Quote from: starspirit on April 24, 2015, 02:31:02 am

BM says "The 24 hour+ delay means you cannot depend upon errors in the price feed to guarantee a profit", which basically reiterates that X needs to be greater than the lag in the price feed, which surely is not that long. Did I miss something there?
Why couldn't the delay be reduced to 10 minutes for example?

I think it's fine to give an opportunity for all timezones to participate and band together against a whale manipulator, and thus provide a truly fair price of settlement. Think of it like a miniature fire sale auctions that last 24 hours.

Do you mean to say that if an extremely large settlement request is made, global community awareness of the pending settlement means that the requestor cannot as easily manipulate the BTS market price at the time that settlement occurs? Hopefully this equally prevents a whale short with the lowest collateral from also being able to manipulate the price to their benefit.

Thanks for the clarifications. While it makes sense to me now that a delay period is the prudent thing to do here, I do wonder what commitment the community actually has to prevent such a manipulation, when it won't be clear whether erratic price movements are manipulation or the real thing. While I'm not recommending this, I wonder if there are other things that can be done to thwart manipulators, such as having a random element in the settlement timing with an alert once complete.

Quote from: starspirit on April 24, 2015, 12:49:36 am

Quote from: joele on April 24, 2015, 12:43:19 am

I'm thinking that Bitshares will create an hedging BOT to correct the price.
This hedging BOT is an investment offer that earn interest.

I think its the action that is important, not the agent. Do you see how the hedging bot might correct the price in the scenario above?

The BOT has large amount of buy and sell orders, just like NUbits is doing.

In this scenario, the bot would need to make a market such that trades are equally likely to go both ways on its book. So it needs to make a market around the level people are comfortable paying, which is $0.90 in this scenario. Market-making like this cannot force the price to peg, unless the market-maker is willing to sometimes oppose the natural direction of the market and make significant losses. Think about central banks that try to peg their currencies when the market wants to go another direction.

The BOT has short orders that will buy orders below the feed to correct the internal price, the external exchange will follow the internal price then.

or

The BOT will close all sell orders below feed and close all buy orders above feed price.

On these joele, I find your meaning a bit fuzzy, would you care to elaborate? What specific actions would the bot be taking? For example "it would buy the internal bitUSD for $0.90 of BTS, and sell for..[ ] on the  [ ]"

I'm thinking that Bitshares will create an hedging BOT to correct the price.
This hedging BOT is an investment offer that earn interest.

I think its the action that is important, not the agent. Do you see how the hedging bot might correct the price in the scenario above?

Quote from: starspirit on April 23, 2015, 09:42:00 am

Clarify this bit? Which price of BTS (external?) and affected how?

Say BTS on BTS/bitUSD is trading at a discount (compared to bitUSD) on the internal exchange compared to the price of BTS in BTS/USD on an external exchange. Arbitrage means that there is a profit opportunity to buy up the BTS from the internal exchange and sell them on the external exchange, thus affecting the price and therefore the feed price.

OK, but I think its important to be specific here. That's not quite an arbitrage because you've started in one security (bitUSD) and ended up in a different security (real USD), so you need to take account of their relative pricing before you assume a profit. In principle, you then still need to switch back to bitUSD (where you started) at the market price to complete the cycle, and when you do that I don't see where the arbitrage profit is. Let's take a discount scenario...

Let's suppose the market consensus is that a bitUSD is worth less than a real USD, maybe 0.90 USD. On the internal exchange, a bitUSD will exchange for $0.90 of BTS (measured in real dollars). And on the external exchange, a bitUSD will also exchange for $0.90 of BTS or $0.90 real USD. Because they trade the same level inside and outside there is no arbitrage, is there?

Even if traders inside the exchange believed unrelentingly in the parity of the bitUSD, and exchange it internally for $1.00 of BTS, the arbitrage is to buy bitUSD outside and sell it inside, a behaviour that will cause the bitUSD prices inside and outside to converge somewhere between $0.90 and $1.00 of a USD, which is still a discount. And I'm not sure how this has an impact on the external price of BTS at all.

Are we on the same page, or am I on a tangent?

I would like to understand this concept better that I've seen referred to numerous times, because I'm just not getting what is meant by it. Maybe somebody can help make this concept click into place for me?

My starting point is that the holder of a bitAsset requires some assurance that its value (measured against external assets) will track the value of the real asset. Their key concern is that when they sell the bitAsset, an asset within the system, they receive enough funds that they could buy the real asset outside the system, should they wish to.

So this requires an exchange price of at least one asset inside the system to at least one asset outside the system, in order to make this calculation. This is what I understand as an external or exogenous price feed.

