ICBIT Derivatives Market (USD/BTC futures trading) - LIVE

[toffoo]

... with just a few more trading days remaining.

Fireball,

That reminds me to ask, perhaps soon it would be a good idea to clarify exactly how the settlement of this expiring contract will function.  This will be the first futures expiration I've seen since I started to look at this exchange a few months back (I'm not sure if it's the first overall) and it's probably best to spell out here exactly how this works to avoid any unpleasant surprises or attempts at manipulation.

The (relatively sparse) contract specification page https://icbit.se/BUZ2 says:

Last trading day: 15.12.2012
Settlement date: 15.12.2012
Settlement: Positions are settled based on the volume weighted average rate of USD/BTC on the exchange with the most average monthly volume (for the month of contract settlement) during the contract settlement day by transferring variation margin between contract holders.

• Is it fair to assume it will be Mt.Gox's rates?
• Where will the official feed of prices/volumes to be averaged come from?
• What hours/timezone define this "contract settlement day"?
• I believe 20:00 UTC is the daily settlement time I've seen used there.  Will BUZ2 trade until that time?  Will your official VWAP number come out and variation margin settlement happen immediately afterwards or should we expect some delay for calculation/settlement?
• Any other potential "gotchas" you think we should watch out for?

Thanks!

[segabtc]

And will there be a"trading fee" since if we hold till the end, these contracts should just expire, correct. for example, If I am hold 10 futures and the previous close, lets say is 13.00 and the VWAP of Gox, again assume, is 13.50, do i get the var at closing, get paid for the difference, and then the position is closed without trading fees? Just trying to prepare for this coming expiration, and would like to know how you plan on dealing with this final closing. Thanks, doing a great job BTW.

[Fireball]

• Is it fair to assume it will be Mt.Gox's rates?

Yes

• Where will the official feed of prices/volumes to be averaged come from?

Mt.Gox website (https://mtgox.com, "Weighted Avg" field atop of it) which provides sliding window 24h volume weighted average.

• What hours/timezone define this "contract settlement day"?

UTC

• I believe 20:00 UTC is the daily settlement time I've seen used there.  Will BUZ2 trade until that time?  Will your official VWAP number come out and variation margin settlement happen immediately afterwards or should we expect some delay for calculation/settlement?

Yes, trading will stop at 20:00 UTC as usual, and the contract will be settled already after clearing. Some delay would happen, but the VWAP will be used as of 20:00 UTC from MtGox website.

• Any other potential "gotchas" you think we should watch out for?

Yes, depending on the MtGox USD VWAP at the settlement day, there is a probability that profits from the futures will be capped because there are customers with open positions who may not have enough money in their accounts at a given settlement price. We'll provide full disclosure of the situation.

Additionally, today the update to the Accounts tab will be deployed, featuring some visual improvements and support for Google Auth one time passwords.

[Stephen Gornick]

there is a probability that profits from the futures will be capped because there are customers with open positions who may not have enough money in their accounts at a given settlement price. We'll provide full disclosure of the situation.

I sense foreboding.  

So someone that closes a position today gets 100% of the gains but someone holding to expiration gets capped?  And there's no knowing by how much?

[killerstorm]

If I understand correctly, profits can be capped only if settlement price is so far from previous clearing price that it can eat through required initial margin.

I don't know how much is "too far", but probably it is related to price range.

Although it looks like it is possible to bit through price range: (1/11.952-1/14.608)*10= 0.152, but initial margin is only 0.1141. But it's kinda unlikely scenario.

So, well, unless something extraordinary happens on mtgox and if people will be arbitraging on ICBIT to prevent huge price swings there should be no problem.

[Fireball]

I deployed the Accounts page update. Now, along with minor visual improvements, this page is going to require a one time password to be entered for any withdrawal operation, if you have tied Google Authenticator to your account.
Finally, it makes the "one time passwords done right", as Stephen said.

[Fireball]

it cancels the order but also I lose all other orders info on that window til I "reload" the page. Can we get a "reload" button like the "Balances" window has?

It doesn't need a reload button, bug which wipes your orders should be fixed instead.

Ok, this is fixed now (don't forget to reload your page to refresh javascripts), and Your Orders table is updating nicely now. Also some people reported situation that they are getting duplicates in the table - this should not happen anymore.

[picobit]

Hi Fireball,

There seems to be something wrong with the equations giving the variation margin on your web page.  The actual calculation is OK, it is just the web page that gives the wrong expression.

