Bitcoin Forum

It finally closed the week green.

It closed green on the 22nd of June! Where were you?  :D

Uh, no.  In fact, that date is mid-week on the bitcoinwisdom chart.  It's not even a closing date for any weekly candle.

In fact it closed green for the first time on June the 12th, but it turned red right after.

Are you talking about the Bitstamp chart? because I see two green weekly candles now.

That's not what 'closed' means, at least not in the common terminology.

EDIT: and you see two green candles because, as Raystonn pointed out, we closed last week's candle in green, and the current week's candle (not closed yet) is green as well.

Finaly, due to all the irrational FUD around the auction it took longer than expected but at last we are green.

Hope this means liftoff.

It means what it means: one more (momentum based) sign that we are leaving the correction-slash-bear market behind.

Immediate 'liftoff' is not guaranteed. You only need to go back to September last year to find 2 green MACD weeks, followed by another drop.

But as I pointed out in another post yesterday, what is special about the current situation is the confluence of momentum signals across time scales: weekly, 3 daily, daily MACD are all positive right now, which rather strongly points to an upwards breakout from the recent slump.

Yes you are right  ;)

"Was green the 12th of June but closed red".

I haven't looked at the MACD for a few weeks now (just watching the price alone). With the recent chain of good news, I can feel the momentum without the charts  :D

Bitcoin does not 'close', it's 24/7 baby.

I think it's different from September. First of September had a green candle too early (they peaked right at the 233 days or so after the precious peak and before the next peak, had they rallied earlier they would be out of sync).

Also, the green candle was pretty sudden as there wasn't really a strong bearish moment yet, while now we had a very intense bear market already before the candle turned green. It already tried to flip to green for 2 weeks or so but kept failing because it kept being pushed back to red. And we are already pretty late if we want to peak 233 days after the previous peak. I don't think we will make it even if we start the rally now.

So I think liftoff will happen soon, but it will still be later than it probably should have been. But that might also mean we reach a higher high. I hope to see $6000+

This. All these non-continuous candles are just arbitrary.

Arbitrary: yes. Uninformative: no.

Maybe open/close doesn't have quite the same importance as in a market that actually closes, but it still seems to carry some information. Also, note that the cutoff time is arbitrary, but consistent.

As an example of relevant information: closing above a line of resistance appears to be a meaningful signal, even in this 24/7 market. Note that, on average, the extrema of a period (high, low) are more thinly traded, and the actual mass of trading happens near H+L/2, to which (except for uncommon candles) open and close are much nearer. In my experience, a candle with a wick/low below some price, that closes above this price, contains the additional information that, within the time period considered, the market moved upwards such that the mass of trading towards the end of the period took place above the resistance.

These bars should just be continuously calculated as if it would close right now. This would give the best insight in the market (and is not possible with the traditional markets).

That would make it impossible to perform any meaningful analysis.

I'd phrase it slightly different than Raystonn: none (or almost none) of the traditional techniques work without segmenting the data in such a way.

That's perhaps the main difference between TA and quantitative forecasting. The former, as a first step, always breaks down the continuous-looking* data into "human digestible" chunks.


* just "continuous looking" because there is a discrete, atomic element after all, a single trade.

If you believe this I guarantee you that your current analysis in meaningless.

What about this guy who says we won't bubble until January (November at the earliest)?

https://bitcointalk.org/index.php?topic=676333.0

interesting thread, but i'm not sure where he got the dates from.

it seems to be way too late, as if he missed one peak/crash.

He discounted 2012 for some reason...

The peaks have been getting further apart but only slightly, the last two were both 235 days interval.

I honestly can't see it taking that long to climb again...I'm betting we are on the cusp of the next big climb right now.
Every other analysis I've seen or performed myself has us at ground zero for the next liftoff.

Some facts:

April 2013's Bubble:
2013-04-09 Peak to 2013-07-05 Bottom: 87 days.

November 2013's Bubble:
2013-07-05 Bottom to 2013-11-30 Top: 148 days.
2013-11-30 Peak to 2014-04-10 Bottom: 131 days.

The November 2013's bubble has much more liquidity than that of April 2013's, hence the longer time period to deflate.

Assuming proportional similarity...

87/131 = 148/X

Hypothetical 2014's Bubble:
2014-04-10 Bottom to 2014-11-18 Peak: 222 days.

This can be driven by the possibility of positive US regulation news in Q4 2014.
$650-$800 is about right for the summer.

i don't think that's an assumption you can just make like that.

especially not based on so little historical data.