inflationary.  I think you're basically saying that I'm implementing BitUSD interest by taking money from shorts and giving it to longs, and eventually the shorts will run out of money [1].

In that scenario some shorts will be margin called, taking BitUSD out of circulation and reducing the interest.  If no new shorts are willing to enter the market at the current BitUSD exchange rate, but there is demand for BitUSD, then the (USD-denominated) value of BTSX will rise until supply and demand match.  The higher BTSX price will profit short-holders, stopping the margin calls.  It will also allow BTSX holders to short more BitUSD into existence while reducing the amount of BTSX that goes to interest payments.  This BTSX rise is basically a capital gain for the whole BitShares ecosystem, which ultimately pays for the interest and returns any additional profit to BTSX holders.

If you're thinking of the extreme of "but then going long BitUSD for 1000 years means...", keep in mind that when the BitUSD economy gets big enough, and BitUSD serve as a substitute for real USD, we've essentially hijacked monetary policy from the real-life Federal Reserve by printing our own USD equivalent.  In the long-term, the interest will ultimately be paid for by inflation of *all* USD!  But of course this effect is negligible as long as BitUSD market cap is a small fraction of real USD market cap.

[1] Unless we implement some variable interest scheme, i.e. making those interest payments proportional to the amount of money the shorts have left.
« Last Edit: September 06, 2014, 05:54:49 PM by drltc »
BTS- theoretical / PTS- PZxpdC8RqWsdU3pVJeobZY7JFKVPfNpy5z / BTC- 1NfGejohzoVGffAD1CnCRgo9vApjCU2viY / the delegate formerly known as drltc / Nothing said on these forums is intended to be legally binding / All opinions are my own unless otherwise noted / Take action due to my posts at your own risk

Offline arhag

  • Hero Member
  • *****
  • Posts: 1213
    • View Profile
    • My posts on Steem
  • BTS: arhag
  • GitHub: arhag
Re: drltc's interest rate and BitBond proposal
« Reply #4 on: September 06, 2014, 07:48:25 PM »
I remain unconvinced with this reasoning.

I am going to take BitUSD5, which is BitUSD with an continuous compounding annual 5% growth rate, as an example (which is essentially the same thing as BitUSD where the shorts pay the longs a fixed 5% interest rate). If there are not enough shorts to meet demand (because people do not expect BTSX value to grow faster than 5% in terms of USD for example), then BitUSD5 holders would be willing to pay a premium (meaning more BTSX) to get BitUSD5. This means the value of BitUSD5 in terms of BTSX goes up, potentially forcing margin calls on shorts. If there are no new shorts entering the system (because they still do not expect BTSX value to grow so quickly), then the problem continues to get worse, potentially leading to the black swan event: under-collateralized BitUSD5. If that happens, that is how you will finally get variable interest. People will know there is not enough BTSX value backing the issued BitUSD5 to provide the promised value (under the assumption of 5% interest). Instead the market will value BitUSD5 below its peg, effectively becoming a BitUSD with a variable <5% interest.

Offline theoretical

Re: drltc's interest rate and BitBond proposal
« Reply #5 on: September 06, 2014, 08:52:43 PM »
If there are no new shorts entering the system (because they still do not expect BTSX value to grow so quickly), then the problem continues to get worse, potentially leading to the black swan event: under-collateralized BitUSD5.

You assume all shorters want to make long-term leveraged bets that BTSX will grow.  This is not the case, in cases of short-term price movements you may get short-term shorters that bet it will revert to the peg.

In other words, yes, shorters are betting that BTSX market cap will grow.  But if BitUSD has recently rapidly risen in price, then this means BTSX market cap (measured in BitUSD) to shrink.  And shorters can bet that the numerical shrinkage doesn't reflect any economic reality, and BTSX valuation will spring back up.

