Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - starspirit

Pages: 1 2 [3] 4 5 6 7 8 9 10 ... 64
31
General Discussion / Re: Are we going to be the first, bitAPPLE (article)
« on: July 23, 2015, 02:41:55 am »
Quote from: Bitcoinfan on July 22, 2015, 10:24:03 pm
If I understand you correctly: it can still work with MPA, the feed price would just be higher to incorporate value of dividends with value of stock together. In a sense your decreasing the upside of shorters ( effectively removing supply in the supply demand equation).
That's correct, as long as a unit of the MPA reflects a higher quantity of the underlying asset (reflecting a reinvested dividend), so that the price per notional share remains consistent with the underlying asset. See my previous comments.

32
Technical Support / Re: Will bitassets in 2.0 Still Earn Interest?
« on: July 20, 2015, 11:08:13 pm »
If there were a floating rate between longs and shorts, then the free market would decide who pays who and how much. At any time it could be nearly guaranteed that this rate would not be the single point solution 0%. If the free market set the rate, there would be no need for further debate. Supply and demand, taking into account all conceivable factors and arguments, would be balanced at that rate (assuming it is completely unrestrained).

To relegate the free market mechanism to the bond market does not serve the same purpose, because it forces bitUSD to be locked away. Demand for bonds does not necessarily flow over to demand for commerce or everyday transactions. In fact, if bond rates are high enough, the opportunity cost of holding bitUSD at the ready is so high that it is less likely to be kept liquid for use in transactions.

An at-call variable rate deposit market is a solution to bridge this gap. As the variable rate will be economically linked with yields in the bond market, the relative yields will settle at a level that provides a balance between the liquidity demand for transactions and the demand for less liquid term investment.

As I often conclude these comments however, i need to add the caveat that the market- determined interest rate today, especially on currencies undergoing financial repression (esp USD and EUR) might well be very low if not negative. This will not always be the case however, especially with currencies in higher interest rate regimes.

33
Technical Support / Re: Will bitassets in 2.0 Still Earn Interest?
« on: July 20, 2015, 01:49:05 am »
Don't give up on yield. I believe its possible in 2.0 (with some adjustment) to have at-call interest bearing deposits, that you can draw on at any time - it just requires to operate a different way to 1.0. Having said that, with external USD interest rates so low, there would be little scope for interest on bitUSD right at this point in the cycle.

34
General Discussion / Re: Stregths of bitshares market vs vulnerability of BTS
« on: July 19, 2015, 12:03:35 am »
Quote from: starspirit on July 02, 2015, 01:56:08 am
Quote from: Ander on July 01, 2015, 11:18:16 pm
BTS decline 94% from peak to trough, between Sept 2014 and April 2015. 

Despite that decline, holders of bitAssets were still able to get the correct value back out of the system.  (Give or take a couple percent, and transaction fees).  In fact, They could even sell at times near the bottom for a 5-10% premium.

A deep bear market is a good test, but is not the nightmare scenario. The nightmare scenario is a price gap or discontinuity, such as when BTC collapsed from the $500-600 region to the $100 region in a matter of hours in February 2014, due to forced unwinding and insufficient market depth to meet the selling. Such flash crashes do happen from time to time in other markets too. These are rare (yet not inconceivable) events.
For further reason not to underestimate this risk, note that BTC experienced a 47% decline within 15 minutes on the btc-e exchange today. The fact that BTS has not suffered such a rapid decline to date is no evidence to think we are safe from it.

35
General Discussion / Re: Bank of England: Central banks looking at 'hybrid systems' using Bitcoin's block
« on: July 17, 2015, 11:58:09 pm »
Wouldn't central banks most likely run their own chain, using fiat IOUs backed by the central bank?

36
General Discussion / Re: Are we going to be the first, bitAPPLE (article)
« on: July 17, 2015, 05:06:53 am »
Quote from: puppies on July 17, 2015, 02:02:47 am
Please don't institute any of this.  I don't think either option i or option ii. Will work without an external infusion of funds.  These reasons have been stated previously in this thread by myself and others before they were ignored. 

puppies, if you are referring to this comment...

