The Club! | Page 936 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Help Me Establish The GbpUsd After Dinner Club 40 replies

Attachments: The Club!
Exit Attachments

The Club!

  • #18,701
  • Apr 16, 2015 12:00pm Apr 16, 2015 12:00pm
GU ... break above 49 xx tomorrow or has no power?
if so my ST target is 5060 and 5115.
  • #18,702
  • Apr 16, 2015 12:33pm Apr 16, 2015 12:33pm
Quoting infinitus
{quote} Hi FFLeo, thank you for considerations. I have eliminated the gap between the bars. That was easy. I am not able to do the thing with the 5 pip data. In order to let Excel quick and successful calculate the huge amount of data, I have to avoid doing extra calculation. Here are the steps: a) download historical data from Dukascopy b) import them into the calculating spreadsheet c) calculate kind of Pivot out of OHLC-Values d) round them to 4 digits for 1pip width or 3 digits for 10 pip width e) build an Excel Pivot Table with the rounded...
Probably you do not need to calculate the profile for larger pip intervals once you have done it for 1-pip interval. Just bin 1-pip profile into wider pip intervals. For example, for 2-pip interval profile, add the value of volume bars at pip-0 and pip-1 into a single bar and use it as the volume bar at pip-0, and combine pip-2 and pip-3 bars into another single bar and use it as the volume bar at pip-2, and so on.

To bin into a 5-pip bar, just add the 1-pip bars at pip-8, pip-9, pip-10, pip-11 and pip-12 into a single bar and use it as the volume bar at pip-10. It should be pretty fast to bin the data.

bin data points = combine several points into one data point. Write a script to do it automatically
ignore wackos and psychos, including myself. But I can't add myself to iggy
  • #18,703
  • Apr 16, 2015 1:38pm Apr 16, 2015 1:38pm
EU actual volume profile


next layers up will be tougher to be broken
Attached Image (click to enlarge)
Click to Enlarge Name: mm.png Size: 34 KB
  • #18,704
  • Apr 16, 2015 1:40pm Apr 16, 2015 1:40pm
Quoting infinitus
{quote} Hi FFLeo, thank you for considerations. I have eliminated the gap between the bars. That was easy. I am not able to do the thing with the 5 pip data. In order to let Excel quick and successful calculate the huge amount of data, I have to avoid doing extra calculation. Here are the steps: a) download historical data from Dukascopy b) import them into the calculating spreadsheet c) calculate kind of Pivot out of OHLC-Values d) round them to 4 digits for 1pip width or 3 digits for 10 pip width e) build an Excel Pivot Table with the rounded...
since you are doing 3 digits rounding for 10-pip internal profile, you can do it also for 5-pip interval. Just multiply the price value by 2, and then do a 3 digit rounding, and then decide the price value by 2 again to get back to its original price. This way you get 5-pip interval by using a 3 digit rounding
ignore wackos and psychos, including myself. But I can't add myself to iggy
  • #18,705
  • Apr 16, 2015 1:54pm Apr 16, 2015 1:54pm
Quoting ffleo
{quote} since you are doing 3 digits rounding for 10-pip internal profile, you can do it also for 5-pip interval. Just multiply the price value by 2, and then do a 3 digit rounding, and then decide the price value by 2 again to get back to its original price. This way you get 5-pip interval by using a 3 digit rounding

:nerd:


clever!
  • #18,706
  • Apr 16, 2015 2:06pm Apr 16, 2015 2:06pm
Quoting infinitus
{quote} :nerd: clever!
if you multiply price by 5 and then do a 3-digit rounding, and then divide the price by 5 again to get back to the original price, you get a 2-pip internal profile.

