The economics of BitCoin

economic

GreatApe wrote: »

How?

Who is going to lend their bit coin for it to be shorted?
One of the aspects of this 'currency' is that a lot of the owners are rightfully paranoid in losing their coins or having them stolen.
Also how do the futures exchanges even find the people who own the coins to borrow from them to short?

From my understanding these futures exchanges are going to act as glorified spread betting platforms. I can see how they can hedge if they have net longs they just buy more bitcoin to hedge. But how do they hedge if they have net shorts unless the exchanges already own a lot of coins they can sell into the market. Maybe this is part of the price ramp the exchanges trying to buy 5% of the coins to give them the ability to do net shorts. There are two exchanges going live within days and Nasdaq wanting to get in.

Do you have a link to the details of the futures that will be traded on CME? The futures contracts could be cash not not physically settled and so there is no need to even borrow the BTC to deliver. So it will as you say be like spreadbetting.

The main question is about margining and risk management. Given the volatility how will the exchanges manage this? They would need a load of cash from the counterparties to mitigate counterparty risk. If given a 99% confidence that BTC wont move more then 30% in one day, then the margin requirement would need to be at least 30% of the position plus more for liquidity.

But what if BTC gets crazy and moves 50%+ or more on a daily basis (both down and up). You could see huge risks developing by the exchanges.

economic

Found the link:

http://www.cmegroup.com/trading/equity-index/us-index/bitcoin_contract_specifications.html

It looks liek it is cash settled (i thinkt hats what financially settled means). Cant see anything on margin though - that would be the most interesting part of it all.

GreatApe

economic wrote: »

Found the link:

http://www.cmegroup.com/trading/equity-index/us-index/bitcoin_contract_specifications.html

It looks liek it is cash settled (i thinkt hats what financially settled means). Cant see anything on margin though - that would be the most interesting part of it all.

Yes it is cash settled and yes they will offer margin accounts

What I mean is how do they offer net Longs or shorts without buying/selling the underlying coins
So if 100 clients each short and 100 clients each go long the exchange is fine they are effectively acting as a vehicle for these 200 clients to bet against each other

But if 20 clients go short and 100 clients go long they would need to hedge that bet by buying the underlying coins. Or turn away 80 clients going long. Likewise if 20 go long 100 go short they would need to short the actual coins to hedge

economic

GreatApe wrote: »

Yes it is cash settled and yes they will offer margin accounts

What I mean is how do they offer net Longs or shorts without buying/selling the underlying coins
So if 100 clients each short and 100 clients each go long the exchange is fine they are effectively acting as a vehicle for these 200 clients to bet against each other

But if 20 clients go short and 100 clients go long they would need to hedge that bet by buying the underlying coins. Or turn away 80 clients going long. Likewise if 20 go long 100 go short they would need to short the actual coins to hedge

they would wait until the order is matched i think. it'll work the same way as stock futures or bond futures.

if 20 go short and 100 long then the prices will move up to reduce demand for longs and increase demand from shorts.

westernpromise

GreatApe wrote: »

Yes it is cash settled and yes they will offer margin accounts

What I mean is how do they offer net Longs or shorts without buying/selling the underlying coins
So if 100 clients each short and 100 clients each go long the exchange is fine they are effectively acting as a vehicle for these 200 clients to bet against each other

But if 20 clients go short and 100 clients go long they would need to hedge that bet by buying the underlying coins. Or turn away 80 clients going long. Likewise if 20 go long 100 go short they would need to short the actual coins to hedge

You do not understand how futures exchanges work.

GreatApe

economic wrote: »

they would wait until the order is matched i think. it'll work the same way as stock futures or bond futures.

if 20 go short and 100 long then the prices will move up to reduce demand for longs and increase demand from shorts.

Ok I wasn't aware of this i thought the exchanged tracked the underlying market almost continuously with some divination due to the cost of capital. I suppose this is roughly what happens with most futures markets as the underlying is so liquid.

So the exchange can diverge from the underlying market except for the settlement date/time when it settles at the underlying market price

Makes more sense now

westernpromise

economic wrote: »

Found the link:

http://www.cmegroup.com/trading/equity-index/us-index/bitcoin_contract_specifications.html

It looks liek it is cash settled (i thinkt hats what financially settled means). Cant see anything on margin though - that would be the most interesting part of it all.