Quote from: puppies on July 13, 2015, 02:07:19 am
I am not 100% on the structure of MPA's in 2.0, and thats why I was talking about the administrator of the asset.  It really doesn't matter if that is an individual or the network as a whole though.  Unless you're suggesting dilution to pay for the dividends (which I would strongly oppose), the funds to pay the dividends would need to come from fees of some sort.  Paying dividends would necessitate reducing the profit of the network, increasing fees, or a combination of both.  I think paying dividends could be a very effective way of getting people to hold bitassets.  I hope to see lots of experiments to find the optimum level of profit for the network, and incentive to hold.  I don't think the network should take on the responsibility of matching the dividends paid out by a centralized company.  BTS holders should not be forced to pay bitAPPLE holders.

...then I wanted to clarify how this is intended to operate. Building dividends into the NAV increases the obligation against shorts, as it rightfully should. There is no need for the network, BTS holders, or anybody else to fund this. Have I understood your concern, and was that your main issue with it?

Quote from: jcalfee1 on July 17, 2015, 03:33:48 am
Quote from: Bitcoinfan on July 17, 2015, 12:19:51 am
Quote from: eagleeye on July 17, 2015, 12:11:25 am
@fuzzy

Can you ask @bytemaster if he is implementing @starspirit no. 2 of bitAsset valuing in the hangout tomorrow.  And if we are still 3 months away from 2.0 if we are three months away I will start a bounty.

Thank you each individual for your contribution to this thread and for kind words.

It's up to the feed producer and has can be enabled immediately out the gate with bit shares 2.0 presuming it delivers on what it promises
You could publish the value price in the feed with all adjustments baked in.   It doesn't have to match the real price it just has to be correct and verifiable.

What you both say makes sense. As a starting point, the easiest approach is to define the "asset" as a portfolio (that is adjusted from time to time) rather than a single share, and the feed producers could supply the price of that total portfolio rather than on a single share, with the same overall feed mechanics we use today. The same process could apply to more diversified portfolio mixes as well.

The trick is to ensure all the feed producers use a consistent method and timing for adjustments. This raises a few follow-up questions in my mind -

- Should the issuer or feed producers be responsible for alterations to the portfolio composition? Does this legally expose them to having some form of control over the asset?
- How should consensus be reached, especially with new event-types that might require some discussion?
- How can quality be managed, so that valuation inconsistencies do not arise? For example, we would not want feed producers mixing post-event (adjusted) prices with a pre-event (non-adjusted) portfolio for example. This needs to be transparent and able to be filtered out, rather than relying on the median estimate to take care of this.

In the end, these might be pragmatic questions for the issuer as much as they might be questions for the BitShares protocol.

37
Technical Support / Re: Suggestions on market charts in GUI
« on: July 16, 2015, 11:08:58 am »
Quote from: svk on July 16, 2015, 10:29:10 am
By window you mean time intervals? If so they're the same as for the price chart, which you select using the buttons above the chart.

Historic feed prices would require a separate database and api and is not something we're planning on adding atm.
On first point, yes that's what I meant. On second, a shame, but I understand.

38
Technical Support / Suggestions on market charts in GUI
« on: July 16, 2015, 02:47:36 am »
Do you think its a good idea to add an extra line for the history of the feed (or Call Price)? That would allow traders to have a price history showing even if the bitAsset market has traded thinly and lacks a smooth line (or in cases like Shenzhen, any line at all), which means another application is required to be used in conjunction to get this data. Traders like to see the price history and movement, because it affects their speculations.

What are the windows for the MACD lines? Could these be specified or labelled?

39
General Discussion / Re: Are we going to be the first, bitAPPLE (article)
« on: July 16, 2015, 01:26:58 am »
Quote from: Bitcoinfan on July 16, 2015, 12:19:05 am
Starspriit what's your thoughts on the new market mechanism for 2.0.  Do you have any concerns?  what are the biggest weaknesses that will hinder it from working?

This is off-the-cuff, so I might miss a few things.

Positives:

(a) Recently proposed mechanism that Smartcoins are self-created by shorts and then sold into the market, rather than shorts selling directly to create them. Big simplification of market mechanism.
(b) Removal of you-get-what-you-ask-for pricing to approach external market standard. This was an unnecessary cost on users.
(c) Removal of expiries on shorts. Recognition that settlements against shorts can be queued in other ways.
(d) Simplification of collateral rules for shorts, only need to meet minimum.

