Bitcoin Forum

Soo..
Bitcoin doubled value in one month.
How long will this one last?
 ;D

You call this a bubble? Well, last time it was doubling every week ;D

Yeah, this is far from a bubble.  We're simply recovering from 32 and still have a long way to go.

Fundamentals bear here.  It's still overvalued based on current need and IMO the current rally is based on pure speculation...  IE, it's a speculative bubble.

I speculate that Bubble2 won't be as extreme or last as long as Bubble1.  I give it a few months before we hit $2.00 again.

If it busts through $7 it might go to $12

However, I think we will probably need to re-test $2.20 before we start looking for new all time highs (they will come though, in time).

And whats the quantitative "fundamentals"?  I suggest looking at difficulty, which if you compare to the last time bitcoin was at $2-$5, say that its currently undervalued.

Difficulty is an effect of price, not a cause.

I define the fundamental price as what is required to store sufficient value to support current commerce.  The price cannot be sustained lower since non-speculating Bitcoin users will bid it up to fill their needs.  Speculation can drive the price higher, but only as long as there is a constant influx of investment - the sum of (miners speculatively holding) + (net increase in fiat).

"Feeding the miners" (with either hope or fiat from people with hope) becomes more expensive with a rising price - at the current $5, we have to find over $1M per month of new hopes and dreams just to keep things afloat.  Getting up to $30 would require over $6M per month of influx.

So where's the floor?  Read this plus what I linked at the bottom (https://bitcointalk.org/index.php?topic=49396.0).

tl;dr: The floor is a few cents, and a reasonable level of speculation would put us at 10x the floor - perhaps $0.50.  Any price above that requires either wild speculation (what we have now), or an increase in commerce.

Hype will get more of the former and thus a bigger, longer bubble; more people doing real commerce will cause the latter, and enable a sustainable long-term growth.  Unfortunately I see mostly the former and very little of the latter.  Until that changes, I'm keeping the bear hat on.

The relation between price and difficulty is a two-way causality.  The fact that difficulty exploded before there was even a market or price for bitcoin falsifies the one-way hypothesis.

As for a fundamental, your definition only makes sense if you assume that the utility of bitcoin is strictly supporting commerce.  I assume that the utility also includes the buying and selling of dollars, euros, etc. (speculation).  Its a value storage and transfer system; a risky asset a la gold.  Its not purely a transactional medium, its an actual commodity.

Difficutly is not a byproduct of the bitcoin.  It is the essence of the bitcoin.  But we're beating a dead bear now.

Price is just a proxy for value.  Before there was a price on an exchange, there was still stored value.  It was harder to quantify, but it still conceptually existed, and would fluctuate with need.  Without even buying or selling actual goods, the demand for novelty was enough to support the very low value per coin at that time.

At $5, you need a lot more than novelty.

Yes, I'm basically modeling it as a transactional currency, and considering the store of value as one aspect of a transaction (all stored value is intended to be spent eventually).  That's basically true for now: the value is too unstable to be a reliable value store, so people seeking a transactional currency won't store much money in BTC.

That's a weakness in the model: a very large and hard to quantify part of the formula is the amount of BTC held long for speculative value.  The actual transactions are small, and the computed value can vary a lot depending how you fudge the variables.

On the other hand, that's also it's strength: any attempt to compute the value based solely on the MAIN current use of Bitcoins (as a speculation vehicle) is completely detached and self-referential, and can be used to justify any price.  Off to the moon we go.

Edit:  Oh, don't get me wrong.  It's going up short-term.  I just don't see it lasting for more than a few months before reality sets in again.

Its straightforward to quantify the difficulty, matter of fact its built into the protocol.  And not coinicidentally, its mich higher now at $5.


Try computing the value using difficulty as one of the variables.  Prices at the moon are only justified if difficulty follows.  When it doesn't, there are spikes in the price:difficulty ratio, and it stays grounded at realistic prices.

If the price goes up, more miners come online and boost difficulty.  If it drops, they get out because it's not worth the power and difficulty falls.

The compliment happens to a degree: if difficulty is high, some people will buy coins at the market instead of mining them, thus boosting the price.

I think the first effect completely overpowers the second, so difficulty is a lagging indicator of price.  And so far, it looks that way to me:

http://bitcoinx.com/charts/chart_large_lin.png

Do you have any evidence to the contrary?

I said as much a long time ago:

Good, we're on the same page, then.

I look at it like an SMA with some upward bias: after the price levels off at bottom an SMA will keep falling, whereas difficulty goes flat quicker (some miners keep mining even when somewhat unprofitable); when the price then goes up a level SMA will start turning up quickly whereas the difficulty lags a bit more (a few miners are willing to start when it's less-unprofitable, but most wait until the price goes high enough to make profit at a difficulty level).

