- This topic has 227 replies, 14 voices, and was last updated 10 years, 3 months ago by
.
- Posts
Ok, I respect your opinion. And surely, dojis often have high volume. But this is due to the movement from open to high, from high to low, and to close … So, how would you interpret the high correlation of data shown in scatterplot at the bottom? And yes, tick chart analysis will give better results but MT4 tick volume can give a fair estimate and its less expansive (cpu and time consuming).
Nevertheless, what’s your opinion on increased volume near reversal points?
If you look at my ForexGT_VolumeOsc (the Volume Oscillator) you’ll see that 80% of the market (between the +40 and -40 lines) lies most of the market. When the Oscillator is above that – it’s almost shure (and accurate) market swing point, usually with minimal price movement. G.
Ok, this is an answer to my question I just asked. Thanks.
Could you give me an advice how to obtain tick counts?
If you look at my ForexGT_VolumeOsc (the Volume Oscillator) you’ll see that 80% of the market (between the +40 and -40 lines) lies most of the market. When the Oscillator is above that – it’s almost shure (and accurate) market swing point, usually with minimal price movement. G.
Ok, this is an answer to my question I just asked. Thanks. Could you give me an advice how to obtain tick counts?
“Volume”, in Forex, is basically “ticks” – i.e. number of price changes reported to the broker by the Interbank.
Based on that you can count the “Ticks” per second, per minute, per bar, or whatever – based on the incoming “Volume” parameter in the Data-Window or via MT4 readings of that parameter.
To be accurate you’ll need fast PC and fast internet connection. Otherwise, you’ll get your data exchange rate between your PC and your broker’s server…
You can identify this by seeing gaps in the Tick-Chart.
G.
Ok, I respect your opinion. And surely, dojis often have high volume. But this is due to the movement from open to high, from high to low, and to close … So, how would you interpret the high correlation of data shown in scatterplot at the bottom? And yes, tick chart analysis will give better results but MT4 tick volume can give a fair estimate and its less expansive (cpu and time consuming). Nevertheless, what’s your opinion on increased volume near reversal points?
The explanation is quite simple.
Volume IS correlated to price movement, but since it’s a LEADING indicator – it will mostly manifest the correlation BEFORE the price movement.
That’s also explains the “Doji” effect, i.e. High Volume and minimal movement – the price movement WILL come in the next bar or tick.
Your High/Low movement explanation of volume accumulation doesn’t clearly explains the high volume – because similiar “regular” bars, with roughly the same High/Low, doesn’t accumulate the same high volume.
That’s why I’m so heavilly “invested” in Volume indicators, and their further development.
G.
This is just what I was considering an hour ago or so for higher timeframes:Based on that you can count the “Ticks” per second, per minute, per bar, or whatever – based on the incoming “Volume” parameter in the Data-Window or via MT4 readings of that parameter.
Dividing all volume tick counts by Period() will easily show ticks per minute. It’s so obvious as a first linear step towards normalization that I’m amazed I didn’t see it earlier.
Now for M1 and ‘below’ I’ll post a low end prototype for extrapolating volume at bar zero (low cpu load) as a first step later this evening.
s.
A good trader is a realist who wants to grab a chunk from the body of a trend, leaving top- and bottom-fishing to people on an ego trip. (Dr. Alexander Elder)
Sry, can’t see why volume should be leading. You can’t say that volume is leading prize or vice versa. In my opinion both are two sides of one coin. Prize can only change if there’s a disequilibrium of interest in buying or selling. Tick volume can be interpret as measure of this disequilibrium over a particular time (e.g. 1 min). Volume will only increase if prize changes. And prize changes only if such an disequilibrium exists and thus if volume increases. Maybe volume has an effect on future prize changes, but I haven’t surveyed this and also haven’t seen some studies on this topic. Thus, I wouldn’t say that volume is leading and give hints on future prize moves …
This is just what I was considering an hour ago or so for higher timeframes:Based on that you can count the “Ticks” per second, per minute, per bar, or whatever – based on the incoming “Volume” parameter in the Data-Window or via MT4 readings of that parameter.
Dividing all volume tick counts by Period() will easily show ticks per minute. It’s so obvious as a first linear step towards normalization that I’m amazed I didn’t see it earlier. Now for M1 and ‘below’ I’ll post a low end prototype for extrapolating volume at bar zero (low cpu load) as a first step later this evening. s.
Incoming “Ticks” are not depended on price chart TF. You should get the same Tick-Rate/sec in M1 and H4…
G.
Sry, can’t see why volume should be leading. You can’t say that volume is leading prize or vice versa. In my opinion both are two sides of one coin. Prize can only change if there’s a disequilibrium of interest in buying or selling. Tick volume can be interpret as measure of this disequilibrium over a particular time (e.g. 1 min). Volume will only increase if prize changes. And prize changes only if such an disequilibrium exists and thus if volume increases. Maybe volume has an effect on future prize changes, but I haven’t surveyed this and also haven’t seen some studies on this topic. Thus, I wouldn’t say that volume is leading and give hints on future prize moves …
Yes, I can say that Volume IS a LEADING indicator… there are two arguments for that:
1. Freedom of Speech…. (see US ammendments) :-}
2. Analyzing correlation and timing on Tick-chart. I did it on milions of data points, verifying this over and over, again and again.