1-pip internal is too small, probably 2-pip interval, 5-pip interval and 10-pip interval should be good enough.
ignore wackos and psychos, including myself. But I can't add myself to iggy
  • #18,707
  • Apr 16, 2015 2:12pm Apr 16, 2015 2:12pm
Quoting ffleo
{quote} Probably you do not need to calculate the profile for larger pip intervals once you have done it for 1-pip interval. Just bin 1-pip profile into wider pip intervals. For example, for 2-pip interval profile, add the value of volume bars at pip-0 and pip-1 into a single bar and use it as the volume bar at pip-0, and combine pip-2 and pip-3 bars into another single bar and use it as the volume bar at pip-2, and so on. To bin into a 5-pip bar, just add the 1-pip bars at pip-8, pip-9, pip-10, pip-11 and pip-12 into a single bar and use it as...


Interesting ideas - thanks for all the suggestions


at the moment my thinking centres around the 1 pip bins.

The question is: how much value has one ultra-high volume 1 min candle, when there are several high volume candles in the neighbourhood. Which one will be dominant?

1 pip bins may be good for precision , but e.g. 10 pip bins may be much better for general decision.

What should we calculate to find the optimal bin size?
  • #18,708
  • Apr 16, 2015 2:22pm Apr 16, 2015 2:22pm
Quoting infinitus
{quote} Interesting ideas - thanks for all the suggestions at the moment my thinking centres around the 1 pip bins. The question is: how much value has one ultra-high volume 1 min candle, when there are several high volume candles in the neighbourhood. Which one will be dominant? 1 pip bins may be good for precision , but e.g. 10 pip bins may be much better for general decision. What should we calculate to find the optimal bin size?
The volume should be allocated to each pip every candle covers. If a candle cover 10 pips, the volume of that candle should be allocated to all the 10 pips, i.e., decide the total volume for that candle by the candle width, and assign the result to each pip.

So it is better to get tick data or 1-second data or as small internal data as possible and do binning by yourself
ignore wackos and psychos, including myself. But I can't add myself to iggy
  • #18,709
  • Apr 16, 2015 2:27pm Apr 16, 2015 2:27pm
Quoting infinitus
{quote} Interesting ideas - thanks for all the suggestions at the moment my thinking centres around the 1 pip bins. The question is: how much value has one ultra-high volume 1 min candle, when there are several high volume candles in the neighbourhood. Which one will be dominant? 1 pip bins may be good for precision , but e.g. 10 pip bins may be much better for general decision. What should we calculate to find the optimal bin size?
we rarely need 1-pip precision (pretty much noise, and unless for scalp). 2-pip should be good. 5-pip good for daily and weekly. 10-pip good enough for weekly and monthly.
ignore wackos and psychos, including myself. But I can't add myself to iggy
  • #18,710
  • Apr 16, 2015 3:13pm Apr 16, 2015 3:13pm
http://toyoizumilab.brain.riken.jp/h...histogram.html


Optimal bin size
  • #18,711
  • Apr 16, 2015 3:14pm Apr 16, 2015 3:14pm
Comparison of 1 pip bin size vs. 5 pips
Attached Image (click to enlarge)
Click to Enlarge Name: mm.png Size: 67 KB
  • #18,712
  • Apr 16, 2015 3:27pm Apr 16, 2015 3:27pm
Quoting ffleo
{quote} The volume should be allocated to each pip every candle covers. If a candle cover 10 pips, the volume of that candle should be allocated to all the 10 pips, i.e., decide the total volume for that candle by the candle width, and assign the result to each pip. So it is better to get tick data or 1-second data or as small internal data as possible and do binning by yourself


when the sample size is big enough one should be OK with attributing the volume to (H,L,C,O)/4

1 min data deliver a fine picture. Question is, if we can handle all the data, if we use the tick data. Might be a problem when one has a history of several years.

Also the question is, if we need the precision of tick-data, when we build bins of several pips in the next step.


Even for scalping purpose it might not really so relevant, because although the volume levels are there, price can overshoot, undershoot or hit the centre.

