{quote} agree and thanks sort of surprised that it moved more than GJ today think I will buy again if I get a good close above 1.0586
- #7,642
- Mar 16, 2015 8:56am Mar 16, 2015 8:56am
- Joined Mar 2010 | Status: aka QueenB | 4,808 Posts
{quote} yeah G/J has a ranging day, which is ok, took one trade and thats it, im observing it atm, 2morrow maybe a better day; I come up with DP now in few minutes, still working on it.
great!
no close above 86 so on the side also.
going for a walk back in a few
- #7,643
- Mar 16, 2015 9:09am Mar 16, 2015 9:09am
- Joined Jun 2010 | Status: zero | 20,603 Posts
{quote} Hey Infinitus, You were talking yesterday about look-back periods yesterday 8h, 32h, 5 days, 32 days, 160 days, 864 days. I have some very basic question regarding this if you don't mind me asking. When you make a profile what time-frames to use, been trying them on 5 min TF for day trading, but as you discussed they are best in ranging markets, what to do when price moves out of range, do you look at higher TF for further clarity. Say for example: I used 5-min TF and made a profile for 8H, 32H, 5Days, 32days, how long will it be valid for....
Sorry for answering so late - was skiing the whole day
The most reliable levels, S/R-levels, whatever, can be obtained from the 4h TF-chart for daytrading. Somewhere in one of my threads there is proof for that.
Have to dig for it later. Since my volume profiles are in synch with my Murrey Math, I would take the data from 32x 4h TF, that is around 5 trading days.
I would make this profile the master-profile.
One TF below would result in a 32h profile, one above in the 32 trading days profile.
I would recommend using all three from a MTF perspective.
My feeling is, that the profiles that are shorter than 24h do not make sense. Maybe it is because not all time zones would be inside of the profile.
I use rolling profiles. New data ticks in, therefore the oldest data gets kicked out. Problem solved with when to create a new profile. In doing this, one can see that it is important to have at least an eye on one TF higher profile.
What to do, when price moves to the far left or far right tail of the profile? I do calculate for each profile a Gaussian distribution. Even when the profile has no good Gaussian shape this method has merits. So we know, when price is in the e.g. 2 SD (standard deviation) range, 3 SD range and so on.
From a mathematical point of view, every time price moves beyond the + or - 3 SD area, chance is overwhelmingly high, that we are in a trending move. So I would at this point not necessarily bet on mean reverting, maybe partial reverting. When the 3 SD gets killed, it is at least time to look into the higher TF profile.
A profile can not get invalidated. It simply is there. Question is, if price is about to explore a new equilibrium (macrostate). This can happen in the very same profile. Sometimes we see profiles with two or more Gaussian like distributions in it.
A profile consists out of high volume zones (there, where price was traded a lot), and the low volume zones (there were traders could hardly agree on price).
Both are S/R-zones
Now to the question regarding the pip difference of each bar:
Let us first go one step back:
It is important, from a statistical point of view, that your sample size for a profile is at least between 600-800. Below will result in poor results, above will not necessarily improve the result.
Keep that in mind.
Now we can decide to collect the data (price and volume) on every tick, or we do simplify the process in collecting the data via 1min candle-data, 5 min candle data etc.
The method with the tick-data would result in the most accurate data. But it comes with a heavy burden:
a) you would need a data vendor, that would provide you with 32 days of tick data (if building the 32 days profile), not to mention the 160 days or 864 days profiles.
b) even when your data provider would have that data, you surly would need to invest a lot in computer infrastructure to handle all this data
So we likely will take the way of simplifying.
For the
8h profile I take 480 1min bars
32h profile I take 1920 1min bars
5.3 days profile I take 1536 5min bars
32 days profile I take 768 1h bars
160 days profile I take 960 4h bars
864 days profile I take 864 daily bars
The result of this simplified profiles will look very identical to the ones gathered via tick data. You will lose some resolution, but the shape will be very same in the end. Remember: you need between at least 600 to 800 samples. Looking on my profiles, you can see, that the 8h profile has only 480 samples. That is, because MT4 does not deliver below 1 min data. Thus the 8h profile has poor quality.
Now to the question regarding pip difference of each bar (let us call it trough size). My spreadsheet takes the high and low of the profile duration.
The calculation area can sample in 300 troughs. Now Excel decides either to take the smallest step or to fill the 300 slots equidistant.