In the end, it's just a lagging indicator of price, just with some extra factors thrown in ($/kW, MH/J, miners' inertia to start or stop mining, available MHps offline after a price drop, the $/MHps for new hardware) that aren't relevant to the fundamental value of a coin.

So for pricing purposes, how is it any more enlightening than an SMA45?

You both seem to be off in the weeds arguing a logical fallacy concerning correlation and causation.

Ludwig von Mises (http://books.google.com/books?id=bqhZRn5zWA4C&lpg=PR2&dq=human%20action&pg=PA55#v=onepage&q&f=false):

This should be tattooed on the hand of any person that is about to speak authoritatively about economics.  Teachers, TV news fodder, drunks in bars, internet idiots, etc.

Trace Mayer or Bill Rounds?

Welcome to the World of the Gold/Silver Replacement:  Bitcoin.

edit:  don't miss this ramp Trace.  this time it isn't coming back.

Your quote is about the fact that the market is too complicated to model accurately, and too fleeting to easily test a theory with any respectable degree of scientific rigor.  I entirely agree - it's very hard to show causation when any given correlation could be the side effect of a near-infinite supply of other factors.

Nonetheless, it is not a logical fallacy to suggest that there are causative relationships in the market.  They're there!  They're just hard to prove.  The best we can do is theorize, then see how well our theories predict the future.  That's the whole basis of analysis.

Difficulty clearly correlates strongly with price.  It's not direct due to all the other factors I mentioned, but there is a causative relationship somewhere in there.

Post hoc ergo propter hoc. I am going to isolate the relevant sentences and bold the key part so you can read and hopefully understand since you missed it the first time.


We are talking about economics with human action and not physics. But I will concede that despite the complexities and unknowables we can know something and from that something derive tendencies.

spoken like a lawyer.  probably Bill.  welcome.

All the lawyers disappeared.


The NDAA was signed into law. The BTC price in USD went up. Therefore, when legislation is signed into law the price of BTC in USD goes up.

oh so encouraging.  i especially liked this part:

"If President Obama were committed to Constitution and international legal norms, he would veto this bill. Instead, he seems more concerned about consolidating the power of the Executive Branch at the cost of our legal and human rights. As “terrorism” and “radical Islam” have come to replace “Communism” in the federal government’s lexicon of fear, the United States continues its spiral toward a new era of McCarthyism. The NDAA is one more step down that road."

Buy more Bitcoin.

I do understand the fallacy.  Correlation != causation.  I've said it in great frustration many times.  But now I feel like I've fallen into a bizarre alternate reality where that message got repeated too often, and everyone has now concluded: "... and therefore, causality does not exist."


I agree with everything you just said there.  I absolutely agree that the relationships are not constant.  But working within your concession, and with a year of good data that coarsely supports the theory, I think it's fair to postulate ergo propter hoc, and that I'm not running afoul of a fallacy in doing so.

I sincerely doubt it.

2.00 IMO was a residual effect of the pain and anguish from previous months.  The market wouldn't offer coins for lower prices despite some really stagnant trading days.

I'd be really surprised to see 2.00 hit again ever. 

Plus... the press is out there.  Once the market cap flys high and people drop a couple hundred million into mining equipment, expect to see fewer fluxuations and thus viability brought to it as a transactional medium (which ought to further stabilize price).

For the record, however, I will be selling probably 25ish percent of my holdings at $30.

Worst case, I'll be up 100% on my investment :D

I would hope that the price remains stable at $ 6.50
that's a fair price for all

If Bitcoin is widely adopted, yes, but remember where that $200M comes from: inflation (IE, devaluing your coins) and to a smaller degree, transaction fees.  We have to get to $10B in commerce before that cost drops under 2% of all money flow.

It's a plausible outcome, but it's an optimistic one.  I consider these scenarios equally likely: Bitcoin remains an interesting but niche currency, and just wallows around unable to sustain that high of a market cap; or another cryptocurrency with more appealing properties comes along and all but replaces Bitcoin.


Worst case, we have either of the above scenarios and never reach $30, and you lose as much as you can tolerate before you capitulate.

In a better case, wild speculation doesn't get us to $30, and we have to to slowly, gradually achieve it with commerce.  You make your investment goal, but it takes 5 years.  15% ROR is OK, can't complain, right?  It's better than losing it, but it's not exactly stellar success for a very high risk investment.

