Which is a pretty good indicator which part of the market cycle we're still in. But I guess that was your point as well.

There's way too much unwarranted hate directed towards K128kevin and accusations of FUD. To which I'd like to say: Shut the fuck up, guys. Be grateful that at least /some/ people in here post minimally more intellectually challenging stuff than pictures of bears or rockets going to the moon.

That said, I'm very skeptical a simple ANN trained on price diff is of much use for actual trading decisions. It'll probably work okay for days with little volatility, but it's almost certainly too crude to in any reliable way predict major trend changes. Sorry OP.

Works.

You're fast, man. If I ever need a developer, I'll try to hire you :D

According to the site: "+" = 171, "-" = 173.

Yeah, pretty much what windjc said: "US based exchange", "all of us are using" ... doesn't compute, for me at least.

Sorry to jump in like that. Just found the thread by chance. Especially big thanks to johnsl for a super insightful answer.

Question though: The link give by googlemaster, to
https://bitmaintech.com/productDetail.htm?pid=00020140107162747992Ce5uBuxW06D6

... I have to say, I've never been tempted to start mining because I always figured that, since BTC profitability is my goal, all mining rig proposals I've considered would have been BTC-negative ROI, and only (marginally) USD profitable if price increased, so buying/speculating with BTC directly was always the way to go for me.

But the Bitmaintech rigs above... they look pretty different. For, say, 100 btc I could get  upwards of 18 TH/s ... o_O

Are they realiable? Does anyone know if they actually deliver, in time?

The calculator at http://www.vnbitcoin.org/bitcoincalculator.php gives me a ROI of about 60 days, based on 15% diff increase after 2016 blocks... too good to be true?

EDIT: is there a BTC calculator that returns the BTC TOTAL generated by a miner (i.e. doesn't calculate profitability based on USD costs/income, which is irrelevant IMO, since like I said, the goal is BTC profitability)... if I approximate this right, with the bitmaintech miners, investing 100 BTC should return ~160 BTC over their life time (assuming 15% diff increase per period). I'd be very happy to take that .

hm. "-" works. "+" doesn't. strange.

Great analysis. Thanks, arepo. Very thorough, and convincing, as always.

I guess in the end we have a slightly different approach to trading: I'm guilty of being a bit of a "trend chaser", and you seem to be more a "predictive trader". And if I look at the larger charts picture, I can mainly see trends that slope down, be it 2 weeks, 2 months or 3 months. And all of those do so with pretty high accuracy. The only trend that falls out of this picture is a ~1 month trend. Depending on how you look at it, it is however also the least "established" one in, my view. So I tend to be skeptical that the more established downwards trends are really over, and the less established upwards trend is really the dominant one.

Based on my long-term assumptions about BTC price, I'm most of the time all-in in BTC with my trading position (except for periods that are absolutely certain to be a longer running down trend), but I freed a bit of fiat a day ago because I suspect the downtrend will continue, though I'm not sure of this. I will say that much however: if in the coming 3 days price doesn't go back into the territory where it falls back in line with the so far "frail" uptrend, by which I mean: break through 600 and stay there, I'm almost certainly going to free more fiat.

Very cool! Thank you.

Just checked it though, doesn't seem to work in Firefox 27.0.1. Checked in Safari, where it works. Any idea what's the reason?

Any news on the "zoom in/zoom out" button? Wouldn't the easiest (and most elegant) solution be to bind it to a keyboard key (like "+" and "-"), or is that something that requires a different style of webpage (something AJAX heavy, I don't know)?

Thanks for the update!

Perhaps I'm leaning a bit more towards a bearish scenario than you at the moment, with ~550 (EMA trend), or even 518 (fibo level) as possible targets for the coming days, mainly because I see clear signs we're gathering momentum to the downside (e.g. the daily MACD you mentioned, or looking at a Fisher transform indicator), and the divergences you see on the oscillators might not have the strength to counter that momentum.

So this is maybe one suggestion I have: your analysis is rigorous, and plausible, and it might well come true (and it'll turn out my more bearish view is wrong). But I also think to be profitable as a trader in the long run it is more important (and more likely) to be *prepared* than to be *right*... what I mean is that several options/scenarios should always be available to a trader, and he/she should pick the most likely one based on what he sees as time progresses.

So maybe you could look into the possibility of a continued (and gathering momentum) downward move, and what signs would point to that one playing out. To be clear: I'm not saying you should change your view, just that it'd be interesting to see your rigor applied to a (maybe less likely, in your eyes) scenario that diverges from the view you presented above.

