Topic: [BTC-TC] Deprived Mining Speculation (DMS) - page 90. (Read 198958 times)

Yes that's the estimate.

When capital rises over 410 days of dividends for MINING, SELLING gets a dividend to reduce capital back to 400 days of cover.  That maintains over a year of cover at current difficulty for MINING (which will obviously last a LOT longer with difficulty rising) and returns unneeded backing cash to SELLING.

As difficulty rose massively this time, SELLING gets a pretty big dividend.  If difficulty rises by under about 3% on a change (exact percentage can vary slightly) then SELLING would get no dividend at all (in fact from memory think it's around 5% after a change where it got a dividend - as it has to make up not only the dividends to MINING but also the 400-410 day gap).

~.011 is a closer estimate, but yeah, that's right.

The current estimate is for a ~.01 dividend per share of DMS.SELLING? I must be understanding this wrong

Ah right - my misunderstanding.

The price used is whatever the system sets (I don't edit it).  I believe it uses the most recent 7-day average - but aren't honestly sure.  Looking at recent outgoing ones in my records it definitely looks more like its a 7 or 30 day average than a 24-hour one (as we haven't been going over 7 days the 7 and 30 day averaeges are the same - so can't tell which it is).

I think he means the price that is used as reference for portfolio calculations.

Although from the contract MINING should receive the new lower dividend tomorrow it won't.

When I wrote the contract I decided to go with a flat calculation based on difficulty 16 hours earlier - meaning that, on average, MINING would be slightly better off (compared to a PMB) than if I calculated exactly based on the time difficulty changed.  Of course sod's law says that the first (and likely biggest in impact on dividend) drop would occur in the other 8 hours.

I also note that TAT.VM pays on 24 hours previously - so would be keeping the old dividend.  If we pay differently it may confuse those trying to compare - which obviously I wouldn't want.

I will therefore be paying at the old difficulty BUT the difference between the contractually obliged dividend and the one paid will be refunded from my own pocket.  I will do this on ALL future MINING dividends where the same issue arises (which should be about one third of the ones immediately after a difficulty change where difficulty rose).

In short : The report will be as though the new dividend rate was paid.  The old dividend rate will actually be paid.  The difference between the two will come from me - not from SELLING investors' dividend.  I'll report the amount refunded by me obviously.

Every PURCHASE sent to me receives 1 MINING and 1 SELLING in return.  Prices have no effect on that.

Deprived, if I transfer the PURCHASE, what will be the price of MINING and SELLING?
Split in half or based on the current Asks price?

The below is NOT an actual report - it's an estimate of what tomorrow's report will look like

If the report were produced now this is what it would look like.  In practice it's likely more PURCHASE will sell - so starting NAV/U will be very slightly higher and the dividend to SELLING slightly higher.  That difference will likely be very tiny.

Sold   1056 (377 trade-in for LTC-ATF.B1 and 679 market sales)
Price   0.066149
Total   69.853344
Less Fee   69.71363731
Man Fee   2.091409119

Management fee of 2.0914 BTC paid.

BTC Balance (BTC-TC)    797.31840000
12500 LTC-ATF.B1    125.00000000
TOTAL ASSETS    922.31840000
   
Outstanding MINING   14176
Outstanding SELLING   14176
Outstanding PURCHASE   446
Effective Units   14622
   
Block reward   25
Difficulty   19,339,258
Hashes per MINING   5000000
   
Daily Dividend    0.00013002
50 days (Min Liquid)    0.00650111
100 days (Forced Close)    0.01300222
365 days (Buyback)    0.04745810
405 days (IPO)    0.05265898
400 days (Post SELLING div)    0.05200887
410 days (Pre SELLING div)    0.05330910
   
NAV Post MINING Div    920.41721561
NAV/U Post MINING Div    0.06294742
Days Dividend Post Div   484.13
SELLING Dividend    0.01093855
NAV Post SELLING Div    760.47375584
NAV/U Post Selling Div    0.05200887
PURCHASE selling price    0.05460932
PURCHASE buy-back price    0.05096870

This is provided for guidance - it is not necessarily the actual payments (MINING should be correct).  No management fee has actually been transferred.

I don't mind people being resistant to lowering the price of MINING. In fact, I expect that people will value MINING higher than its actual value. I know I did at first, when I just strolled in and went "hey, 5 MHash PMB for cheap, cool" without looking further. I had some coins invested in LTC PMBs, which are a lot more stable in a way since the LTC difficulty isn't spiking in quite the same way. So I just extrapolated that to BTC without giving it further thought. I expect that many people will still see MINING as a lucrative "PMB", when compared to other BTC PMBs.Time will tell.

Understandable reasons, but I wouldn't have minded some debate either

Anyway, 4 blocks away from the diff-readjustment, so it's extremely likely to happen before midnight GMT. I'm very curious to what happens tomorrow and the days after with the prices of the securities. My compliments for making this security, since regardless of whether I win or lose, it's been fun thinking about it and running some numbers.