Furthermore, I can even show you the “reverse” effect, i.e. when price moves WITHOUT support of Volume, it will soon enough fall… that’s “fools money” chasing “trend” that’s already finished, and it can easily identified by a good Volume indicators.
A childish example of Volume effect:
If you and I are the only persons in the market, you got EUR and I got USD, we will soon enough establish an exchange rate which will not vary much as time goes by.
Now here comes another fellow, pockets full of either EUR or USD. Even before the exchange among us start – the exchange rate will start moving significantly from its previous rate.
The reason: new VOLUME comes into the market.
G.
A rather interesting topic. So if you say that prize can move without support of volume, I think in my mind’s world it’s the equivalent of maximum rise in prize with given volume. For instance, if volume for a particular time interval equals 10 it means that prize may have changed at most -/+1 pip (= 10 ticks).
Yes …
Incoming “Ticks” are not depended on price chart TF. You should get the same Tick-Rate/sec in M1 and H4…
I’m aware of that fact – thanks!
s.
A good trader is a realist who wants to grab a chunk from the body of a trend, leaving top- and bottom-fishing to people on an ego trip. (Dr. Alexander Elder)
A rather interesting topic. So if you say that prize can move without support of volume, I think in my mind’s world it’s the equivalent of maximum rise in prize with given volume. For instance, if volume for a particular time interval equals 10 it means that prize may have changed at most -/+1 pip (= 10 ticks).
Increased volume (either up or down) will move price from its current rate.
That’s the equivalent of supply & demand, basic laws of any “free” market. The declaration itself that “I want to buy” will create the demand.
On the other hand and in extreme cases, TOTAL lack of Volume will create chaos in the FX market. See the EURCHF in January (Swiss bank drops the promissed peg @1.2000) where everyone wants to sell but nobody wants to buy. Price fell like a rock until exchange rate met someone willing to buy.
G.
To prove to yourself the above just watch major announcements, minutes or seconds before the events.
You’ll see price “spikes” without much movement in volume. Those are the “fools money” trying to guess the direction and get quick pips.
If it was some major financial institution behind it – you will see the volume move in that direction.
G.
Maybe, i´m just dumb…but i got a little question regarding who generates the volume we see on our brokers platform. If it´s only the customers of the broker..and those orders doesn´t hit the market..then it can´t influence price..right? And even if all retail traders trades would be in the market..all together on the same trade at the same time..would that be enough to move the price for lets say 1 pip? I don´t know how much is needed for a 1 pip movement.
Isn´t it true that almost no retail trader money reaches the real market? We´re almost always just dealing with our broker..alone. If so , where does this “fools money” come from?
i´m a weirdo..
"A dream you dream alone is only a dream. A dream you dream together is a reality." (John Lennon)
Hi all,
about ForexGT_OscVol, we need use ticks and try to put the highest percentaje possible in a range?
So we have to normalize tick in a period = session?
I try to study the volume and their correlations. It is interesting.
Thank you very much G . Open the eyes i bit more!
Ok, here we go.
Please remember that this is a preliminary and purely experimental indicator – to be continued and extended (hopefully!).
All we have now is plain volume (in ticks per minute) and an average, in particular my version of an FIR filter based on John F. Ehlers’ publications. At the moment, the details of this filter are not important, it only serves as ‘some’ visual anchor.
G.: we discussed volume averaging before – your proposals are not forgotten!
What I want to explore now is that dancing red dot at bar zero while ticks are coming in: this is an ultra-simple linear extrapolation of the accumulated tick count of the most recent bar. Linear extrapolation starts 5 seconds after open time, and will be continued until close. A very simple model, that leads to mostly stable predictions after about 20 to 30 seconds (visually). Many volume outbreaks at bar zero can be monitored before the bar is closed by visually following this simple dot. One step in the right direction, IMO.
Obvious problem: when a volume outbreak starts shortly before close, we are ‘blind’ for the next 5 seconds, which may be essential for a trade decision.
Now let’s consider we’re putting fixed sample rates M1, M5, etc. aside. In our code, we’re setting up a compound array that serves to store tick counts per second – ‘every’ second, not every minute. So we would have a ‘rolling’ window in the time domain, at 1 sample per second. My proposal for its depth would be 65 elements, so we could store volume values of 64 past seconds and the current one.
From those 64 past values we would extract the average tick count per time unit for the most recent intervals of 4, 8, 16, 32, and 64 seconds. These will be visualized. If volume is constant, those 5 values should be close together, having no specific order. When volume suddenly increases, tick counts of the shorter intervals would be on top, and below the longer ones for decreasing volume.
We might even think about low pass filtering those 1 second samples in order to display an ‘intrabar volume trend’.
Remarks?
s.