You need to look at the clearing rules, which as far as I can see have yet to be published. In principle though they would just SPAN margin it like anything else. The daily settlement price is based on a daily reference rate compiled in accordance with the IOSCO principles for financial benchmarks.

economic

westernpromise wrote: »

You need to look at the clearing rules, which as far as I can see have yet to be published. In principle though they would just SPAN margin it like anything else. The daily settlement price is based on a daily reference rate compiled in accordance with the IOSCO principles for financial benchmarks.

Yes it would be SPAN which is effectively based on MC but using historical data of BTC to calibrate the MC model?

Yes its the clearing houses that take on the counterparty risk not exchanges (they just take execution risk). Then you have clearing members which would probably slap on a multiplier of some sort on the SPAN margin to charge their clients depending on client risk and the state of the BTC market.

What i am curious is how they model and calibrate the model to produce the SPAN numbers given it is BTC (8 years of history). I presume at the start of the BTC (back in say 2010) we had seen daily moves like -75% or -50%. Will these moves be taken into account? Is it weighted so that the latest data is more influenced in the SPAN numbers? Is it normal or lognormal returns that are being used?

EDIT: Sorry i have not worked on futures or clearing houses, just stuff i had been reading a whilst back, hence the silly questions!

westernpromise

GreatApe wrote: »

Ok I wasn't aware of this i thought the exchanged tracked the underlying market almost continuously with some divination due to the cost of capital. I suppose this is roughly what happens with most futures markets as the underlying is so liquid.

So the exchange can diverge from the underlying market except for the settlement date/time when it settles at the underlying market price

Makes more sense now

If the future doesn't converge on the underlying, then by definition it is worthless as either a hedge or proxy for that underlying. Here it will converge because the settlement price for open positions, the BRR, is derived from trades observed on the venues where bitcoins trade.

Of course this depends on the BRR being accurate and not manipulated, and this is the bit I don't understand. High among the advantages people cite for bitcoins is that they're an anonymous way of moving cash around. If that is true, they are presumably attractive to organised criminals. So the BRR is going to be based in some degree on the activities of criminals. The scope for market abuse must be huge, especially given the illiquidity and anonymity in the underlying. If the market dumps or spikes on tiny volumes, which apparently it does, thejn you could have someone who smashes the price at 4pm on small volumes of bitcoins that flow into the BRR, and are then used to settle a position he holds that is many times larger in CME's bitcoin futures contract. Since bitcoin cash players certainly include criminals, this risk must be heightened.

Exchanges routinely monitor for this by comparing who's active in the cash market with who's got big positions in the future, but CME can't do that because the trades that create the bitcoin cash price are themselves anonymous.

Somehow they've persuaded the CFTC that this is kosher but I don't know how they will investigate if someone claims that the price of their future has been manipulated. They will simply not have access to the information required to verify or refute the claim.

westernpromise

economic wrote: »

Yes it would be SPAN which is effectively based on MC but using historical data of BTC to calibrate the MC model?

Yes its the clearing houses that take on the counterparty risk not exchanges (they just take execution risk). Then you have clearing members which would probably slap on a multiplier of some sort on the SPAN margin to charge their clients depending on client risk and the state of the BTC market.

What i am curious is how they model and calibrate the model to produce the SPAN numbers given it is BTC (8 years of history). I presume at the start of the BTC (back in say 2010) we had seen daily moves like -75% or -50%. Will these moves be taken into account? Is it weighted so that the latest data is more influenced in the SPAN numbers? Is it normal or lognormal returns that are being used?

EDIT: Sorry i have not worked on futures or clearing houses, just stuff i had been reading a whilst back, hence the silly questions!

Clearing houses' risk models have to be approved by whoever oversees them (the BoE in this country) and would be expected to use up to date thinking on how to do this. But broadly I think clearing houses would apply EWMA, which is exponentially weighted moving averages, to historicals when doing their MC on it. This would mean that the 50 or 75% price moves seen 5 years ago would have very little weight compared to yesterday's price move. The net of it is that you don't need much history as once you get past more than about a few months, you're diluting the old values almost out of the picture anyway.

This must be a bit pro-cyclical of course - if something dumps in price overnight you have to margin all the shorts, who may then liquidate to pay the margin and make it dump more.

In a way the bitcoin future is not the significant step; you can have futures in all kinds of exotic or mundane things. What will be a sea change is if the clearing houses start accepting bitcoins as collateral.