The biggest areas I still think we need to work on are:

(e) A mechanism to target parity rather than a floating premium. This would be more consistent with what traders are familiar with on competing exchanges.
(f) Option of a direct fee/yield mechanism between longs and shorts to mediate supply and demand. In some cases shorts may need incentive to support the issue.
(g) An ability to define the asset as a function of other inputs, with adjustments for dividends, carry costs, yields, asset replacement etc (as per this this thread).
(h) More diversified collateral, which I suspect the wider market would prefer to forced reliance on BTS (with early stage risk), or a derivative of BTS.
(i) Alternatives to full market liquidation in black swan events. Currently even minor losses could trigger these and impact on the bitShares brand. I've started a discussion in another thread.
(j) Re-assess the method of feed production. Currently there is reliance on a small set shared models for feed production.

(k) Expand the design flexibility available to issuers of privatised Smartcoins to accommodate a wider range of entrepreneurial solutions. Ultimately this is the final solution for most of the challenges (and possibly others) above - let the free market determine the best designs. I'm looking forward to the path toward flexibility that will be opened under 2.0, though its only the tip of an iceberg, and my goal is to keep promoting this direction. In fact, then we can all agree to disagree on the best approach and let the market prove what's most effective.

40
Technical Support / Re: Will bitassets in 2.0 Still Earn Interest?
« on: July 16, 2015, 12:23:32 am »
Right now, yield on USD and some other major currencies does not matter, and we can get away with it for the other benefits of crypto, only because the external interest rate environment is extremely low due to financial repression. This is not a usual state of affairs economically.

Ultimately, users will need an equivalent of an at-call deposit market where they can deposit their currency to earn interest, and withdraw at their discretion to make transactions. Otherwise it will be a financial disadvantage to hold SmartCurrency for transactional purposes rather than holding the real currency, especially for currencies in higher interest rate regimes, or when financial repression is eventually terminated.

The bond market as proposed does not meet this particular need well because it forces the user to lock away their money. Although this might be a neat way to earn investment return, this takes SmartCurrency out of the transactional system, and disincentivises use for commerce.

If external interest rates are high enough, it is IMPLEMENTABLE to have an at-call deposit market that pays users a yield (although less a percentage to incentivise shorts), combined with the ability to convert to and from non-interest bearing SmartCurrency at will if desired. I have a design for this, although it is not feasible in 2.0 without added flexibility in the parameterisation (which I aim to write a more detailed post on, and possibly submit as a project proposal in 2.0). It is simply not true what many people say that yield cannot be implemented. It's only that the version of yield that was put in place was flawed and failed, leading to philosophical rejection of the idea.

41
General Discussion / Re: Alternatives for dealing with Black Swans?
« on: July 15, 2015, 11:17:34 pm »
Quote from: cylonmaker2053 on July 15, 2015, 05:50:07 pm
this seems to introduce a new contagion risk with forced liquidation of the entire market being a new possibility.
Do you mean the 2.0 proposal to automatically trigger the liquidation event? It is possible, though perhaps unlikely. A user who had the ability to force the collateral price down by the required percentage from the minimum collateral mark could set up a small short at the minimum collateral requirement, and irrespective of total pool collateral in excess of that, force the event. Though it might be very costly to do so, assuming the minimum collateral requirement were set appropriately with regard to liquidity in the underlying collateral.

42
General Discussion / Re: Are we going to be the first, bitAPPLE (article)
« on: July 15, 2015, 11:05:10 pm »
Quote from: topcandle on July 15, 2015, 06:45:01 pm
Quote from: starspirit on July 15, 2015, 03:05:04 am
Quote from: jcalfee1 on July 15, 2015, 02:46:51 am
There is the stock split too.   It is actually a real challenge/expense to get the detailed dividend and split data required without having rounding errors (which compound over time) or violating someone's terms of service.   It is as if we have the new technology and that it is better if we support companies that upgrade.
Yes, in principle bitAssets should allow for ALL possible corporate actions such as dividends, bonus issues, buybacks, takeovers and share splitting.
The impact of any of these should be a fair adjustment to asset value and obligation of shorts.

There are two ways I can think of to do this in the case of stocks:

i) Replicate the corporate actions for owners of bitAssets. That would involve making distributions from the collateral pool for dividends, splitting the number of bitAssets for stock splits etc. I think to work out all these adjustments, while possible, would be very difficult to implement.

ii) Define a bitAsset as representing a varying number of shares. Then the fair value of the bitAsset can be adjusted over time to reflect all of these events. For example, a dividend would result in an increase in the underlying number of shares represented, essentially assuming reinvestment. A stock split would simply represent a change to the number of underlying shares.