A good trader is a realist who wants to grab a chunk from the body of a trend, leaving top- and bottom-fishing to people on an ego trip. (Dr. Alexander Elder)
Maybe, i´m just dumb…but i got a little question regarding who generates the volume we see on our brokers platform. If it´s only the customers of the broker..and those orders doesn´t hit the market..then it can´t influence price..right? And even if all retail traders trades would be in the market..all together on the same trade at the same time..would that be enough to move the price for lets say 1 pip? I don´t know how much is needed for a 1 pip movement. Isn´t it true that almost no retail trader money reaches the real market? We´re almost always just dealing with our broker..alone. If so , where does this “fools money” come from? i´m a weirdo..
Not all brokers have “dealling desk”. Some are ECN/STP. Even those with “Dealling Desk” are backing themselves in the market.
In any case, retail market is ~6-7% of the total FX market.
The Volume or “Ticks” data is coming from the Interbank and shown by the broker in your Mt4 terminal after adding its spread.
G.
Ok, here we go. Please remember that this is a preliminary and purely experimental indicator – to be continued and extended (hopefully!). All we have now is plain volume (in ticks per minute) and an average, in particular my version of an FIR filter based on John F. Ehlers’ publications. At the moment, the details of this filter are not important, it only serves as ‘some’ visual anchor. G.: we discussed volume averaging before – your proposals are not forgotten! What I want to explore now is that dancing red dot at bar zero while ticks are coming in: this is an ultra-simple linear extrapolation of the accumulated tick count of the most recent bar. Linear extrapolation starts 5 seconds after open time, and will be continued until close. A very simple model, that leads to mostly stable predictions after about 20 to 30 seconds (visually). Many volume outbreaks at bar zero can be monitored before the bar is closed by visually following this simple dot. One step in the right direction, IMO. Obvious problem: when a volume outbreak starts shortly before close, we are ‘blind’ for the next 5 seconds, which may be essential for a trade decision. Now let’s consider we’re putting fixed sample rates M1, M5, etc. aside. In our code, we’re setting up a compound array that serves to store tick counts per second – ‘every’ second, not every minute. So we would have a ‘rolling’ window in the time domain, at 1 sample per second. My proposal for its depth would be 65 elements, so we could store volume values of 64 past seconds and the current one. From those 64 past values we would extract the average tick count per time unit for the most recent intervals of 4, 8, 16, 32, and 64 seconds. These will be visualized. If volume is constant, those 5 values should be close together, having no specific order. When volume suddenly increases, tick counts of the shorter intervals would be on top, and below the longer ones for decreasing volume. We might even think about low pass filtering those 1 second samples in order to display an ‘intrabar volume trend’. Remarks? s.
1. For prediction of next move you’ll need the Volume DIRECTION.
2. Only last 1-2 tick-rate time-window/interval and their direction will determine next bar initial move.
G.
Just adding one idea. From results of my experiment I conclude that it should be possile to estimate up and down volume by calculating proportions of up/down ticks in relation to total tick volume. Maybe we could apply it on @simplex indicator to obtain possible volume direction.
Ok:
For prediction of next move you’ll need the Volume DIRECTION.
I assume you’re talking about direction of price?
Only last 1-2 tick-rate time-window/interval and their direction will determine next bar initial move.
That would make the coding task simpler and reduce CPU ressource consumption. I already assumed that the shortest (i.e. most recent) time intervals would make the greatest impact.
s.
A good trader is a realist who wants to grab a chunk from the body of a trend, leaving top- and bottom-fishing to people on an ego trip. (Dr. Alexander Elder)
Just adding one idea. From results of my experiment I conclude that it should be possile to estimate up and down volume by calculating proportions of up/down ticks in relation to total tick volume. Maybe we could apply it on @simplex indicator to obtain possible volume direction.
As already explained: a “Tick”, in the FX “Volume” parameter, is one price change.
If the “Tick” caused price to go up, what is the volume direction of that tick?
G.
It’s up. But that’s exactly what I did. I tried to separate volume to up and down volume. So I can’t see what you’re mentioning.
It’s up. But that’s exactly what I did. I tried to separate volume to up and down volume. So I can’t see what you’re mentioning.
Sorry, it was a reply to simplex last post. I quoted the wrong post.
G.
G:
If the “Tick” caused price to go up, what is the volume direction of that tick?
Ok, got it – thanks!
Only counting, no weighing with strength of price movement, I assume.
s.
A good trader is a realist who wants to grab a chunk from the body of a trend, leaving top- and bottom-fishing to people on an ego trip. (Dr. Alexander Elder)
Hi all,
I never worked with tick so looks like interesting and i want to learn all possible.
I have some doubts:
1- For predicted direction is necessary 1-2 previous ticks?
2- Volume of tick is the direction of this tick: ie, if we have two up ticks, then we have 2 u of volume??
3- We have to counted this ticks and his volume and enter in one time period: ie, if we use M1 chart, we have to count the 60 previous ticks??
4- But the question is, how to have all the market between +40 and -40 because if we count ticks by this way, we still have market like pips.
Thank you very much for this interesting subject
This is just what I was considering an hour ago or so for higher timeframes:
This is just what I was considering an hour ago or so for higher timeframes:
- You must be logged in to reply to this topic.