To some extent your scenario is the Achilles heel of any Bitshares-like system where you have limited-issue cryptocoins that are used to somehow collateralize unlimited-issue BitAssets:  If the nominal value of the cryptocoins shrinks to such an extreme extent that the asset collateralization becomes wildly insufficient, then the system is hosed, end of story.  We try to mitigate the scenario by leverage limits, margin calls, and the black swan fund, but you can always have price movements so extreme that those measures are exhausted and the short side becomes insolvent.  So if we implement a proposal which *isn't* theoretically capable of a black swan scenario, the resulting system would have to effectively be something that is completely different from BitShares.
BTS- theoretical / PTS- PZxpdC8RqWsdU3pVJeobZY7JFKVPfNpy5z / BTC- 1NfGejohzoVGffAD1CnCRgo9vApjCU2viY / the delegate formerly known as drltc / Nothing said on these forums is intended to be legally binding / All opinions are my own unless otherwise noted / Take action due to my posts at your own risk

Offline arhag

  • Hero Member
  • *****
  • Posts: 1213
    • View Profile
    • My posts on Steem
  • BTS: arhag
  • GitHub: arhag
Re: drltc's interest rate and BitBond proposal
« Reply #6 on: September 06, 2014, 10:21:07 PM »
I want to first make sure I am not misunderstanding you. You are not saying that BTSX value will always grow fast enough to back the collateral of any BitAsset we can come up with, correct? We can precisely define all kinds of crazy BitAssets in which the real world value of the peg would increase (according to the formula) way too quickly. It is impossible to expect the real world value of BTSX collateral backing it to always grow along with it to fully back it. Therefore, the result is partially-backed BitAssets that will be trading at a value below the peg.

If the BitAsset peg is growing at a unreasonable rate, why would new shorts enter? The BTSX value would not be able to grow fast enough in the medium-term to catch up with and surpass the peg; thus, the shorts would just be putting up margin to be wiped out. If new shorts dry up and BTSX real world value cannot increase fast enough to keep up with the peg, the BitAssets will soon enough not be backed by enough BTSX value as collateral to maintain their peg value. Either all BitAsset holders trade for BTSX (covering all existing shorts before they lack enough collateral and thus causing the BitAsset to disappear) in which case those BitAsset holders do not get to realize all the future gains that the BitAsset promised, or the BitAsset holders who stubbornly continue to hold will soon reach a point where there is not enough BTSX collateral backing their BitAssets and thus the BitAsset will trade below the peg anyway. There is no free lunch. The growth in value that BitAssets can realize is necessarily limited by the growth in real world value of BTSX.

So my concern is that we might price a fixed interest rate on BitUSD a bit too high. If the demand for that BitAsset is too high and it is a time when for whatever reason BTSX value is not expected to grow (or grow fast enough) for a long-enough period of time, then there will not be enough new shorts (at the mandated currently too high fixed interest rate) to meet the demand. The high demand for BitAssets but lack of shorts could lead to a higher likelihood of the undesirable black swan event. On the other hand, if we were to allow the interest rate to be dynamic, then during times like this new shorts would enter (and get matched) with low promised reward payments to BitAsset holders. Also, the lower rewards would result in lower interest rates on the BitAssets thus reducing the demand for the BitAssets. This is a proper negative feedback loop that will result in balancing the supply and demand and reduce the likelihood of partially-collateralized BitAssets.

Now, maybe you think that at least a 5% interest rate is very reasonable to expect without risk of the black swan events and having the BitAsset below the peg, based on your expectations of BTSX growth. If that was the case, we could offer exactly a 5% interest rate to BitUSD holders (assuming black swans don't happen which we cannot really ever guarantee). The rest of the capital gains in BTSX could be enjoyed by the shorts or perhaps shared with others through a variable-interest BitBond for example. Or, if we were to instead share those gains with the BitUSD holders directly as I would like to do with my proposal, it would result in a variable interest rate on BitUSD but with a "minimum" of 5%. I say "minimum" because it could be less if black swans happen. But if they don't then BitUSD holders could be promised at least 5% interest. If you want to provide exactly 5% for simplicity, then you can just have a variable interest rate on BitUSD with a cap of 5% (excess could go to insurance fund or shareholders). If BTSX growth really ends up being as awesome as we hope, then the variable interest will always be above 5% and thus from the regular person's perspective they consistently see a 5% interest on their BitUSD.