Now that's a good example of a contrarian indicator post. As long as guys like you keep posting "we only going down because MANIPULATION", there's still plenty of room to the downside left. Btw, same holds true for bearish conspiracies in an uptrend ("The China rally isn't sustainable!!!", @~700)

Sold ~1/5 of my on-exchange BTC @590-595

Here's what I wrote before about that, in the wall thread:

This will take some time before it will influence price. It's not the bullish equivalent to a flash crash event... what I mean is, I'm not expecting price to turn around 180 degree tomorrow. But this combination of solid 'old' money chasing the new one, not just the size of the Pantera/Fortress fund at registration, but also that little nugget about Bitstamp, makes me pretty confident that those scenarios we discuss here once in a while, like yesterday, those that paint the picture of a painfully drawn-out bear market that can last several more months to a year (you know the people who see it like that, Mat, or Lucif) simply are not a realistic option anymore.

* * *

Also, confirmed or not is not the point. A lot of "unconfirmed" bearish news influence price, and the same holds for bullish items. The difference is that this one will take time to sicker in. It'll probably work in the mid term to give back some confidence that at least some exchanges are (mostly) trustworthy, which in turn will (mid term) increase buying pressure. But more swings to the downside are completely possible until then, in my opinion.

tbh, short-to-mid term, the situation might remain somewhat depressing. I'm not sure if we can hold 600. But every pessimist mood came to an end so far, and I'm betting on that happening earlier (say, in the range of at most a couple of months of lackluster price development) rather than later.

Question (since you seem know your stuff)... how does LTC handle volume? I didn't trade any alts yet, always thinking 'probably too much slippage to be worth my while', but from what I can tell you don't trade in tiny volume either. Is it worth it? (I'm aware, asking another trader if he thinks I should trespass into his territory :D ... just think of me as another pig that can be slaughtered)

http://www.bloomberg.com/news/2014-03-17/fortress-linked-pantera-said-to-invest-in-top-bitcoin-exchange.html

http://finance.fortune.cnn.com/2014/03/18/bitcoin-hedge-fund/

You know, this will take some time before it will influence price. It's not the bullish equivalent to a flash crash event... what I mean is, I'm not expecting price to turn around 180 degree tomorrow.

But this combination of solid 'old' money chasing the new one, not just the size of the Pantera/Fortress fund at registration, but also that little nugget about Bitstamp, makes me pretty confident that those scenarios we discuss here once in a while, like yesterday, those that paint the picture of a painfully drawn-out bear market that can last several more months to a year (you know the people who see it like that, Mat, or Lucif) simply are not a realistic option anymore.

Perfectly valid observation. It's similar to the longest bear market scenario I can reasonably conjure myself, and probably also matches the EW theory motivated scenario that keeps floating around, in which the bear market lasts into 2015.

That said, I'll start worrying about it once it's clear that the following two, already rather pessimistic, triangles don't play out ultimately, i.e. a clear reversal towards the end of one of the 2 (and a half) triangles. I still consider those  more likely than the scenario above.

A note on volume

Seems to be a common complaint right now: "volume is just too low", "back in 2013 the reversal had much better volume". I'm not so sure about that. Let's see...

Excluding the day of the bottom itself (I'm leaving it out because I am interested in the "post bottom" volume), we have 20 days to look at. So I'm looking at 26.02.-17.03 for the current period. Same (post bottom) period in 2013 is 06.07.-25.07. Here's a visual of the time period I'm talking about:






Now, volume data (via btccharts). I'm summing over Mtgox, Bitstamp and btc-e in 2013, and Bitstamp and btc-e in 2014. Note that I'm ignoring Chinese exchanges entirely, which contribute a lot more volume today than back then, which works against my own case. (but tbh, I simply don't trust their volume can be taken at face value).

06.07.-25.07.2013
Mtgox: 57.3M USD
Stamp: 21.3M USD
btce: 8.4M
= 87M USD

26.02.-17.03.2014
203.5M USD
114.2M USD
= 318M USD

That's volume in USD, alright. Has to be, as that's the only meaningful way to compare volume across eras with vastly difference prices.

Volume in dollar:
318M:87M = 3.7:1

So, $volume is 3.7 times higher now than it was in the same "reversal" period in 2013.

Now, to mention this as well, price now is ~640 USD, back then it was ~90:
640:90 = 7.1:1

Summary: we trade at a price around 7 times higher per coin right now than back then, and are seeing about 3.7 times as much volume in USD compared to 2013. Keep in mind that expecting the relation between volume and price to be 1:1 a constant is most likely wrong, so based on volume alone I don't see a reason to question the trend reversal.

To be fair, he was one of the relatively few dissenting voices during the ~800 "stability" phase that warned (and I'm using the word "warn" loosely here -- more like "shouted without interrupt") that we'll go down further. Probably got quite a few ignores for that, undeservedly so.