Further futures on MPEx
One of the more important economic activities in Bitcoin is of course mining, so either a hash rate based or difficulty based financial instrument is absolutely needed.
MPEx had for a timei a hash-based option, but that didn't work so well for the marketmaker and was discontinued. The experience has convinced me that designating marketmakers for hash-based derivatives is not functional, as the financial and technical knowledge required to successfully run such a market for long intervals is not available to humankind (currently, if at all). Still, because BTC can not be without a mining future, here's a novel attempt in the form of a cash-settled difficulty index set of futures :
- March : X.IDIFF.MAR
- June : X.IDIFF.JUN
- September : X.IDIFF.SEP
- December : X.IDIFF.DEC
They all settle on IMM datesii as the value of difficulty pure and simple. For instance X.IDIFF.DEC will most likely settle for 0.03370182 BTC exactly, seeing how difficulty only changes in ~1600 blocks, which is about 11 days, which is past the 19th.
Other than the settlement dates there are two major differences from other MPEx futuresiii, specifically :
- No insurance fee. This has the obvious disadvantage that contracts may be forcibly liquidated if the difficulty of contracts sold exceeds the seller's posted collateral, but there's no practical alternative. It is however possible that as the mining market matures some reasonable insurance scheme may be introduced.
- 2.9x capital per quarter. Other futures require 1.8x of last settlement value posted as collateral, X.IDIFF.* require 2.9x for each quarter out. Thus currently X.IDIFF.DEC requires 0.097735278 BTC per contract whereas X.IDIFF.SEP 2.38366569 BTC per contract.
Note that because of the forced liquidation explained above it is not advisable to pay more for a contract than its collateral value. Thus paying 3 BTC per X.IDIFF.SEP contract today will result in a guaranteed loss of at least 0.61633431 BTC per contract, because should difficulty indeed exceed 23.8mn those contracts will be liquidated for that value.
I hope these will prove useful for miners, their investors, and the BTC public in general. Good luck.
———- First trade on Sun, 22 Apr 2012 22:08:01 GMT, last trade on Sat, 25 Aug 2012 08:09:37 GMT. [↩]
- That is the third Wednesday of March, June, September and December, before midnight GMT. If anyone is curious why the other futures do not settle on a sane date (such as the IMM dates) the explanation is quite simply that it did not seem desirable to have MPEx settlement dates coincide with IRL settlement date for IRL-referenced futures, but it seemed preferable to always use the close on a future which is still at least a few weeks out. Hence the 4th Friday rule (which also conveniently matches MPEx option expirations). [↩]
- Three! Three major differences! [↩]
Friday, 14 December 2012
I believe that any sufficiently large miner can make relatively safe profits by selling these contracts in an amount derived by the following formula:
([easiest difficulty shares per minute] * ([current block reward] / [current difficulty])) * [minutes until settlement] * 2
They should sell at a price per contract greater than the projected expenses for the time period divided by the number of contracts.
The amount received in excess of projected expenses should be risk free profit. Which, btw, is received in advance of the work being done. Which enables the miner to buy more equipment and repeat the process.
Friday, 14 December 2012
Quite an interesting point. Let's try an example : suppose Bob has a farm doing 30Gh/s, and costs are 1 BTC / Gh / month for him. How much should he sell ? At what price ?
Friday, 14 December 2012
You'd need to convert Gh/s to shares/minute. Most mining software displays an average shares per minute which should be a more reliable number than the estimate I am about to use. I think 30 Gh/s equates to roughly 419 shares per minute.
If my assumptions are correct (and I haven't spent much time scrutinizing them) this boils down to:
(419×(25÷3370182))×8603×2 ~=54 contracts
(1×30)×.199369213 ~= 6 / 54 = 0.11076067 BTC per contract to break even.
This seems high, but that might be due to the cost / GH / month you've chosen. I you use .3 in place of 1, it calcs a break even number at 0.03322320 BTC per contract.
Friday, 14 December 2012
Pretty good stuff!
Friday, 14 December 2012
(I was using the X.IDIFF.DEC in that example)
Friday, 14 December 2012
Turns out that I was working with flawed assumptions :o
All that stuff in my previous comments is wrong.
Saturday, 15 December 2012
Come back, 1 yr!