Offline theoretical

Re: drltc's interest rate and BitBond proposal
« Reply #7 on: September 06, 2014, 11:49:38 PM »

Now, maybe you think that at least a 5% interest rate is very reasonable to expect without risk of the black swan events and having the BitAsset below the peg, based on your expectations of BTSX growth. If that was the case, we could offer exactly a 5% interest rate to BitUSD holders (assuming black swans don't happen which we cannot really ever guarantee).

This is exactly what I'm saying.  If the interest rate is too high, we may have some fundamental problems.  Even 5% may be too high once the market is large enough.  But bytemaster computed somewhere that APY so far from fees alone is 10%.  Add short interest into the mix and 5% should be very reasonable for the next several years at least.  My proposal also includes "buffering" some amount of pending interest payments in case there are some underwater shorts that will be unable to pay.  So if 5% is ever too high, even in a terrible black swan the buffer will pay longs 5% for months.  In those months we can hopefully work out a community consensus for a hard-fork to fix whatever the actual problem is.  Or if no hard-fork solution finds consensus, at least the long-holders will have plenty of warning that their interest rates are going to drop.

The rest of the capital gains in BTSX could be enjoyed by the shorts or perhaps shared with others through a variable-interest BitBond for example. Or, if we were to instead share those gains with the BitUSD holders directly as I would like to do with my proposal, it would result in a variable interest rate on BitUSD but with a "minimum" of 5%.

I really think a fixed interest rate is simpler to market and more attractive for investors.  And I really want to have a way to reward investors for making a long term commitment.  My BitBond proposal accomplishes both of those goals.  BitBonds as I propose them are actually fixed-rate investments; each bond purchase has a clearly defined APY based on the purchase price (which the purchaser obviously knows in advance).  (The term "variable-rate" is usually applied to financial products that can change their interest rate after the investment is made, which my BitBonds cannot do.)
BTS- theoretical / PTS- PZxpdC8RqWsdU3pVJeobZY7JFKVPfNpy5z / BTC- 1NfGejohzoVGffAD1CnCRgo9vApjCU2viY / the delegate formerly known as drltc / Nothing said on these forums is intended to be legally binding / All opinions are my own unless otherwise noted / Take action due to my posts at your own risk

Offline arhag

  • Hero Member
  • *****
  • Posts: 1213
    • View Profile
    • My posts on Steem
  • BTS: arhag
  • GitHub: arhag
Re: drltc's interest rate and BitBond proposal
« Reply #8 on: September 07, 2014, 12:52:54 AM »
This is exactly what I'm saying.  If the interest rate is too high, we may have some fundamental problems.  Even 5% may be too high once the market is large enough.  But bytemaster computed somewhere that APY so far from fees alone is 10%.  Add short interest into the mix and 5% should be very reasonable for the next several years at least. My proposal also includes "buffering" some amount of pending interest payments in case there are some underwater shorts that will be unable to pay.  So if 5% is ever too high, even in a terrible black swan the buffer will pay longs 5% for months.  In those months we can hopefully work out a community consensus for a hard-fork to fix whatever the actual problem is.  Or if no hard-fork solution finds consensus, at least the long-holders will have plenty of warning that their interest rates are going to drop.

I'm glad we are finally reaching an understanding. My issue is I really don't like price fixing and would prefer a design that can last. I rather avoid requiring some other consensus mechanism to constantly tweak parameters if there is another market-based solution that can take care of it. Of course, the trade-off with that as you mention is that the BitUSD holder has difficulty predicting what kind of returns he will get.

I really think a fixed interest rate is simpler to market and more attractive for investors.  And I really want to have a way to reward investors for making a long term commitment.

