Quote from: Findus on March 28, 2014, 04:09:48 PM

Robitnik, I'm really greatful for your summarizing the history.

I think that the crux of what I don't understand is here:

Quote from: robitnik on March 28, 2014, 02:49:24 PM

If COG.F/2 shares were instantly converted to ordinary shares at the time of purchase (i.e. when we made the deal for the hardware), the value of the hardware per share would have been 650/14420 = 0.045 BTC per share.

that means that when we made the deal for the hardware, each 5 BTC COG.F2 share immediately became 20 x 0.045 = 0.9 BTC worth of COG shares?
Because the investment of COG.F2 investors was dissolved with previous COG owners? Is that fair?

No problem. It's a complicated situation.

If they were converted immediately (they weren't) they would have been worth that much in hardware. The share price is usually worth more than just the price of the hardware though. COG.F and COG.F2 were essentially a bet that the share price would be greater than 0.25 BTC when they converted which seemed very plausible at the time. There are a few factors that run the price up.

That hardware (if it was running) produces bitcoins, so the dividend yield increases the price of a share.
Acting as a group allows us to make better deals (i.e. the $500 discount per unit) than if we were buying alone, this increases the share price.
Because we have many machines running rather than say just 1 each, a single machine's failure is not catastrophic & the risk is spread.
We have an economy of scale when it comes to other things like hosting, power and management which all add to the value.

As I said in a previous post, historical data for BTC mining securities suggest that people typically value an asset like this at between 3 and 10 times its annual yield (an "interest rate" of 10-30% in BTC terms).
i.e. if the dividend changes, the share price will change to ensure the following is true: 0.3 > ((weekly div)*52/(share price)) < 0.1
This has generally been the case with some exceptions resulting in spikes up and down depending on the disaster of the week.

At the current difficulty, 20 TH/s (if running) will produce ~14 BTC per week which (after reinvestment) yields 0.5 mBTC per week. If there was faith in Cognitive going forward, that would traditionally have supported a share price of between 0.078 and 0.26 BTC per share. We also have another 10 TH/s to be delivered.

Where it went wrong was Garrett waiting too long for the Washington data center. The second Cointerra started delivering, we should have been deploying. There should have been an intermediary datacenter in place beforehand. That's what is being done now. If we decide to keep this show going, we'll have to take measures to ensure something like that doesn't happen again. As I said elsewhere, I expect the difficulty growth rate to slow, which means our current hardware will be running (and paying off) for quite a while. If we can manage even one more good deal and deploy it quickly, Cognitive will be set for a long time.

But all that depends on whether shareholders decide to liquidate now for ~0.02 BTC per share or not. If cog keeps going, it will also depend on us figuring out a better way to manage it. YoYa suggested setting up a board consisting of Finance, Mining and PR which seems like a good idea to me.