I prefer (ii), at least initially, because I think it is a lot easier to deal with. There are no additional mechanics required for implementation of bitAssets, apart from the flexibility to define fair value as a function of stock price that depends on an adjusted number of shares over time.

Your # 2 is the right idea.  It would be the equivalent of a NAV calculation that a Mutal Fund / Actively manage fund would do at the end of the day.  A feed producers responsbility will be to make these calculations everyday, report and submit.  In the end the Spot Prices will be different between a MPA and a Stock, but the Fair Value will be nearly the same.

Exactly. This was the basis for my idea on ETF-style tracking funds here...(although if you read this thread, note that although the primary concept is unchanged, I have since then changed the application of both yield and settlement to what I think is a more efficient approach)...
https://bitsharestalk.org/index.php/topic,16672.msg213412.html#msg213412

Quote from: Ander on July 15, 2015, 07:03:10 pm
If you want an asset to track a stock over time, it needs to track X shares of the stock, where X increases when a divident is paid, a stock split, etc.

For example, I have an asset which initially represents 1 share of XYZ stock which is priced at $10.

XYZ gives a 10 cent dividend (1% of its value at the time of the dividend).  Now, my asset represents 1.01 share. of the stock.  Later the stock does a 2 for 1 split.  Now, my asset represents 2.02 shares of the stock.  And so on.

If you don't do it this way, any stock split will ruin the asset, and it will lose ground over time as the company issues dividends.
We are in agreement.

Assuming (ii) is the preferred model for now, a follow-up question is then how best to implement this NAV style calculation. At this point I tend to think the easiest way is to trust the feed producers to provide all the relevant inputs, rather than trying to code in all the possible permutations of corporate actions at the outset. Feed producers could provide multiple inputs (for example Price, Quantity, and Time), with guidance provided by the issuer and collectively as to how best to adjust these quantities with upcoming events. From this the NAV could be updated automatically.

NAV, rather than price, would then serve as the basis for settlements and liquidation events also.

Quote from: cylonmaker2053 on July 15, 2015, 01:56:16 pm
the more we flush out this theoretical asset, the more i'm starting to think it's maybe too problematic to worry about, the possible rewards not worth the headache; especially since we have so many possible billions of dollars of capital flows from the simple pegged products, like bitUSD, bitCNY, bitEUR, bitBRENT, bitGOLD, bitSILVER, etc. mimicking corporate equity, which means trying to account for all the possible decision permutations management could make, seems like a mess when we're sitting on a gold mine in possible value with our current products. by no means is this impossible--maybe something to keep flushing out for 3.0--but i sure hope our devs don't divert any brain power or labor hours at this point.

Yes, shares are a level of difficulty up from currencies, commodities and indexes. I agree it is better we set our sights on the lower hanging fruit first and build up. These are issues we will still need to deal with even in the simpler asset classes, although the event-types are more limited. For example, deposits pay yield. Commodities have costs of carry and funding rates to contend with. Indexes also have dividends (but usually there is shandy accumulation index to use).

43
General Discussion / Alternatives for dealing with Black Swans?
« on: July 15, 2015, 06:36:51 am »
The current proposal for 2.0 is that when the least collateralised short has insufficient collateral remaining to cover 100% of their debt, then a liquidation event is triggered. The liquidation event settles all shorts, and leaves the pool of remaining collateral for longs to settle against at a time of their choosing. In effect, the shortfall loss from any under-collateralised shorts is shared by all the longs, and thereafter the token no longer reflects price movements of the original bitAsset, but only the collateral.

I've been thinking about a couple of issues with this. First, even an immaterial under-collateralisation loss would result in complete liquidation of the market.  That could be detrimental for popular Smartcoins (e.g. bitUSD or bitCNY) and affect BitShares' brand even if it occurred once or twice (especially if there is a rush to sell the collateral, leaving big losses). Second, the full collateral backing of the token is not available to longs in such an event - they bear a loss as soon as the least collateralised short becomes under-collateralised, equal to the shortfall. This is even though total collateral in the pool may still be much greater than 100% of total debts. (As an aside, the collateralisation levels we generally quote probably give a false sense of security, because adequately collateralised shorts also have a claim on that collateral).

I would like issuers of privatised Smartcoins to have more flexibility in the rules governing black swan events, so that they can manage these issues in different ways they think is appropriate - without making judgements in this post on what ought to be done, because people will have different preferences here.