So if you make the assumption that BTSX growth will always be good enough that you can sustain a 5% interest rate without depleting the insurance fund, then is it fair to say that you are basically taking the free-market variable interest paid by shorts, buffering it into a fund, and then paying that interest out capped at 5% to BitUSD holders, and then using the excess interest to pay for the BitBonds? If so, that means if BTSX growth slows (thus reducing interest rates shorts are willing to pay) and/or transaction and ask/bid overlap fees shrink, you risk having BitBonds become worthless (since they wouldn't pay higher interest than just holding BitUSD), the insurance fund providing no protection against black swans, and in extreme situations perhaps not even being able to pay the full 5% interest. I would tweak the "graceful degradation" to prioritize the insurance funds before the the 5% interest. Even then, I still don't care for the BitBond concept all that much. I don't see why we should be rewarding people more for long term commitment. Why not just take the excess interest and give that value to the shareholders instead?
« Last Edit: September 07, 2014, 12:55:32 AM by arhag »

Offline theoretical

Re: drltc's interest rate and BitBond proposal
« Reply #9 on: September 07, 2014, 02:23:54 AM »
I'm glad we are finally reaching an understanding. My issue is I really don't like price fixing and would prefer a design that can last. I rather avoid requiring some other consensus mechanism to constantly tweak parameters if there is another market-based solution that can take care of it. Of course, the trade-off with that as you mention is that the BitUSD holder has difficulty predicting what kind of returns he will get.

As I alluded to, what we do is pay 5% when we can, with any surplus going to bond-holders.  When we can't pay 5%, we shut down writing new bonds and pay whatever we can afford.

Shorts are, of course, free to charge a premium above the peg, which they can use against their interest obligation.  That is fairly close in practice to a variable interest rate.

My proposal here has the shorts paying a fixed interest rate.  I have an idea for charging shorts variable interest based on their leverage ratio, allowing a limited number of shorts to be initiated with high leverage (not enough to jeopardize the system) and giving people incentive to maintain lower leverage ratios (in the form of reduced interest on the portion they are "borrowing.")  However that idea is more complex, and I want to focus on one problem at a time.

So if you make the assumption that BTSX growth will always be good enough that you can sustain a 5% interest rate without depleting the insurance fund, then is it fair to say that you are basically taking the free-market variable interest paid by shorts, buffering it into a fund, and then paying that interest out capped at 5% to BitUSD holders, and then using the excess interest to pay for the BitBonds?

I'm not sure if I made this clear, but I intended in this proposal to have all shorts pay the same interest rate, I was thinking 7%.  That interest would be paid by an appropriate exponential increase in the short position's short BitUSD balance, which would gradually lower their margin call price.  Thus you should always be able to pay 5% to longs and have the 2% spread plus fees go to insurance and BitBonds, unless and until you get a margin crisis.

If so, that means if BTSX growth slows (thus reducing interest rates shorts are willing to pay) and/or transaction and ask/bid overlap fees shrink, you risk having BitBonds become worthless (since they wouldn't pay higher interest than just holding BitUSD),

Re-read the sections on BitBond mechanics.  The BitBonds are fully backed by BitUSD.  If the BitBond fund is depleted, the network will stop writing new BitBonds.  BitBonds already in circulation will pay their full face value in BitUSD, and can be sold in user-to-user transactions for less than face value if someone doesn't want to wait for the interest payment schedule (the buyer would receive an implied interest rate based on the price at which this transaction occurs).


the insurance fund providing no protection against black swans, and in extreme situations perhaps not even being able to pay the full 5% interest. I would tweak the "graceful degradation" to prioritize the insurance funds before the the 5% interest.

It was my intention to prioritize the insurance fund over 5% interest, unless the insurance fund already has sufficient capital.

Even then, I still don't care for the BitBond concept all that much. I don't see why we should be rewarding people more for long term commitment. Why not just take the excess interest and give that value to the shareholders instead?

If by "shareholders" you mean "BTSX holders", we would have to convert the BitUSD-denominated funds into BTSX to be destroyed.  Which would require a network-controlled "bot" doing open market operations on the BitUSD / BTSX market.  Yes, my auction mechanism is basically a bot that operates in the BitBond / BitUSD market, but in that market we have a very clear idea of what the "fair" value of a BitBond is (one BitUSD) and even the most abnormal market conditions cannot change that.