Or Bitcoin could be wildly successful.  I'm optimistic, but I still need to see real, widespread adoption before I become confident.  Don't lull yourself into thinking that it's too good to fail.

The only reliable correlation between any cause and BTC price I've ever seen are google trend vs price-charts posted by some older users here.
Much too big of a coincidence to be that, a coincidence. BTC price is, at least historically, strongly linked to it's popularity in mainstream media.

The news references and price increases/falls go almost hand in hand. Someone can probably post a chart, I don't have any saved.

You don't see any correlation in the price vs difficulty chart I posted?

And yes, I know how the news correlates...  I'm responsible for a couple of those charts.  :)

Your graph clearly does not show the relationship that you are claiming.  Difficulty does not equal price times a factor, even with a delay.  It doesn't even equal price times a polynomial.

You need a more complex model.  There are more than two quantities, and the connections between them aren't linear.  For one thing, the amount of mining that happens is correlated to the value that miners see in bitcoin, which is only weakly related to the price.

The simple model has no predictive power.  The price has just gone from $2 to $5 for the second time ever.  Look closely at your chart and ask it what the difficulty is going to be, and when.  Is it going to triple in a few weeks?  Or is it going to increase by a hundred thousand or so?  Seven months ago, tripling and increasing by about 100,000 were the same thing.  This time, they are not.  And look at November and December, the price and difficulty seem to be decreasing and then increasing, but this time together.  What happened to the lag?

An increase in price can cause an increase in difficulty, but it is far from clear that it will always.  Like ol' Ludwig said, your graph shows a unique historical fact, not an eternal principle.

Miners are turning back on their idle rigs.

I didn't say it was linear, constant, or that there were only two quantities:


Those are the major current ones, but more could be added.


That's why I list "available MHps offline after a price drop" as an input.


My chart shows that a relationship exists.  I claim that price is one variable that feeds forward into difficulty, not that it's a simple linear relationship.  I do not claim it's an eternal principle.  Quit trying to spin it like I am.

...

And you're entirely missing the point.  Here's what started this whole thing:


bitcoinBull thinks I should be using difficulty to compute price.

All I'm saying is is that there are both logical reasons and rough empirical evidence showing that price feeds forward into difficulty (not that it's the only factor; not that it's an eternal principle; not that it's a linear relationship; simply that it's one variable that feeds forward) - that there is some kind of causal relationship in that direction.

...  But that I don't see any evidence that difficulty has any predictive power for price, and I don't see any logical reason why it should be a significant variable in an idealized formula for fundamental price.

And I'm asking why he thinks it should be.

If you look carefully at the "mining factor" chart, in the month of May there are two spikes.  Both spikes were the price rising, but the valley between them was not due to a dip in price, but to a rise in difficulty.  The second spike in May, may not have happened without the prior rise in difficulty (cum hoc (http://en.wikipedia.org/wiki/Post_hoc_ergo_propter_hoc)).

Similarly, the current rally is blowing past the SMA45, which topped at $20 and bottomed at $3.  But if you compare price to the difficulty, which topped at 2.0m and bottomed at 1.1m, its still only catching up.  

Lately I've been thinking of difficulty as having a "pull" effect on the price, rather than a "push".


It makes good sense to me that difficulty has predictive power.  If the network had decayed and difficulty plummeted much more than the ~50% from its peak, wouldn't that have made the recent price reversal all the less likely?

If price is a measure of demand, then difficulty is a measure of supply (high difficulty = low supply).  But its also a measure of the network security and robustness.  The higher it gets, the less chance there is of, e.g. a 51% attack.  If the "value" of bitcoin is the resilience of its distributed sytem, then difficulty is its best measure.

I price bitcoin according to its ability to store value in a decentralized system where no central authority can seize, freeze, or inflate my coins.  I care much less about the ability to do everyday commerce with bitcoins.  I want an alternative to PMs, real estate, and shares of public companies in which to invest/trade my money.  So more important to me than the number of merchants, is the liquidity (which need not be enormous since I don't have a ton of money).

So an idealized formula for my fundamental price P_f would factor in difficulty, exchange volume (or net volume) and bid depth, and maybe bitcoin days destroyed (another measure of decentralized activity - I look at it as bitcoins exchanged between exchanges).

Because I'm mostly interested in trading dollars/euros for bitcoins, not bitcoins for stuff, the bitcoin "economy" I'm concerned with is primarily the exchange volume.  Somebody else might care more about the number of merchants.