Often I have the following:
8h 1pip
32h 2 pips
5.3 days 2 pips
32 days 2-3 pips
160 days 8-10 pips
864 days 9-11 pips
Last bu no least:
you said: "Also the longer the bar, the more significant the levels becomes for reversing but again that needs to be taken carefully"
I am curious what you may have meant.
The length of a bar in a market profile says, how often the specified price was traded. In volume profile the length of the bar says, how much volume was traded at the specified price.
Hope that helps....
{quote} Sorry for answering so late - was skiing the whole day
insanity: Doing the same thing over... and expecting different results -AE
- #7,650
- Mar 16, 2015 10:46am Mar 16, 2015 10:46am
- Joined May 2011 | Status: [Forex]X-Ray | 4,835 Posts
- #7,651
- Mar 16, 2015 12:08pm Mar 16, 2015 12:08pm
- Joined Mar 2012 | Status: Trader | 10,087 Posts
{quote}do not understand the reasoning for even scalping on an MT4 account. Scalping should be the worst to do on MT4
It works very well and i see no reason here not to trade as on FXCM or Oanda or another broker. This is my view and Tested since more than 2 years.
Another reason is i can draw my fibo's how i want.
- #7,653
- Mar 16, 2015 12:35pm Mar 16, 2015 12:35pm
- Joined Jan 2014 | Status: Less is more ... | 9,059 Posts
Fix up, trade sharp ...
{quote} Sorry for answering so late - was skiing the whole day
... you said: "Also the longer the bar, the more significant the levels becomes for reversing but again that needs to be taken carefully" I am curious what you may have meant. The length of a bar in a market profile says, how often the specified price was traded. In volume profile the length of the bar says, how much volume was traded at the specified price. Hope that helps....
Thank you for the reply, it makes complete sense and gives me a better understanding of it. Will dig it a bit deep. What I meant there was (my understanding, could be wrong), if a specified price was traded a lot then chances of price moving on to the next high level bar increase (negotiating process) - the tug of war between the buyer and sellers!
"Once the market defines a range with excesses at each end, it negotiates within that range to establish value. The market trades between the established excesses until it either trades above the high excess or below the low one."
I think I will go back again and read it once more to have a clear understanding but thanks for the lovely reply.
- #7,656
- Mar 16, 2015 2:12pm Mar 16, 2015 2:12pm
- Joined Jun 2010 | Status: zero | 20,603 Posts
{quote} Thank you for the reply, it makes complete sense and gives me a better understanding of it. Will dig it a bit deep. What I meant there was (my understanding, could be wrong), if a specified price was traded a lot then chances of price moving on to the next high level bar increase (negotiating process) - the tug of war between the buyer and sellers! "Once the market defines a range with excesses at each end, it negotiates within that range to establish value. The market trades between the established excesses until it either trades above...
Exactly!
With this being said, the chance is high in ranging markets, that we can bet on reverting to mean when price is at the tails of the distribution
- #7,658
- Mar 16, 2015 4:51pm Mar 16, 2015 4:51pm
Wave one produced 340 pips Wave 3 is up to 560 pips so far............ Wave 3 is the money wave and normally the longest wave................. {image}
You wrote a couple of days ago about an Elliot wave formation in the H4 chart, with price now at 180 would you say today's small bull leg would qualify as the 4th wave?
Cheers
- #7,659
- Mar 16, 2015 5:34pm Mar 16, 2015 5:34pm
- Joined Nov 2012 | Status: Trader | 13,922 Posts
The More You Learn, The More You Earn.....
- #7,660
- Mar 16, 2015 5:43pm Mar 16, 2015 5:43pm
- Joined Nov 2009 | Status: Trader | 27,245 Posts
{quote} Hi Schickeria, thanks for the video. Really impressive how easy MT4 pricing can be manipulated. Regarding the historical data: do not understand what you want to show with the two sources. FxTop gets the historical data from the central banks. Regarding the three data sources: who are they from? Which broker? When I do compare the volume output from FXCM and the one from Dukascopy, they have different output values. That is dependent of how the volume is calculated and/or put into the volume/tick-volume indicator. Absolute values here are...
...volume oscillators: chaikin money flow, chaikin oscillator, ease of movement (EMV), elder force index (EFI), money flow index (Oscillator), twiggs money flow, volume oscillator, volume rate of change, transaction volume, up and down volume indicators, (modified) arms index, NPDD (ninety percents downside days), 10-to-1 up volume days and 9-to-1 down volume days, value of volume statistics, volume brakout filter, volume dips, volume oscillator, volume spikes, -climaxes, -explained, -on breakout, -shock spiral, volume dips, VPCI, VWA...
FXcube