Here are some key options I would like to see:

i) Ability to let the pool of longs absorb the loss, while leaving the market open, and only allowing business-as-usual settlements at the reduced value (equivalent to what the liquidation value would be at any time under the 2.0 approach). This maintains equity amongst tokens that settle and those that remain trading. However, it also means that the shortfall amount (under-collateralisation loss) that has not been diluted through settlements, continues to live on as a deduction against fair settlement value for remaining longs, unless there is some way to plug this gap (either through (iii) below, or if the market cap grows significantly and the percentage discount gets low enough, through (ii)).

ii) Ability to have the pool of shorts take the first loss, and leave the market open. That is, spread losses from under-collateralised positions across remaining shorts, via a pro-rata increase in the obligation of each short. This protects longs and allows the market to keep trading at parity as long as the total collateral in the pool is sufficient to cover all debts. In this case shorts share a collective interest in the insurance of the Smartcoin.

iii) Initiate a transparent reputation system for shorts, that incentivizes shorts to make good on margin called positions that were under-collateralised, or the broader community to voluntarily rectify under-collateralisation events. So for example, a short that is margin called and under-collateralised, has a debit against their name for the shortfall amount. This might prevent them from taking short positions on other assets, assuming issuers have the ability to grant or reject authority to individual shorts. If these shorts make good on this amount, within an acceptable time, that debit is removed, and their reputation maintained. Anybody else in the community could also voluntarily contribute to plugging an under-collateralised position, and receive recognition via a credit. This process would require a mechanism for users to contribute to the general collateral pool.

I don't have all the answers here, just trying to see what options are available other than a forced liquidation event. Feel free to express your views or other ideas.

44
General Discussion / Re: Are we going to be the first, bitAPPLE (article)
« on: July 15, 2015, 03:05:04 am »
Quote from: jcalfee1 on July 15, 2015, 02:46:51 am
There is the stock split too.   It is actually a real challenge/expense to get the detailed dividend and split data required without having rounding errors (which compound over time) or violating someone's terms of service.   It is as if we have the new technology and that it is better if we support companies that upgrade.
Yes, in principle bitAssets should allow for ALL possible corporate actions such as dividends, bonus issues, buybacks, takeovers and share splitting.
The impact of any of these should be a fair adjustment to asset value and obligation of shorts.

There are two ways I can think of to do this in the case of stocks:

i) Replicate the corporate actions for owners of bitAssets. That would involve making distributions from the collateral pool for dividends, splitting the number of bitAssets for stock splits etc. I think to work out all these adjustments, while possible, would be very difficult to implement.

ii) Define a bitAsset as representing a varying number of shares. Then the fair value of the bitAsset can be adjusted over time to reflect all of these events. For example, a dividend would result in an increase in the underlying number of shares represented, essentially assuming reinvestment. A stock split would simply represent a change to the number of underlying shares.

I prefer (ii), at least initially, because I think it is a lot easier to deal with. There are no additional mechanics required for implementation of bitAssets, apart from the flexibility to define fair value as a function of stock price that depends on an adjusted number of shares over time.

45
General Discussion / Re: What will happen when bitcoins reward halves to 12.5 BTC?
« on: July 14, 2015, 01:16:53 am »
I think the impact on the bitcoin price is not easily predictable.

At least in theory, a known event should not have an abrupt impact on the bitcoin price at that time, because all parties (bitcoin market participants and miners) reflect that information in advance when weighing up the economics of their current decisions. There may be an action required at that point (such as miners turning off equipment), but, again in theory, the potential for all those actions and their consequences should be built into the current price in a fully informed market. In fact the entire future schedule of reward reductions should, in principle, already be built into the market's valuation of bitcoin in this way.

That means that any resulting movements that do occur at such times are not directly from the event itself, but from endogenous market reactions as talk of the event becomes elevated and market participants try to collectively trade off the sentiments of all other players and less informed participants. This endogenous action may (or may not) trigger some increased volatility around the event, but, again in theory if the market were efficient (a big "if") at pricing in the events, it would not be possible to profitably predict the price reaction one way or the other.

Having said all that, the bitcoin market is very immature, so in practice it may be possible to speculate on whether the market is actually pricing in the event efficiently or not.

Pages: 1 2 [3] 4 5 6 7 8 9 10 ... 64

Page created in 0.063 seconds with 23 queries.