That's fair enough - my own investments/holdings mainly came from solo-mining LTC for a few days then trading it up.  I think you'd find there's a lot more resistance to a significantly lower MINING price than you anticipate.

Part of the motivation for developing DMS was to see whether I could (and what happened when it was used) - when I paid the 15 BTC to register the assets I was by no means certain there'd be enough interest to even recoup that (though obviously I hoped to get that back in management fees and make some change trading it myself).

I'd love to debate your conclusions - but I can't as:

a) I DO have some reasonable-sized investment in MINING and SELLING (the 125 BTC is exchanged in for LTC-ATF.B1 plus some more I bought with cash) - so last thing I want to do is convince everyone to agree with me
b) It would ruin the experiment from my perspective.
c) My views may well get given more credibility because I run the fund - which would be wrong (as the math of how everything works is transparent - so I have no advantage in knowledge) but understandable.  If people get it wrong they must do so based on their own (or other people's) predictions - not mine.

I've taken the position that I want to take in the securities so I'm good for now. My BTC investments are basically just play-money (only ~2 BTC that I got mining some crappy altcoins for 2 afternoons). If I invest a significant amount of my savings (and not get murdered by my wife ), it could shift the price significantly (since volume is quite low) and potentially ruin most of my profits (not to mention that I prefer not to put too much into something as volatile as BTC, since the life of me and my family still runs on euros, not BTC).

Also, I'm a mathematician, I like the puzzle more than the what I can do with the result

One comment on this - and any similar posts which claim to calculate a value for MINING that's significantly different to market price.

IF you're sure you're right then why are you telling everyone rather than using that knowledge to make some profit?

I don't have a problem with anyone posting their calculations here - just don't expect any feedback from anyone who believes you're wrong and is actually investing.  The +EV thing for them is to agree with you if they believe you're wrong - hope other people believe you - and bet against them in the market (doesn't matter which way they think you're wrong - they can bet against you).

It won't acutally end up the cheapest as there's one or two PMB-like things around (shares paying out based on a fixed amount of promised hashing) where the issuer has either scammed in the past or is believed to be a scammer.  Those will always be the cheapest.

MINING will stay cheaper than most - largely because, as you pointed out, people can be sellers as well as buyers here.

Someone gos to a PMB and believes it'll never return capital at the current price all they can do is make a post in the thread.
Someone looks at MINING and believe's it'll never return capital they can buy a PURCHASE, convert it and sell the MINING part to someone else.

Obviously they have to believe there's enough of a margin to cover my fee and to justify tieing up their cash.  And they can, of course, be wrong - and be donating to whoever buys MINING from them.

There's also a few other reasons why MINING will stay cheaper than comparable PMBs.  One of the big ones is psychological.  MINING is explicitly referred to by me as speculation - I don't pretend it's a bond and I don't ever claim investing in it will make a profit at any specific price.  So people read thread - or even just the title - and think "this is risky".  And rightly so - it IS risky.  But they'll then go to some PMB which gives the same returns and buy it for more - as it gives the misleading impression of being some sort of bond where capital is safe and they just sit back and get dividends for free that the issuer is (for whatever charitable reason) throwing in their general direction.  Being honest about the speculative and risky nature of MINING (you buy at the wrong price you make a loss) is definitely something which will keep its price below comparable PMBs.

Whether it's profitable or not at any price point isn't something I'm ever going to comment on (other than that I've explicitly capped its maximum value at the price of a PURCHASE - and it should never trade above 365 days dividend).  But I can safely say that it'll make more profit or lose less than an equivalent amount of hashing from a PMB bought at a large markup in terms of cost per MH/s.

Lets assume that the difficulty increases by a factor X every 2016 blocks (so X = 0.1 would mean +10% difficulty). An increase of X means that the last 2016 blocks were done in 14 / (1 + X) days (target is 14 days per readjustment, and the new difficulty is calculated based on the actual elapsed time to solve the previous 2016 blocks).

Starting with the first day after a readjustment, say MINING has a dividend of D, which is paid daily until the next readjustment at which point it is updated. So we have 14 / (1 + X) days of a payout of D. After that, the payout becomes D / (1 + X) as the dividend scales inversely with difficulty. Again 14 / (1 + X) days later, payout drops to D / (1 + X)^2. And so forth. Our total payout for MINING (or any PMB without reinvestment or other special properties) is:

14 / (1 + X) * D + 14 / (1 + X) * D / (1 + X) + 14 / (1 + X) * D / (1 + X)^2 + ...
= (14 / (1 + X)) * D * Sum_N=0^infinity 1 / (1 + X)^N
The summation is actually the sum of a geometric series and a convenient closed formula exists for it:
Sum_N=0^infinity r^N = 1 / (1 - r)
(for 0 < r < 1)
So in our case, the summation part of the equation adds up to:
1 / (1 - (1 / (1 + X))) = (1 + X) / X
which makes our final value of MINING dividend from now until the end of time:
14 * D / X
where D is the initial dividend payment per day and X the factor with which the difficulty goes up every 2016 blocks.

