Short Order Refactoring

[xeroc]

The only real change I see in the OP is that a short-sell now has to potentially pay two (different) fees in contrast to one fee atm .. though this only makes the DAC more proditable (depending on the ACTUAL fees)

[arhag]

{Disclaimer} I did not get toast/Runes big break through idea that interest is bigger than collateral, so...take the above with a grain of salt!!!

Yeah, I'm not sure what any of that has to do with the changes discussed in this thread.

As far as I understand (and I am really not sure if I do), he is just talking about how things break down and require changes when something like BitUSD becomes the unit of account for goods and services (meaning when we have succeeded in world domination ). That is a very interesting discussion to me and would love to hear more about your thoughts, toast, about what the new system would need to look like and how the transition from the old system to new system could happen.

[tonyk]

Ha ha, several of you went to some pretty big exaggerations in this thread, for sure...

Calm down!

This changes nothing... but simplifies the market (order) matching by combining operations. The rest of the perceived benefits are just in your heads boys.


{Disclaimer} I did not get toast/Runes big break through idea that interest is bigger than collateral, so...take the above with a grain of salt!!!

[toast]

This was the first step in a line of reasoning that caused Rune to conclude that BTS and maker are both central banks and the collateral is incidental. It's just that leveraged trading is one of the only cases where we can get excess collateral for a loan put down voluntarily. At a large enough scale (like took-over-the-world scale), any particular bitasset must detach and the exchange CFD mechanism has to be replaced with a mechanism that lets price float but controls interest based on consumer price feeds.  (edit: actually mb you can use the exchange CFD against the price index itself if the stuff in the index is super liquid)

[svk]

Not having the third order box for shorting is great as well because it makes the trading interface look more like a traditional interface, and frees up some real-estate. Of course we need to present the new borrowing feature but I much prefer this flow.

[xeroc]

Breaking out the shorting operations into discrete transactions is fantastic. The single short/cover operations of old could still be implemented in the client. I like the idea of breaking out what could be a client side feature/batch and what the blockchain should field. The more that can be pushed to client side the less bloat in the blockchain. Reminds me of programing in assembly where you have to do the extra work of building each operation but it is extremely lean and fast.

This.
back to the KISS principle!

for the OP

[Riverhead]

Breaking out the shorting operations into discrete transactions is fantastic. The single short/cover operations of old could still be implemented in the client. I like the idea of breaking out what could be a client side feature/batch and what the blockchain should field. The more that can be pushed to client side the less bloat in the blockchain. Reminds me of programing in assembly where you have to do the extra work of building each operation but it is extremely lean and fast.

[starspirit]

Great. This makes it easy (and cheaper) to better simulate a short sell order without excessive soft-collateral funds. You can put up nearly all of the BTS (excluding some to pay for fees) into a short position and get a good amount of BitUSD out of it (not too much to have a high enough collateral value to debt value ratio that makes it unlikely to be margin called or force called from redemption). Then you can sell the BitUSD for more BTS which is then moved back into the short position to increase the debt and then take that extra BitUSD and sell it again. If we assume the value of the initial debt is 50% of the value of the initial collateral and the prices don't change much during this process, then the BitUSD amount will halve with each round. In just a few rounds, it becomes negligible to extract and sell any more debt from the short position. Furthermore, adjusting the ratio at a later point only requires transactions dealing with one short position rather than multiple ones (for example, a new short position created with each round in the process described earlier).

arhag, you just filled a gap in my own thinking. I realised that self-creating a short and selling the long for BTS meant that the short did not have as much leverage to BTS as under the current approach of short-selling directly into the market. I felt though that the equivalent outcome should still be possible, and now you've described that process so thanks. So in theory there should be no loss from the new approach even if the borrower's motive is still to leverage BTS as much as possible (although its a bit less convenient).

[starspirit]

I support this change. I also believe we need to shift the paradigm even further:

1. The old CFD metaphor is obsolete and will only confuse things going forward. Smartcoins are no longer brought into existence by BTS bulls wanting leverage. They are brought into existence by borrowers of the asset, who simply offer BTS as collateral. They can then sell the asset and use those funds for any purpose they desire. Buying BTS for leverage is just one of those many uses. Perhaps we should be promoting SmartCoins in a more relevant way.

What is a bitUSD? Its a transferable token that represents a collateralised USD loan to a set of borrowers.

2. A primary use of such borrowing should be making markets in the Smartcoins. By self-creating the Smartcoin, the user can switch their long position between the Smartcoin and the real underlying asset as relative prices move, without ever changing the number of BTS they hold (and now use as collateral). Their only requirement is to ensure their collateral remains sufficient.

I've previously commented on each of the above, for example here... https://bitsharestalk.org/index.php/topic,16427.msg210011.html#msg210011

There are some further possibilities this approach opens up. For example:

3. There is scope for much greater flexibility on collateral, given it is controlled exclusively by the shorts. That is, the collateral is completely independent of what the Smartcoin trades against in the free market. It could be BTS, or a BTC or USD substitute, or even a portfolio of collateral tokens. This would open up the architecture of Smartcoins, increasing the potential range of users on both sides of the coin. For further comments on the concept of Flexible Collateral, see:
https://bitsharestalk.org/index.php/topic,16326.msg208798.html#msg208798, and

By the way starspirit, I don't see how the BitAssets backed by a mix of collateral types would be fungible unless a fixed mix ratio was specified as part of the BitAsset definition that all shorts of that privatized BitAsset had to satisfy. And in that case, I would imagine the only practical way to short new BitAssets with mixed backing collateral into existence would be through a self-short. The logic for margin calls would also get more complicated with mixed collateral.

Correct. With self-shorting and self-cancellation, shorts get to control the mix of collateral they want. I've been working on just such a structure. Collateral can then be completely independent of the markets in which the token trades. Also margin calls could be satisfied by applying each collateral token in sequence to covering the debt until it is satisfied, and then returning the residual collateral tokens to the short. This sequence could even be determined by the short.

Also, in case it wasn't clear in my answer, no, you don't require a fixed mix ratio. The shorts could change the mix of collateral as they please, as long as they met the minimum coverage conditions.

[Edit: Offering collateral flexibility like this admittedly comes with some complexity though, notably in dealing with settlements and covers].

Bottom line - the OP is the right way to go, but there is a lot more possibility than you are thinking yet.

[monsterer]

Any way this can be made symmetrical, so you can lock up bitAssets to create BTS at the feed?

[maqifrnswa]

This could be the mechanism that would allow arbitrage towards the feed - it could be the missing link!

When BTS:BTA is greater than the feed, I can generate BTA for cheaper than I can buy them on the market, then sell them on the market for profit, which returns the price to the feed! You are rewarded for "fixing" the feed when BTA is under-supplied!

[svk]

Great idea! Worth implementing just because of how much easier it becomes to explain the process!

[merivercap]

Margin calls due to insufficient collateral will execute up to the "call limit" which is a second price provided in the feed.   This allows each BitAsset (Smartcoin) to define how much the book may be walked when executing a margin call.

Hmmm... sounds good, but I need some clarification.  I assume all the margin calls should be settled and it seems the 'call limit' allows walking the book until a certain price.  And thereafter? 

[roadscape]

Sounds great... Is there any conceivable downside to this refactoring?

edit: After re-reading it seems the main caveat is that to be completely short requires multiple iterations of selling your newly-created longs for extra collateral, as arhag described.

[nomoreheroes7]

Yup, I always had trouble figuring out what exactly the hell was happening when a short was opened, etc. I mean I had a general idea, but the way it was presented was intimidating for newbie traders. I'll support anything to make it less intimidating/confusing to the noggin.