It's been said by many that when liquidity is significant inside the system, that we can move toward an endogenous price feed.

When people refer to an endogenous price feed, being "within the market", what exactly is meant by this? Does this involve a comparison of only the values of tokens that all reside inside the system? If so, I can't see how this provides any information about how any internal token is to be valued outside the system.

Am I understanding the concept correctly?

Thanks.

Quote from: starspirit on April 23, 2015, 02:15:42 am

Quote from: luckybit on April 23, 2015, 12:48:08 am

Quote from: starspirit on April 22, 2015, 11:49:22 pm

Quote from: luckybit on April 22, 2015, 02:44:13 pm

Quote from: lastagile on April 22, 2015, 02:17:03 pm

If I'm bad guy, and I have 1 million USD. I want BTS price keep falling down.
What I need to do is, change all my money to BITUSD, Then wait the monthly expire of short order. I'll hold all my BITUSD until we have no enough shorter, the expired short order have to get the BITUSD to cover. Shorter need to sell BTS to get BITUSD, while all the BITUSD is hold by me.This will keep the price falling. 

As a bad guy, I'll lose nothing, while earn some interesting.

I think the rule need be changed. The one hold BITUSD should pay people short for them.

In an bank I borrow money, and I can use the money to do more meaningful things, and I have to pay interesting.
In BTS, I short BTS and create BITUSD, but I can not use the BITUSD, other people get my BITUSD, and use BITUSD, and I have to pay the interesting to him. What a fuck???

BITUSD is for trade, not for people to hold. I'd like suggest people who short, set interesting to 0.

This idea must come up from some coder. But what we need is economist.

The last thing we need is an economist who has no experience with this space. Bytemaster has an economics background himself, he's not just an ordinary developer. That is why Bitshares has the advantages it has.

Bitshares started out as an experiment to test a hypothesis that Bytemaster put forth. It's just a testing of that hypothesis and we learned from that. Bytemaster could pursue a Phd in economics honestly just from his work on Bitshares so who would you be able to find who knows as much about this space and economics who could help?

There was an economist on the team at one point early on as an adviser. Advisers are good and can always help. What happened to the economist Charles Evans? http://www.pecuniology.com/who.php
 
https://www.youtube.com/watch?v=2ueyO72j_Iw
 
https://www.youtube.com/watch?v=ZT9ICMfUDjk
 
But I think it would be a dramatic error to appeal to the authority of aliens. Aliens to the space we are in have no experience, no trust, and while they might have knowledge which could be helpful it's unlikely they understand the mechanics of this particular environment to give advice. It's kind of like going to a lawyer to help with smart contracts, it's not helpful unless the lawyer is someone who really knows how to read and write code and also understands the space.

One man, no matter his intelligence, cannot design such a vast system alone, at least if it is to become as vast as his vision demands.

It's not one man, it's an entire community. At the same time there isn't any man with more knowledge and skills required for the task in this community than Bytemaster and I doubt anyone from some other community would be able to learn enough to be of much help in a short period of time.

It would take months for someone to figure out Bitshares and Bitshares is constantly changing.

luckybit, I understand this. I still don't understand all the inner workings, so it would take a long time for an outsider to learn it. I just don't think its true that coders can exclusively contribute value to the design of the system.

Who said only coders contributed? I brought up CWEvans and also many other random people on the forum have contributed via discussion and feedback.

However a random economist who hasn't studied or used Bitshares or even Bitcoin probably wont be very helpful. They simply would have no idea or reference for what a decentralized autonomous corporation is or what Bitshares is and everything would have to be explained.

By the time they catch up it would be months later. What exactly could they contribute that CWEvans can't contribute ? CWEvans actually knows the community.

At this point Bitshares is in the phase where Bytemaster has a vision for it and the community agrees with his vision. It's just a matter of getting the components to work in code.

I'm sorry if I misunderstood you. At the risk of doing so again, you now appear to be saying that other non-coders have contributed in the past, but the need for that has gone now because at this point its only about putting the envisaged design into code.

Yet we can see the design is still constantly evolving, e.g. BitAsset 3.0. And this is only a sliver of the total architecture eventually required for what Bitshares will ultimately become. And this is all as it should be, because there is an everlasting need to Reimagine Everything ahead of other keen innovators in the external world. Therefore I don't see that the need for such contributions ever truly disappears.

Wouldn't you agree it would be valuable if we drew more effectively on the collective skill set within the existing community, who are already very familiar with Bitshares? I bet that includes economists, lawyers, investors, financial engineers, marketers and other useful professions that could each place valuable perspectives on design proposals within the system.