Except that you are wrong; your chart shows the exact opposite.  Do you see the huge area between the red and blue lines to the right?  It tells you that the correlation is very low.  Actually, worse than that, it tells you that the amount of correlation actually gets worse as one of the variables changes.  Price seems to lead difficulty when difficulty is low but appears almost totally unrelated when difficulty is high.  Do you see it?  You've found an indicator that predicts nothing more than its own invalidity.


Well, obviously, he's wrong too.   ;)

He's wrong because he forgot the first law of technical analysis:  Your perfect indicator becomes useless as soon as you explain it to someone else.  He was using difficulty as a heuristic to figure out if market moves were "legit" or not.  Actually, that indicator was crap from the start, but the joke was still funny.  If you look at your chart, you can see that this heuristic would have signaled "buy" all the way down from $30+ to $6.

Oh, here's a fun one.  Try to look at your chart and predict what the price will be when (or before) the difficulty passes the previous peak.  Will it be like $8 or $10, just enough to make rigs operationally viable with even relatively expensive power costs (ignoring capital costs)?  Or will it be closer to $30, the price that (maybe sorta) sparked the massive capital investment that allowed the difficulty to attain the previous peak the first time?  Or will it be more like $50 or $100 or beyond, high enough to make the last bubble seem tiny and pathetic and convince everyone's grandmother that she needs to get in on this too?

I would be really nervous about making that prediction, because the two charts show a spike and an almost square digital 0 -> 1 transition.  With the right scaling factors, those two will always look like the early one caused the late one, and the price spike almost certainly really did cause the difficulty jump.  But what happens next?  Was difficulty primed before, like supercritical water, just waiting for a nucleation site to boil over, but now spent?  Or is mining primed now, with the people that missed the first bubble now waiting for the ROI to get short enough to buy their own rigs?  Or is it something else?

Try not to take this so seriously man.  I wasn't trying to be insulting, I just don't think that you know the things that you think you know, or at the very least, you don't know the things you do know for the reasons you think you know them.  I grew up on fractals.  Check out the chaos that emerges from the incredibly simple Lotka-Voltera (http://www.nahee.com/spanky/www/fractint/volterra-lotka_type.html) equations, and shudder at what equations you'd need to really model the financial, technical and psychological properties of bitcoin.

Well, no, the supply is constant.  Difficulty is a measure of demand.


The 51% attack is only one threat to Bitcoin.  The network would be adequately defended against double-spends at 5% the current difficulty (it had better be, or we'll have big problems when generation starts dropping).

IMO key management is a much bigger threat to security, and by proxy, confidence.  It's not measured in the difficulty at all.

Thanks though.  I at least get where you're coming from: that greater security of the system creates more intrinsic value in the system, and therefore justifies a higher price.  I'm not sure if that's properly called a Pf, but it's an interesting concept.  I'll give it some thought.

To me, it says that MH/J (more efficient hardware) and J/$ (miners finding lower cost power) has increased.  It's a scaling factor for the red line.  It's been increasing over time.  Yes, that means the correlation is not constant or linear.  It'd line up closer in line if we corrected for it, but I don't have hard data to plug in.

Instead, apply a general correction:  take the derivative of each and you'll find they correlate pretty well (fudging for some variable lag).

Here's a log chart so you can see the old data better:

http://bitcoinx.com/charts/chart_large_log.png

Every bump up in the blue line is followed in about 4 weeks by an increase on the red line.  Drops in blue cause red to flatten, though it doesn't fall easily since a) miners' sunk costs mean they generally stay online during mild drops, and there was a long overshoot in June-July which took a long time to correct; b) MH/J (and other efficiency and mining cost factors) is increasing.

It's also particularly skewed in late 2010 due to the switch from CPU to GPU mining.

None of that is corrected in the charts, which is why I say that price is ONE factor feeding difficulty, not THE factor.



Don't look at low vs high.  Look at rising vs falling.  On the Difficulty side it's more rising a lot vs only rising a little (again, the MH/J skew, among several other factors).



I'm not insulted.  I do appreciate the criticism.  It was just a frustrating distraction since it didn't matter whether I was right or wrong; I just wanted to hear how bitcoinBull's logic went, since I've heard that general theory before and it seemed bizarrely nonsensical.




Oh, I know better than to try to model the market, but I do think my vague hand-waving model is reasonably correct for modeling difficulty - again, using all the factors I mentioned (and a few more, but again, this is all beside the point), not just price.

I dig this in principle, but disagree on the floor - I reckon its probably higher. Though the exact value is not so important. (we aren't trying to time tops and bottoms are we, now that would be naive ;) !)

