Calculating Acc/Dist | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

how many lots on a 35k acc 6 replies

WADS (William Acc/Dist) Make Money off of the Trend and The Reversal! 0 replies

Opening an Acc with Alpari 6 replies

mt-4 demo acc 6 replies

Calculating Acc/Dist

  • #1
  • First Post: Jun 30, 2005 3:03pm Jun 30, 2005 3:03pm
Hello Everybody,

In my recent studies, I got stuck into a problem which I couldn't solve. I need to measure the Accumulation and Distribution of a certain currency pair.

I am trying to get out an idea about trend continuations and reversals, that's why I am looking for the accumulation and distribution measure.

Now the real problem is the abscence of volume in the forex. All methods of accumulation/distribution calculations requires volume.

Does anyone know of a method for calculating the accumulation and distribution using only the 4 data prices available (Open, Low, High & Close) ??


Thanks All.

Nader.
  • #2
  • Jul 1, 2005 7:22pm Jul 1, 2005 7:22pm
nader, can you post the formula you are using? i will see if we can get it to work without volume.
Relax and be happy.
  • #3
  • Jul 2, 2005 2:31am Jul 2, 2005 2:31am
MetaTrader has an Accumulation/Distribution indicator. Here is the MetaTrader code:
//+------------------------------------------------------------------+
//| Accumulation.mq4 |
//| Copyright 2005, MetaQuotes Software Corp. |
//| http://www.metaquotes.net/ |
//+------------------------------------------------------------------+
#property copyright "Copyright 2005, MetaQuotes Software Corp."
#property link "http://www.metaquotes.net/"

#property indicator_separate_window
#property indicator_buffers 1
#property indicator_color1 LightSeaGreen
//---- buffers
double ExtMapBuffer1[];
//+------------------------------------------------------------------+
//| Custom indicator initialization function |
//+------------------------------------------------------------------+
int init()
{
IndicatorShortName("A/D");
//---- indicators
SetIndexStyle(0,DRAW_LINE);
SetIndexBuffer(0,ExtMapBuffer1);
//----
return(0);
}
//+------------------------------------------------------------------+
//| Accumulation/Distribution |
//+------------------------------------------------------------------+
int start()
{
int i,counted_bars=IndicatorCounted();
//----
i=Bars-counted_bars-1;
while(i>=0)
{
double high =High[i];
double low =Low[i];
double open =Open[i];
double close=Close[i];
ExtMapBuffer1[i]=(close-low)-(high-close);
if(ExtMapBuffer1[i]!=0)
{
double diff=high-low;
if(0==diff)
ExtMapBuffer1[i]=0;
else
{
ExtMapBuffer1[i]/=diff;
ExtMapBuffer1[i]*=Volume[i];
}
}
if(i<Bars-1) ExtMapBuffer1[i]+=ExtMapBuffer1[i+1];
i--;
}
//----
return(0);
}
//+------------------------------------------------------------------+
sxaxlxvxaxtxixoxnxbxyxgxrxaxcxexdxoxtxoxrxgx
  • #4
  • Jul 2, 2005 2:42am Jul 2, 2005 2:42am
Quoting narafa
Hello Everybody,

In my recent studies, I got stuck into a problem which I couldn't solve. I need to measure the Accumulation and Distribution of a certain currency pair.

I am trying to get out an idea about trend continuations and reversals, that's why I am looking for the accumulation and distribution measure.

Now the real problem is the abscence of volume in the forex. All methods of accumulation/distribution calculations requires volume.

Does anyone know of a method for calculating the accumulation and distribution using only the 4 data prices available (Open, Low, High & Close) ??


Thanks All.

Nader.
A&D cannot be measured by OHLC except in terms of movement. Low open, long range, high close is clearly accumulation if the time frame over which it happens is statistically significant. But when it happens during low volume is it real accumulation or just speculative opportunism? Additionally one of the most significant A&D indicators of continuation (rather lack thereof) is when there is high volume with little price movement. Completely undetected by OHLC and yet one of the most statistically relevant reversal indicators.

I've often thought that forex trading activity might be measured by designing a tick meter. Measure ticks per second or ticks per minute and combine that with some type of ROC measurement and you would be able to synthesize market acceleration and deceleration. The more data sources you monitored the more statistically significant your readings would be.

Say you had a Tick-ometer for each of 10 data sources measuring ticks/timeunit for each as well as the speed of ticks incoming and the rate of price acceleration/deceleration -- you could measure market behavioral patterns in ways that are not currently available in any conventional software that I know of. I think this would lead to some very interesting clues to market behavior particularly if you had data sources from the major exchanges and from several forex sources. I think this is getting as close to actual volume measurement as you can without actually having volume figures.

