Yes, but we can't ignore the fact that just in the last month the payout has dropped about 40% so whether we make ANY return from our investment after a year is a big question.  Further to the charts I posted on June 1st I have managed to extract further data from the blockchain which puts it into better perspective.  The following chart is compiled from my own data (starting May 10th) combined with payouts to three other GM user addresses going back to around February 21st.  It seems to confirm the point that I called "Stefan's Data Point" in my previous post.



Clearly the average downtrend has not always been as steep as it has been for the last month, which is good news.  Maybe there is some reason to hope that the past month was not typical and the payout will rise again - or at least drop off at a shallower angle.

As a matter of interest does anybody know why the volatility has dropped off in the last six days?  This is a chart of my payouts since I started with GM.  Something seems to have changed six days ago. 



Not complaining - just curious.

The last pay-out in my transactions list on the GM website is dated 06/06/14, but the actual pay-out was 06/07/14 (transaction ID ad720b79cbd7d4a640bea245523bdc7177040a2b58b4e18ce3a7be6dbea44b5f). If this was also your last pay-out everything is normal.  Expect a new pay-out later today - the exact time varies.

I hope you are right.  My point is that, with the current rate of returns at an annual rate which would more than double the initial investment, it is tempting to let greed get the better of caution.  It is now clear to me that making a profit from mining, even with a good, honest service like Genesis Mining, is by no means a certainty.  I was about to double my investment until I crunched the numbers, but now I will hold back and see what happens for at least a month or two.  If the current down-trend in returns is short lived I will be back for more.  I would also be much more inclined to increase my investment if GM would make a firm commitment to proportionally increase the hashpower of existing customers if and when you drop the prices for new contracts.  You did exactly that on May 15th, which seems to me a very good way to keep your customers loyal, and if I could be sure that this was company policy, not just a one-off, I would be buying new contracts now.

Could be that GM have not already prioritized showing historical data because it is actually not all that encouraging.  I have been mining with GM since May 10th and I am prepared to share my data with the community.  I started with a 5MH/s contract on May 10th which was upgraded free of charge (thank you GM) to 9 MH/s on May 15th.  I then bought another contract to give me extra hashpower on May 16th.  In order to make the results meaningful I will divide the BTC payouts by the hashpower on the particular day and multiply by 1000 to present them in mBTC/MH/s.  My mining has been set to 'auto' all the time.  Please note that I will not include the pay-out in fiat because that will complicate the results due to exchange rate variability.  I am not interested in the exchange rate because I invested BTC and I am being paid in BTC.  Some would argue that the results are better than I am showing because the price of BTC has gone up in the past week, but I could have kept my BTC in cold storage and received the same benefit.  So here are the returns to date (mBTC per MH/s):

5/10/2014   3.19
5/11/2014   3.77
5/12/2014   3.17
5/13/2014   4.02
5/14/2014   3.34
5/15/2014   3.34
5/16/2014   3.01
5/17/2014   2.91
5/18/2014   3.64
5/19/2014   2.42
5/20/2014   2.45
5/21/2014   2.73
5/22/2014   3.07
5/23/2014   2.56
5/24/2014   3.09
5/25/2014   2.33
5/26/2014   2.86
5/27/2014   2.86
5/28/2014   2.53
5/29/2014   2.04
5/30/2014   3.01

That is a total of 62.33 mBTC for each MH/s that I bought.  That would be 1083 mBTC after twelve months at this rate.  My average cost per MH/s was 325 mBTC so that would be better than 230% ROI.  Nice!

... Or is it?

Let's look at a graph of the results:



The trend is pretty obvious I think.  The trend line was drawn by Excel and it is the best fit to the data.  I always expected the returns to drop off over time, but not to that extent.  Let's put it in context.  GM are pretty coy with their historic data, but I have found one point of reference which I will call "Stefan's data Point" (SDP).  I found a screen shot of a GM dashboard posted by Stefan from Genesis Mining: https://bitcointalk.org/index.php?topic=467587.20.  On it you can see the hash power (35 MH/s), the output (roughly $120 per day) and the date.  Note that the returns graph was going negative at the top so I am assuming the screen shot was taken on March 8th but the last payout was March 7th.  The price of BTC averaged about $627 on March 7th so the pay-out would have been about 191.38 mBTC (0.19138 BTC).  Divided by 35 the pay-out would have been about 5.46 mBTC for each MH/s.  In the following graph I have extrapolated the trend line back to March 7th.  I have also drawn a line from SDP through my data and extrapolated it forward to the 0.89 mBTC/MH/s line.  The significance of 0.89 mBTC/MH/s is that I need to average above this line in order to make a profit over the year.  The cost of each MH/s was 325 mBTC.  325/365 = 0.89.



In fact a straight line decay is not likely because it would imply that the returns go to zero, which would only happen if GM went out of business.  More likely would be an exponential decay.  The following graph shows the best fit exponential decay from SDP through my data and extrapolated for the year beyond the start of my data.



So, if this turns out to be what will happen, what will my returns look like?



In this scenario the exponential curve is exactly the same as the one in the previous graph (using the same decay factor) except that the start point is where I began mining with GM and I am showing the accumulated pay-outs.  I have overlayed  my actual returns so far.  The end point in one year shows total pay-out of 365.72 mBTC per MH/s which is just a 12.27% ROI on my initial investment of 325.75 MBTC per MH/s.

I make no claims to be a mathematician or statistician.  If anyone can spot mistakes please accept my humble apologies in advance, and if you can interpret these figures better than me please go ahead.  I am happy to share my Excel spreadsheet with anyone who can help make this better. 

I have had an email exchange with Marco Streng to discuss this.  He has always replied politely (always a positive sign) but did not respond to all my questions.  He did not dispute my figures but pointed out that mining is a risky business and profits are not guaranteed!  I am aware of the dangers of extrapolating data forward a whole year based on less than one month of information, so the reality could be much better or worse than these graphs suggest, but I think it is worth sharing what I have seen so far.  12% ROI looks a bit marginal and even that could easily be wiped out.

Hi watcoin, nice idea.  I presume you are doing your arbitrage crypto:crypto, not crypto:fiat - right?  Otherwise how would you economically transfer fiat between exchanges?  Do you use any software aids to help find arbitrage opportunities?  In your experience are there any particular exchanges and currency pairs that are particularly good for arbing?