What's important is that this hypothetical floor is unlikely to be a fixed value. I think it must inevitably rise - as time goes on, more people get involved, and that means more people putting buying pressure at more points on the way down, more people want coins, and there is more money in the game, the harder it falls the higher the number of people see a speculative opportunity

Predicting the ups and downs of a random walk is mug's game. The only question to answer is whether bitcoin will stick. If it will then its inevitable that it will go up in value in the long (long i.e. years/decades) term for simple reason of supply/demand. Im sure the price will oscillate, and some will win and some will use.

I read in another thread someone saying 'buy and hold is dead' (or something like that) if you believe bitcoin will stick then this is absurd. This is possibly the best chance ever to get in at the bottom of what will likely be one of the most profitable secular bull runs in history.

So ask yourself if you think it will stick. There's a few killer reasons it might not - fundamental crypto flaw, fundamental legal issue. Maybe there is something else fundamental that I can't think of right now, but I would cover that under the third possibility 'black swan'.

For these three reasons you should be cautious for sure, assess your appetite for risk and take your position accordingly.

Provided you haven't bet the farm, and are aware that it could all turn to dust, and you are happy with that possibility, then buy and hold is surely the most guaranteed way to 'not lose'. If you don't lose, the winning will take care of itself.

I do a bit of trading here and there - but I know I'm probably on a loser, its entertainment, gambling even. That's entirely different from what I have bought to hold. They will be held - I'm guessing at least a decade.

Each to their own though. As much of a long term bull as I am, I think its fairly clear market is overbought right now (even after the small drop today). My play money is back on the sidelines, and my orders are in on gox (i even have a speculative lowball bid in on bitcoinica with some hefty leverage! cover all bases!).

Let the games continue!

I agree with all of that.  :)

A point that bears repeating is that these formulas aren't for predicting bottoms.  All the variables are too fuzzy to pick an exact number.  Further, the floor is very soft and acts over a long time period, so the price will never bounce off of it.

It's like the P/E on a stock.  There's no "right" value that the price will track tightly, but it's still good for a sanity check: if you see a P/E over 100 and there isn't hard evidence of coming growth, it's overvalued.

And when Bitcoin's at $15 and there's not a major new market in the pipeline....  Well, some speculation is reasonable, but at some point it's overvalued.

The reason for the "Buy and hold isn't a smart trading method", is because everyone KNOWS the price is going back down eventually. Therefore, if you know it's going down, sell and buy 2x, 3x, etc. more BTC at the lower price for holding.

The only thing you should really know for sure, is that you don't know.

What if it doesn't 'go back down eventually' just like you envision. What if it spikes down to a penny above your bids then immediately back up over $10 never to return to single figures?

What if it spikes down, you buy in at $2, the govt makes it illegal and it immediately tanks to a few cents... forever destined to remain an underground currency never worth more then $0.10

What if something happens that nobody has even considered yet that takes it to the moon/zero.

You don't *KNOW* shit. Neither do I.

In the short term, you especially don't know shit. Some geek who was mining back in 2009 may wish to finally buy himself a Porsche... Some 1%er may wish to diversify and throw a few of his millions into this Bitcoin thing.

In the short term anything is possible.

In the long term, we have a deflationary, uncounterfeitable "commodency", so ultimately it can only go up.

Agree with what you've said, part from I would say ultimately should be probably ;)

I love debunking this fallacy.

Remember IxCoin (https://bitcointalk.org/index.php?topic=36218.0)? In the long term, it is a deflationary, uncountrerfeitable "commodency". But in what direction is it going to go?

As long as Bitcoin is still alive, the death of Ixcoin or any other clone only buttresses Bitcoin's standing. It has huge first mover advantage over anything that came after it.

What Ixcoin had, aside from uncounterfeitability and deflation, was also the massive hurdle of having to compete with Bitcoin.

I just debunked your debunking... I love it as much as you :-)

The Dow just went positive after being down all morning.  UST's selling off. 

RISK ON!

+1

I'm sorry how do you know what I need? Perhaps I need a savings account that pays more than %1.3 or maybe I need a place to store my stuff where the local warlord can't get it, you know like one of those expensive off shore accounts? Perhaps I need to side step overly priced wiring fees imposed by government supported monopolies? Maybe I dont want chargebacks anymore? Perhaps I may even need my world economy to stop tanking every-time someone in a position of power decides to fart?

While all the speculators contemplate their navels those using the system are getting away with financial murder, think Scarface where he's bitching about 10% for his transactions, does Tor need Bitcoin?  

I could even be speculating that my local fiat economy is about to fall off a cliff, after all some countries in the world have more debt than they have ever seen in history.

The Internet has needed Bitcoin since before Paypal,

I needed Bitcoin when it wasn't there.