Maybe this is getting out there a bit but often times the money is made by venturing where others have not already flocked. Most likely this is being done by someone but I've yet to see it for forex data with tick level monitoring. I know there are sources where you can subscribe to data that displays in real time the average of several hundred banks and it is fairly jumpy -- but it is only the average and no acceleration or deceleration data is available. Also you'd need sub-second time stamps on the data and this is not available with a lot of the data sources out there. I was going to keep this to myself but since I've not done anything about it perhaps someone will. If you do please pm me and I'll be your first subscriber.

FT
  • #5
  • Jul 2, 2005 11:35am Jul 2, 2005 11:35am
Quoting merlin
nader, can you post the formula you are using? i will see if we can get it to work without volume.
The formula I know for calculating the accumulation/distribution is the normal one which is as follows:

X=(Close-Low)
Y=(High-Close)
Z=(High-Low)

Summation of {(X-Y/Z)*Volume}

As you can see, the volume here is a major element of the formula, so it can't be eliminated unless the formula itself gets some change.

What do you think ?
  • #6
  • Jul 2, 2005 11:52am Jul 2, 2005 11:52am
Quoting fijitrader
A&D cannot be measured by OHLC except in terms of movement. Low open, long range, high close is clearly accumulation if the time frame over which it happens is statistically significant. But when it happens during low volume is it real accumulation or just speculative opportunism? Additionally one of the most significant A&D indicators of continuation (rather lack thereof) is when there is high volume with little price movement. Completely undetected by OHLC and yet one of the most statistically relevant reversal indicators.

I've often thought that forex trading activity might be measured by designing a tick meter. Measure ticks per second or ticks per minute and combine that with some type of ROC measurement and you would be able to synthesize market acceleration and deceleration. The more data sources you monitored the more statistically significant your readings would be.

Say you had a Tick-ometer for each of 10 data sources measuring ticks/timeunit for each as well as the speed of ticks incoming and the rate of price acceleration/deceleration -- you could measure market behavioral patterns in ways that are not currently available in any conventional software that I know of. I think this would lead to some very interesting clues to market behavior particularly if you had data sources from the major exchanges and from several forex sources. I think this is getting as close to actual volume measurement as you can without actually having volume figures.

Maybe this is getting out there a bit but often times the money is made by venturing where others have not already flocked. Most likely this is being done by someone but I've yet to see it for forex data with tick level monitoring. I know there are sources where you can subscribe to data that displays in real time the average of several hundred banks and it is fairly jumpy -- but it is only the average and no acceleration or deceleration data is available. Also you'd need sub-second time stamps on the data and this is not available with a lot of the data sources out there. I was going to keep this to myself but since I've not done anything about it perhaps someone will. If you do please pm me and I'll be your first subscriber.

FT

Your idea about measuring the ticks/timeunit is excellent, but requires many data sources as you mentioned, which is practically very difficult and would be very costy.

The problem which I see with ticks/timeunit is that it will be deceptive some times in giving you an idea about market acceleration or deceleration. Sometimes a huge amount of buyers and sellers are exhanging a currency at a very confined trading range. This is a signal of either accumulation or distribution, and an early signal of market acceleration, although it won't be monitored in the tick/timeunit most of the times.

If you want my opinion, the best tickometer is the traders eye by watching the tick by tick chart. This of course is very time consuming and exhaustive.
  • #7
  • Edited 10:10pm Jul 2, 2005 9:40pm | Edited 10:10pm
narafa]
Quote
Your idea about measuring the ticks/timeunit is excellent, but requires many data sources as you mentioned, which is practically very difficult and would be very costy.
I've heard it said "you have to spend money to make money". I've found that it is usually true. Time can also be worth a lot of money. Therefore if this proves to be a valuable tool it will be worth a lot more than the cost. Also the exercise itself will generally reveal unexpected bonuses. That is often the way of research.

Quote
The problem which I see with ticks/timeunit is that it will be deceptive some times in giving you an idea about market acceleration or deceleration. Sometimes a huge amount of buyers and sellers are exhanging a currency at a very confined trading range. This is a signal of either accumulation or distribution, and an early signal of market acceleration, although it won't be monitored in the tick/timeunit most of the times.
Yes I already made that point. However tick count will normally increase when volume does even within a narrow range whereas OHLC will reveal absolutely nothing. Show me any way to gather forex volume that gets any closer without actually gathering volume data directly from a number of banks and averaging it.

Quote
If you want my opinion, the best tickometer is the traders eye by watching the tick by tick chart. This of course is very time consuming and exhaustive.

Chart staring is probably one reason why you posted a query about how to measure A/D.

Regardless my approach was not intended to eliminate the traders eye but to compensate for its inabilities to objectively quantify aspects of chart movement in real time.

First as far as A/D goes there will be nothing of A/D that a traders eye will reveal that is not made visible to the eye except by real time tick activity or historical charts of OHLC. If there is huge volume at one price only -- then it will not show up to the traders eye or by ticks or by any other measure than actual volume measurement.