The variation margin should be the price difference times the size of the futures contract.  As the actual contract is a futures contract for -10 USD, traded in BTC, but the price is listed as the price in USD/BTC, a reciprocal comes in, confusing matters a bit, and the equation becomes (on your web page):

Vm = (1/p2 - 1/p1) * W/R

Here 1/p is the "real" price (in BTC/USD), and W/R should be the contract size, i.e -10 USD, giving a variation margin in BTC.  However, on the web pages https://icbit.se/BUZ2 and https://icbit.se/BUH3 you list W = 0.0001 USD and R = -0.001 USD, giving W/R = 0.1 (no unit).  So both the magnitude and the unit is wrong.  The example at the bottom of the page correctly uses W/R = -10 (but leaves out all units).  And the actual variations margins calculated in your trading platform are correct as well, fortunately

I have not looked closer into the Gold and Oil options, but since the prices here are listed in BTC I am a bit surprised to the reciprocal of the price in the equations.

[Fireball]

I have not looked closer into the Gold and Oil options, but since the prices here are listed in BTC I am a bit surprised to the reciprocal of the price in the equations.

Thanks, I will check the equations on the web to make them match the real calculations, but as for Gold and Oil, they don't use such sophisticated pricing. Price is quoted directly in BTC, so they are classic linear futures (the 1/x non-linearity is not present there).

[guruvan]

it cancels the order but also I lose all other orders info on that window til I "reload" the page. Can we get a "reload" button like the "Balances" window has?

It doesn't need a reload button, bug which wipes your orders should be fixed instead.

Ok, this is fixed now (don't forget to reload your page to refresh javascripts), and Your Orders table is updating nicely now. Also some people reported situation that they are getting duplicates in the table - this should not happen anymore.

Yes this was fixed

My poor browser still shows hundreds of megs of ram consumed after a short time & I'm noticing that it's eating 25% cpu (sustained, on my little netbook)

Thanks for your hard work.

EDIT:

I have not looked closer into the Gold and Oil options, but since the prices here are listed in BTC I am a bit surprised to the reciprocal of the price in the equations.

Thanks, I will check the equations on the web to make them match the real calculations, but as for Gold and Oil, they don't use such sophisticated pricing. Price is quoted directly in BTC, so they are classic linear futures (the 1/x non-linearity is not present there).

The equations for both are VM = (1/PriceClose - 1/PriceOpen) * W/R;

[Ichthyo]

The equations for both are VM = (1/PriceClose - 1/PriceOpen) * W/R;

Shouldn't that be VM = (PriceClose - PriceOpen) * W/R;  ?

Since the prices are in BTC. Obviously, the BTC/USD rate is indirectly factored into these rates "Oil/BTC" or "Gold/BTC" but that's another story 

[guruvan]

The equations for both are VM = (1/PriceClose - 1/PriceOpen) * W/R;

Shouldn't that be VM = (PriceClose - PriceOpen) * W/R;  ?

Since the prices are in BTC. Obviously, the BTC/USD rate is indirectly factored into these rates "Oil/BTC" or "Gold/BTC" but that's another story 

yes. they're incorrect on the website.

[Fireball]

yes. they're incorrect on the website.

Ouch, copypasta :-). Fixed already, thanks.

[ThePok]

Is the settlement for free? Or do we have to pay 0.005 too?

[Fireball]

Is the settlement for free? Or do we have to pay 0.005 too?

No fees for BUZ2 settlement, so it'll be free.

[smickles]

Is the settlement for free? Or do we have to pay 0.005 too?

No fees for BUZ2 settlement, so it'll be free.

Do you mean to imply BUZ3 will have a fee?

[Fireball]

Is the settlement for free? Or do we have to pay 0.005 too?

No fees for BUZ2 settlement, so it'll be free.

Do you mean to imply BUZ3 will have a fee?

No, I did not mean that the next contract will have it. In case it would have a settlement fee though, this is going to be announced well before teh contract's settlement date.

[ThePok]

Thats not acceptable too. It should be announced befor the Future is started!

[Fireball]

Thats not acceptable too. It should be announced befor the Future is started!

"well before teh contract's settlement date" == contract launch day, maybe :-)

Seriously speaking, there is no need to introduce the settlement fees in ICBIT now.

[Fireball]

Trading will be stopped up to an hour after today's 20:00 UTC clearing.

About debts problem:
we had a long discussion with Tycho and there are two ways to solve problem of debt, when there is a customer who opened a position, market moved against her, and there is a lack of liquidity (you can still see huge BUZ2 buy "wall" at ~12 which is a liquidation of debtors' positions, there is no such liquidity on the market to forcefully close his positions).

To fix such situation there are two ways:
1. Profit capping. In simple words, if you made profit buying/selling/leaving contract for settlement, your profit will be reduced, but you still get a profit, as much as counterparties could pay.
However, Tycho said its inacceptable and this option will not be used.

2. Mutually closing positions of debtor against equally distributed counter-positions holders. In simple words, this means if the market does not have enough liquidity to close debtor's position, and you hold profitable open position, your position will be partially closed (at a rate which still brings you profit). This is what's going to be used.

Soon after BUZ2 setllement I will open up a separate thread for arbitraging strategies discussion with arbitrage bots samples source code provided, because ideally we need to solve the problem of liquidity so that neither points 1 or 2 above would be needed at all.