If we plug in some actual numbers... Assuming that the difficulty goes up to 19,000,000 tonight (which is a reasonabe, possibly conservative estimate), D = 0.132344 mBTC (I use mili-bitcoin or mBTC, because numbers close to unity are easier to work with mentally for most people). That means that with a 20% difficulty jump every readjustment moment, the total payout from MINING will be:
14 * 0.132344 mBTC / 0.2 = 9.26408 mBTC.

And that is all you should pay for MINING if you believe that 20% increase per period is what'll happen. In fact, you don't want to get MINING for this amount, because it'll take forever (literally) to earn it all back and you'll end up with no profit in the end. The NAV/U of the fund will be approximately 54 mBTC after tonights adjustment and the dividend payment to SELLING, so the price for SELLING should be roughly 45 mBTC *after* the dividend payment.

If you take the very conservative estimate of a 10% increase in difficulty, the same calculation gives us a lifetime payout for MINING of 18.53 mBTC.

Now these calculations assume exponential growth of the difficulty, which in the long term is not very realistic. However, it doesn't matter that much, because for the short to mid term this exponential growth is a decent enough assumption and by the time the growth of the difficulty will flatten, the expected payouts from MINING will already be very low. With a depleting asset such as a PMB, the largest gains are had at the start, so the behavior of the difficulty on the short to mid term is important right now, while long term is mostly relevant if there's anything left after the dust clears.

Since the start of the year, the difficulty has increased by 15% every increment on average if you assume an exponential growth. However, the %-growth seems to have picked up recently.

DMS.Mining will most likely get to be the best-priced PMB there is, because there is more pressure to find a fair price for it.
This also makes it very difficult to predict. How many people will drop their PMB for this one?
The very long description and unusual structure will scare of some of them.

If you want to wonder what people are thinking, consider the ones paying double or more what TAT.VM costs (and an even bigger markup compared to MINING) per MH for 'real' PMBs - that's ones where they have to rely on hardware not failing for the operator to be able to afford to pay ANYTHING out.  Think the absolute worst ones (for value) are on Bitfunder - PMBs which aren't even mining yet lol.

For that matter you can even compare MINING to shares with reinvestment - reinvestment doesn't work that much differently to buying more MINING with your dividends (with a good manager the shares will perform slightly better).  And if PMBs are bad value there's some shares which are actually WORSE value (at the prices they currently trade at) - as the majority of their real value was lost by holding their capital as USD in pre-orders for BFL machines whilst BTC rose.

The absolute worst (in terms of value per hash) are ones where you buy a share of actual fixed hardware.  Those act like PMBs - except they stop paying when the hardware fails and they don't have any guaranteed payments at all (so they're even less perpetual and even further from being bonds).  Those should sell for LESS than PMBs per hash - but somehow manage to get away with selling for more sometimes.

I really like to know what someone, who buys MINING for > 0.03 (or e.g. TAT.VM for 0.007), estimates about the average rise in difficulty in the next months. It has to be < 10% per change, otherwise there'd be a very little chance he will ever make back what he paid IMHO. And right now we have changes > 20%, without having some other big players beside Asicminer to enter the market. I guess as soon as this happens, difficulty will rise even more and all the PMBs will implode...

On SELLING you can only safely assume it to be profit if the total dividends you've received exceed what you paid for your SELLING.  In theory at any point difficulty can fall, stop rising (or rise by less than about 3% per change) meaning no dividend for SELLING.  That won't happen in the medium-term (unless you believe BFL will never ship, Avalon chips won't arrive, batch 3 Avalon won't be sent AND none of the other ASIC manufacturers will deliver).  In the short-term (next dividend or two) it's possible - as ASICMINER aren't shipping at present.  In the long-term it likely WILL happen (when ASICs reach saturation point) unless BTC rises vs USD significantly.

Only realistic scenarion in which SELLING receive no more dividends ever is if there's some massive problem with BTC causing the exchange-rate to fall to the level where it's not profitable to mine with ASICs other than with cheap power.  That scenario is one of the few where PMBs WOULD pay (a LOT - in that case) more than MINING (assuming the operator paid out even when he couldn't mine) - but that would be little comfort if BTC was basically dead.

Yes - and that's true for EVERY SINGLE SECURITY (other than scams where ANY price is too much).

The price you buy for is what determines whether it's a good move buying it or not.  It's not a Good investment or a Bad one or a Profitable one or an unprofitable one unless you talk about at a specific price.  The WORST question I see asked regularly is "Is X a good investment?"  Without a price it's totally impossible to answer - yet the forum is full of idiots who will proclaim something as good without stating the price at which that claim applies (it's fine during an IPO where there's a fixed price).