The volume profiles are no timing tools (e.g.: the volume level was hit, I will trade reversion). It has to be combined with other methods.
  • #18,713
  • Apr 16, 2015 3:37pm Apr 16, 2015 3:37pm
Quoting chess king
GU ... break above 49 xx tomorrow or has no power? if so my ST target is 5060 and 5115.
Without 4870 test maybe. But with down direction is what ill play. 4880 first needs to give with LL, then lower. Without 4880 i am not shorting. And even though, bulls are printed in so, even if 4880 is given, i dont think it will go bellow 4740, at leasts thats what structure is telling.
  • #18,714
  • Apr 16, 2015 4:36pm Apr 16, 2015 4:36pm
Quoting janpec1000
{quote} Without 4870 test maybe. But with down direction is what ill play. 4880 first needs to give with LL, then lower. Without 4880 i am not shorting. And even though, bulls are printed in so, even if 4880 is given, i dont think it will go bellow 4740, at leasts thats what structure is telling.
we will see.
tomorrow the news can help the game..
  • #18,715
  • Apr 17, 2015 12:30am Apr 17, 2015 12:30am
Quoting infinitus
Comparison of 1 pip bin size vs. 5 pips {image}
the 5-pip bin size looks smooth enough and still has good resolution.

Regarding sequence (timing) look at the following chart. In a trend (or micro trend), the HH and HL will be located at each peak or ledge of the market profile.

Attached Image (click to enlarge)
Click to Enlarge Name: mm.png Size: 21 KB
ignore wackos and psychos, including myself. But I can't add myself to iggy
  • #18,716
  • Apr 17, 2015 12:48am Apr 17, 2015 12:48am
Quoting ffleo
{quote} the 5-pip bin size looks smooth enough and still has good resolution. Regarding sequence (timing) look at the following chart. In a trend (or micro trend), the HH and HL will be located at each peak or ledge of the market profile. {image}
when price proceeds to a major support or resistance (deep grove or large peak of the market profile), if there is weekly divergence, most likely the price will reverse.

Attached Image (click to enlarge)
Click to Enlarge Name: euro548.png Size: 833 KB
ignore wackos and psychos, including myself. But I can't add myself to iggy
  • #18,717
  • Apr 17, 2015 12:55am Apr 17, 2015 12:55am
Quoting ffleo
{quote} when price proceeds to a major support or resistance (deep grove or large peak of the market profile), if there is weekly divergence, most likely the price will reverse. {image}
another example:
Quoting ffleo
{quote} remember this post. {image}
Attached Image (click to enlarge)
Click to Enlarge Name: Cable-sell - i7860-TDA_273.png Size: 545 KB
ignore wackos and psychos, including myself. But I can't add myself to iggy
  • #18,718
  • Apr 17, 2015 1:34am Apr 17, 2015 1:34am
Quoting ffleo
{quote} the 5-pip bin size looks smooth enough and still has good resolution. Regarding sequence (timing) look at the following chart. In a trend (or micro trend), the HH and HL will be located at each peak or ledge of the market profile. {image}
May I place my 2 cents into discussion?

Yes, extremums are completely at the same prices that have maximal volumes.
It is fundamental thing I learned at forex markets.
The matter is that these ptices are the POC prices at different intervals of time.
The POC-prices (or I must additionally say, PVP-prices) are themselves support/resistance prices - that is why the first hitting of them bring us to the an extremum, the second hitting bring us to either its breakening out or reversing at that price level.
Price is traveling from one POC to the other POC taken at different timeframes.

Quoting ffleo
{quote} when price proceeds to a major support or resistance (deep grove or large peak of the market profile), if there is weekly divergence, most likely the price will reverse. {image}
Agreed on that.
Any gap (either real or hidden) is the place where supply/demand shift takes place. That is why these price areas should be filled with new upcoming prices.

Best,
D.P.
Hunting High and Low
  • #18,719
  • Apr 17, 2015 3:19am Apr 17, 2015 3:19am
Quoting chess king
{quote} we will see. tomorrow the news can help the game..
No 4870 test, and higher it went. Now to open 474 again it needs to break bellow 4910. Without 4910 test upside is open.
  • #18,720
  • Apr 17, 2015 3:29am Apr 17, 2015 3:29am
Very good and important info from ZH:
http://www.zerohedge.com/sites/defau...150416_BoJ.jpg

Eyes on JPY pairs, or rather BOJ in next days. Finger ready on buy if they announce QE of some sort.
6 traders viewing now