The main feature of BitAsset 3.0 that I like is that redemption is possible at the feed price, which underpins valuation throughout this and external markets. My main issue has been that it eradicates interest, which is particularly important for currencies. I believe that if interest rates are high on a currency in external markets, it basically eliminates usage of the bitAsset as a transactional currency, because it will only pay to have it hoarded in the bond market. However, this weakness could be overcome if one of the bond market tokens represented an at-call deposit instantly interchangeable for currency (like an ATM machine) or with check payment facilities.

I've just posted a white-paper that describes the architecture of an at-call market integrated with a bill market. That is here...https://bitsharestalk.org/index.php/topic,15880.msg203776.html#msg203776
This envisages currency being instantly created and destroyed with floating interest as the clearing mechanism.

However, that system can be adapted to accommodate BitAsset 3.0 and potentially create a more powerful system.

i) a Cash Creation Market, along the lines of BitAsset 3.0 (although I would suggest some modifications), which does not bear interest, and has a floating level of collateral as the clearing mechanism, plus
ii) an at-call Deposit Market, whose tokens pay a floating interest rate from longs to shorts (like the Currency Creation Market in the whitepaper), and
iii) a Bill Market at various terms (akin to zero coupon bond tokens, as described in the whitepaper).

The Deposit Market would use exactly the same structure as I outlined for the Cash Creation Market in that white-paper, except that now it would be settled in the currency token instead of BTS, and the bids and offers would be for DEPOSIT tokens. This guarantees instant interchangeability between deposits and currency, with a floating rate determined by the market.

A challenge with this particular market is the zero bound on interest rates. The whitepaper outlines a couple of possible approaches.
 
What I like about this total approach is that it would now be much closer to the complete functionality of the current money market system, while obviously also having the usual block-chain based advantages.

I want to think about all this some more, but if it pans out, as soon as I get a chance I will update the whitepaper to v2 to incorporate this thinking.

As a follow-up thought to this, I was thinking more about the purpose of non interest bearing cash, when its possible to create interest-bearing currency. I can think of two possible purposes:

i) anonymity - there could be mechanisms in place to ensure anonymity of holding and spending (like some other crypto currencies).
ii) security - cash could be heavily collateralised in compensation for lack of interest

The question I have is do we ideally want a system with:

a) anonymous or extra secure cash + interest bearing deposits + bills/bonds, OR
b) interest-bearing currency + bills/bonds.

The option of (c) non-interest cash + bills/bonds, is lacking a critical ingredient as I've discussed, if we want these currencies to have transactional utility.

Any thoughts?

[edit: One more thing, by bond market, do we mean a counterparty-risk-free bond market, collateralised by BTS (akin to a government bond market), or do we mean a risky credit market for loans to counter parties? I am assuming the former.]

The main feature of BitAsset 3.0 that I like is that redemption is possible at the feed price, which underpins valuation throughout this and external markets. My main issue has been that it eradicates interest, which is particularly important for currencies. I believe that if interest rates are high on a currency in external markets, it basically eliminates usage of the bitAsset as a transactional currency, because it will only pay to have it hoarded in the bond market. However, this weakness could be overcome if one of the bond market tokens represented an at-call deposit instantly interchangeable for currency (like an ATM machine) or with check payment facilities.

I've just posted a white-paper that describes the architecture of an at-call market integrated with a bill market. That is here...https://bitsharestalk.org/index.php/topic,15880.msg203776.html#msg203776
This envisages currency being instantly created and destroyed with floating interest as the clearing mechanism.

However, that system can be adapted to accommodate BitAsset 3.0 and potentially create a more powerful system.

i) a Cash Creation Market, along the lines of BitAsset 3.0 (although I would suggest some modifications), which does not bear interest, and has a floating level of collateral as the clearing mechanism, plus
ii) an at-call Deposit Market, whose tokens pay a floating interest rate from longs to shorts (like the Currency Creation Market in the whitepaper), and
iii) a Bill Market at various terms (akin to zero coupon bond tokens, as described in the whitepaper).

The Deposit Market would use exactly the same structure as I outlined for the Cash Creation Market in that white-paper, except that now it would be settled in the currency token instead of BTS, and the bids and offers would be for DEPOSIT tokens. This guarantees instant interchangeability between deposits and currency, with a floating rate determined by the market.

A challenge with this particular market is the zero bound on interest rates. The whitepaper outlines a couple of possible approaches.
 
What I like about this total approach is that it would now be much closer to the complete functionality of the current money market system, while obviously also having the usual block-chain based advantages.

I want to think about all this some more, but if it pans out, as soon as I get a chance I will update the whitepaper to v2 to incorporate this thinking.