Also, if other cryptocoins are anything to go by, there will be some time until we have people setting up businesses using BTSX or BitUSD to sell things that people want to buy.  Relatively few people want to put their money in the BitShares ecosystem because there are few products; relatively few people want to create products because there's not enough consumer buy-in.  This is true of all cryptocurrencies, even Bitcoin -- and triply true for brand-new ones like BTSX / BitUSD.  It would be helpful to the ecosystem if we had some product which can only be bought with BitUSD, is useful to any consumer in arbitrarily large quantities, and whose producer can fill arbitrarily large orders.  BitBonds are that product.

Finally, I think BitBonds will prove to greatly increase the value of BTSX.  I'd suggest you re-read the sections on BitBond mechanics.  The existence of N BitBonds means N BitUSD, most of which come from investor principal, are "temporarily burnt" and are guaranteed to remain out of circulation for (on average) six months.  This means the BTSX collateral for those BitUSD is also guaranteed to stay removed from circulation for an identical period.  Indeed, you can view the BitBond auction as the network converting its BitUSD balance to BitBonds at a 1:1 ratio, then giving away those BitBonds to incentivize investors to voluntarily remove their BitUSD from circulation by converting it to BitBonds at a 1:1 ratio (and using an auction mechanic to focus on the investors who accept the smallest incentives relative to the BitUSD they're willing to convert).

Locking up large amounts of BTSX in collateral for long periods of time makes BTSX more scarce and thus increases its value.
BTS- theoretical / PTS- PZxpdC8RqWsdU3pVJeobZY7JFKVPfNpy5z / BTC- 1NfGejohzoVGffAD1CnCRgo9vApjCU2viY / the delegate formerly known as drltc / Nothing said on these forums is intended to be legally binding / All opinions are my own unless otherwise noted / Take action due to my posts at your own risk

Offline GaltReport

Re: drltc's interest rate and BitBond proposal
« Reply #10 on: September 07, 2014, 02:47:00 AM »
I'm glad we are finally reaching an understanding. My issue is I really don't like price fixing and would prefer a design that can last. I rather avoid requiring some other consensus mechanism to constantly tweak parameters if there is another market-based solution that can take care of it. Of course, the trade-off with that as you mention is that the BitUSD holder has difficulty predicting what kind of returns he will get.

As I alluded to, what we do is pay 5% when we can, with any surplus going to bond-holders.  When we can't pay 5%, we shut down writing new bonds and pay whatever we can afford.

Shorts are, of course, free to charge a premium above the peg, which they can use against their interest obligation.  That is fairly close in practice to a variable interest rate.

My proposal here has the shorts paying a fixed interest rate.  I have an idea for charging shorts variable interest based on their leverage ratio, allowing a limited number of shorts to be initiated with high leverage (not enough to jeopardize the system) and giving people incentive to maintain lower leverage ratios (in the form of reduced interest on the portion they are "borrowing.")  However that idea is more complex, and I want to focus on one problem at a time.

So if you make the assumption that BTSX growth will always be good enough that you can sustain a 5% interest rate without depleting the insurance fund, then is it fair to say that you are basically taking the free-market variable interest paid by shorts, buffering it into a fund, and then paying that interest out capped at 5% to BitUSD holders, and then using the excess interest to pay for the BitBonds?

I'm not sure if I made this clear, but I intended in this proposal to have all shorts pay the same interest rate, I was thinking 7%.  That interest would be paid by an appropriate exponential increase in the short position's short BitUSD balance, which would gradually lower their margin call price.  Thus you should always be able to pay 5% to longs and have the 2% spread plus fees go to insurance and BitBonds, unless and until you get a margin crisis.

