Topic: A MACD Analysis Thread (Read 3178 times)

seams we need to address this point macd wont tell you where the freaking top or bottom is if that shit even existed we wont have a market no forex no stocks nothing caput dead muerto mort. Macd just tell you how is the present price compared maximum of past 26 period (default settings )

it ain't no bottom yet

hi troll

Uh oh! macd says get out at the bottom.

1h macd is about to have a bearish cross so anyone who joined the market from the last bullish cross will have to exit
note that hull macd is very sensible
http://yfrog.com/hsouuvp



remember if any of this is usefully for you donate

macd say get out

http://yfrog.com/hswkicp

1d macd is not doing anything
http://yfrog.com/ockuboep



1d hull macd is about to cross but remember hull MA is extremely sensible
http://yfrog.com/nx3kqpp

By this logic, ALL indicators are flawed, since all indicators/oscillators are a retold story of the past. The MACD is nothing more than a pair of exponential moving averages and the convergence/divergence thereof. Similar to the (10, 20, 50, 100, 200) EMA crosses used as an entry level, textbook indicator for determining shifting market trends.
Nothing is perfect in TA. You can only make the most informed decision based on what has already happened.

I hate to break it to you, but the MACD has a fatal flaw. In essence, the MACD is really a lagging indicator. But people seem to think that by shifting the values into future by inconsistent amounts, they can turn it into a leading indicator. That is stupid, of course, but that's what people do. By inconsistent, I mean that the two moving averages used to calculated the MACD are offset by different amounts (of time), and that generates lot of noise which most people interpret (unfortunately) as signals.

Ever wonder why moving averages never seem to be aligned with the data they are averaging? That is because usually when they are drawn, they are shifted into the future by N/2 periods. For example, a 26-day MA is shifted into the future by 13 days and a 12-day MA is shifted by 6 days. This is why the yellow and purple lines on your MACD chart look similar but are not aligned.

If you combine a MA shifted 13 days with a MA shifted 6 days, you get noise. If you remove the shifts, then two things will happen.

1. You will get real signals, and no noise.
2. The signals will all happen after-the-fact.

The MACD is really a lagging indicator, which means that it can't be used to predict anything.

MACD dont give us much clue only the fast MACD histogram points out lower it can also be a false signal because it is over sensible 
http://yfrog.com/mgoaqcp

look at the dated on the image, the yellow line is the current level of macd 0.268 but the 3 lines are there to show when was last above this level

I am looking at a 1 h macd that looks quite different   (bitcoincharts.com)

chart 1 h period  the yellow line show the current macd level
http://yfrog.com/khu18yp

bullish cross above ZL and ppl who did the 5 Histo bar front running you risked and got a better entry
what i said on the above post get confirmed here 
http://yfrog.com/h6yfuep

the ZL can provide a bounce and a bullish cross but is way to early to tell there is room for side ways for another week and then the bounce of cross can happen
http://yfrog.com/od8u1grp

soon there should be a move based on the macd
http://yfrog.com/oea2tkp

so far is a small divergence on 1d chart
http://yfrog.com/kjru2up


macd on 1h chart did X above ZL
http://yfrog.com/esdtkep

yes
yes
because MACD is a oscillator

It seems that in the "layered" MACD histogram, the darker ones (slower?) are lagging compared to the lighter ones (faster?), but they tend to line up very tightly at local tops/bottoms. This may be good for helping in the decision making for entry/exit. Obviously reinforced by other technicals. I do sense the possibility of over trading on shorter bar periods (looks like you used hourly?) If used on a daily chart, it may minimize some of the noise and false triggers, at the expense of a smaller profit.

first one need to remember what is the meaning for histo bar
where are the histo bars above or below
is there any divergence ?

macd is a oscillator ( https://bitcointalksearch.org/topic/oscillator-vs-indicator-know-them-before-you-speculate-131184 ) so in your scenario i will consider a change IF the bars and macd lines are above the ZL and after 3 histo bars heading lower 5th bar will reinforce this and there should be more analysis in lower and larger time frame using macd, then checking the trend, momentum indicators, BB, EW, MA and FIB levels

What do people think about using the MACD histogram to front run MACD crosses?  For example when the histogram bars stop rising upwards and the most recent bar is lower than the last bar (in an uptrend).  Any good indicator to confirm the "rate reversal" you would see in the MACD histogram?

i will make threads about them soon™

Good thread to remind others basics of TA.

IMO, witchcraft around only one oscillator (macd) doesnt make sense. MACD ZL cross/X must be confirmed along with others TI/trends/MAs/EWs/etc.

Only complex look on current situation gives best prediction.

yep that's a quite achievement since my English is bad


Yet another hammer
http://yfrog.com/nxmikfp




if anyone wants the 3 macd chart posted above http://www.putlocker.com/file/3E5B0C1F82E17B64

aye mang at least you spelled grammar right 

before using macd one needs to understand why the need for macd ? because trading using moving average sux sometimes
so the main macd line ( the fast one ) is a the difference between 2 MA 12 periods and 26 periods  by default on SC since this are the post used numbers
so we got a line that is the diff of 2 MA and since that's not twisted enough there is another MA based on this line and you get "the slow" macd line by default is 9 periods
also there is the histogram that show the gap between the 2 lines and traders who like to take risk use them to do front running since the macd can lag (so many damn MA) so most of the time is considered that 3 histogram lines show a trend and above 5 give a confirmation for said trend
also there is the ZL that give major support and major resistance nothing more

Convergence cases
is when the price go up for  and macd go up
is when price go down and macd go down
divergence cases
is when price go up and macd go down
is when price go down and macd go up


weekly chart
there is only one problem the X(cross) was above the ZL that means any rally will have short live, but like always is a probability and there can be exceptions
http://twitter.yfrog.com/h3xydzp


a perfect example for a exception was the gold ( use google and check for fundamentals)
http://twitter.yfrog.com/g0ekmp

1 day
the price goes up, it take off a trend line, histogram goes under ZL, macd X above the ZL and this is a bearish cross since the fast macd goes blow the slow macd
http://twitter.yfrog.com/esd6hvp



1h
chart is a perfect example of ZL hammer when macd was about to go bullish above the ZL BAAAAAAMM!!!! hammer  
http://twitter.yfrog.com/h62gyep



how to avoid data noise ?
SC by default use last price (the close price) so using a average of OHLC can take off some data noise but also can delay a signal :/
other way to cut the data noise made by close ?
is using volume, here is a chart suing 10k volume per bar
http://twitter.yfrog.com/obrbcqp



maybe i will fix typos and grammar later