If so, that means if BTSX growth slows (thus reducing interest rates shorts are willing to pay) and/or transaction and ask/bid overlap fees shrink, you risk having BitBonds become worthless (since they wouldn't pay higher interest than just holding BitUSD),

Re-read the sections on BitBond mechanics.  The BitBonds are fully backed by BitUSD.  If the BitBond fund is depleted, the network will stop writing new BitBonds.  BitBonds already in circulation will pay their full face value in BitUSD, and can be sold in user-to-user transactions for less than face value if someone doesn't want to wait for the interest payment schedule (the buyer would receive an implied interest rate based on the price at which this transaction occurs).


the insurance fund providing no protection against black swans, and in extreme situations perhaps not even being able to pay the full 5% interest. I would tweak the "graceful degradation" to prioritize the insurance funds before the the 5% interest.

It was my intention to prioritize the insurance fund over 5% interest, unless the insurance fund already has sufficient capital.

Even then, I still don't care for the BitBond concept all that much. I don't see why we should be rewarding people more for long term commitment. Why not just take the excess interest and give that value to the shareholders instead?

... It would be helpful to the ecosystem if we had some product which can only be bought with BitUSD, is useful to any consumer in arbitrarily large quantities, and whose producer can fill arbitrarily large orders.  BitBonds are that product.
...

This is interesting.  I just realized from your post that these financial instruments (Interest Earning Accounts, CDs , Bonds etc...) are PRODUCTS. WE will have PRODUCTS that can be marketed and SOLD.  We don't have to wait until some vendor starts accepting BitUSD etc...Wow.
« Last Edit: September 07, 2014, 02:49:27 AM by GaltReport »

Offline arhag

  • Hero Member
  • *****
  • Posts: 1213
    • View Profile
    • My posts on Steem
  • BTS: arhag
  • GitHub: arhag
Re: drltc's interest rate and BitBond proposal
« Reply #11 on: September 07, 2014, 04:05:39 AM »
drltc, first let me just say, thanks for all your quick responses. This has been a good discussion.

You have addressed most of my initial concerns. But there are two things left about the proposal that are unsatisfactory to me.

First, I really think it needs a mechanism to adjust the interest rate shorts need to pay. I don't like the idea of setting a fixed number like 7%. Shorts shouldn't be selling below the peg, so they need another way to compete for bids, and I think that should be through the rewards they are expected to pay to the funds.

Second, I am okay with the BitBond concept, but I still don't prefer it over alternative strategies of using the value of the excess interest. For one, I am not convinced how interesting BitUSD with full variable interest or BitBonds would be to someone who wants to grow their wealth over the long-term in these early stages of the DAC (and in the saturation stages there wouldn't be enough growth occurring to fund BitBonds anyway). It makes more sense to me to just hold the wealth as BTSX. If it is over the long-term and they expect BTSX to grow in value enough to provide such nice returns in the first place, the small short-term volatility should not be a concern to them. Assuming they have a little flexibility in when they sell it, it should give them higher returns than BitBonds.

Thus, I don't think of BitBond as a compelling product. BitUSD is the compelling product because it can be used as a currency (at least when we finally have merchants accepting it). Adding a reasonable interest rate (5% if we can manage it) makes a lot of sense because it outcompetes what people can get in their regular bank accounts. Adding more to it probably doesn't add much extra demand (people would want to keep the majority of their savings in the higher growth BTSX). If BitUSD is not useful right now, then the only reason people would be investing in BitBonds is if they believe BitUSD will be useful someday. This means they already believe in potential of BitShares X and that BTSX will grow. Why not just hold BTSX then?

So, what should be done with the excess interest after ensuring the insurance fund has enough reserve and the 5% interest is paid to BitUSD holders? Again, I would say it makes the most sense to just give that value back to the shareholders. Yes, that means having the DAC use the BitUSD to buy up (and then destroy) BTSX on the BTSX/BitUSD exchange at market price. Alternatively, I think it would be cool if some of that BitUSD was first used to pay delegates and other designated workers (who fund projects like development and marketing) their USD-denominated income. Then any remaining excess funds could be used to buy up and destroy BTSX to provide a dividend to shareholders.



 

Google+