However if there is huge volume and the price fluctuates rapdily in a small range there is a clue that there is a lot of activity there and probably a lot more volume than light trading where you see a tick and wait 5 minutes for the next tick. This sort of thing will be seen by a traders eye whether he is looking at a tick meter or a chart. One major difference is you could chart synthesized A/D patterns and look through history if you combined A/D with ROC to look at acceleration. Then you could research for patterns of acceleration/deceleration to identify possible continuance/reversal characteristics. Then apply statistical distribution models to determine the reliablity of the patterns. Basically I'm just talking about typical development scenarios but applied to information gathered in a way that is not widely being done, if at all.

Whatever your eye can tell you is driven by ticks. It would not matter whether you were watching a tickometer or a tick chart ... except a tickometer could reveal additional elements such as accurately quantified accelearation and deceleration which would make the acceleration and deceleration objectively measured instead of by the emotional filters we bring from day to day. One day we may be unimpressed by a certain level of acceleration and other days it may grab us a certain way and yet both degrees of acceleration might be precisely the same.

Another type of example: If the scale and aspect ratio of your chart are different from session to session due to previous volatility then you may react more strongly to a candle shooting up than on another day when the aspect ratio and/or scale are different because so many ratios have changed but you probably will not have made precise adjustments in your inner calibration of those ratios so that a candle shooting up one day may look like a little blip and another day may look like a giant sequoia. Either way it can help to use tools that maintain consistency. Acceleration and deceleration are very difficult to maintain standard frames of reference for when staring at candles. The very next candle can change the scale of your chart and if you were looking away at that second you may not have even noticed the change.

Anyway you asked how to measure A/D in Forex. If you've come to the conclusion that chart staring is the best way then no point in continuing the thread. I've spent a lot of time trying to come up with a better way and that is the best I've thought of or heard of. Other than trying to get a bunch of banks to cooperate by providing volume information. It could be done especially if you offered them a lot of money from volume data subscription revenue and there were 100 or more banks all sending their real time volume data to you. You could resell the agregate volume data back to mm's and banks everywhere, charge lots of money for it, and then it would gradually get factored into the market and soon lose a lot of relevancy. Or you can try to synthesize it with tools that are not widely in use and perhaps captalize on it before someone produces actual volume data.

FT
  • #8
  • Jul 3, 2005 1:28am Jul 3, 2005 1:28am
Quoting narafa
Your idea about measuring the ticks/timeunit is excellent, but requires many data sources as you mentioned, which is practically very difficult and would be very costy.
TradeStation measures tick volume automatically.

you could also use the volume from EC (euro futures contract). the futures contract mimics the EURUSD almost exactly. there is a futures contract for every major pair.
Relax and be happy.
  • #9
  • Jul 3, 2005 8:29pm Jul 3, 2005 8:29pm
Quoting merlin
TradeStation measures tick volume automatically.

you could also use the volume from EC (euro futures contract). the futures contract mimics the EURUSD almost exactly. there is a futures contract for every major pair.
The issue with TS when they were with RJO is that its forex volume has been far less than representative of the market. Now that they are moving to Gain it may be a bit more representative but I wonder about that.

Also TS has yet to incorporate subsecond time stamps on its ticks so measuring tick volume below one minute is not possible. When a billion dollar s of volume trades on the forex it can be bought or sold in much less than a minute when the liquidity is strong. Whereas the retail mm's are not likely to see any of that activity. Really big dollars don't often hit the retail markets. If China is buying US dollars they don't go to GC or Oanda.

Having said that one thing that is probably the simplest way to use TS to understand tick volume is to make two charts for the same pair one above the other and chart one in tick candles and the other in time period candles and compare the two. That would give a rough comparison of tick volume in a given time period.

Have you ever looked at a graph of the spot market VS the futures markets? The cash price VS the futures prices can diverge, cross, and converge. This would imply that volumes can differ substantially. Prices are defnitely not that representative. However I would not argue that some sense of volume may be gleaned from the futures markets and that it is more likely to be representative of forex volume than retail forex mm's.

FT
  • #10
  • Jul 23, 2005 3:47pm Jul 23, 2005 3:47pm
Quoting fijitrader
The issue with TS when they were with RJO is that its forex volume has been far less than representative of the market. Now that they are moving to Gain it may be a bit more representative but I wonder about that.

Also TS has yet to incorporate subsecond time stamps on its ticks so measuring tick volume below one minute is not possible. When a billion dollar s of volume trades on the forex it can be bought or sold in much less than a minute when the liquidity is strong. Whereas the retail mm's are not likely to see any of that activity. Really big dollars don't often hit the retail markets. If China is buying US dollars they don't go to GC or Oanda.

Having said that one thing that is probably the simplest way to use TS to understand tick volume is to make two charts for the same pair one above the other and chart one in tick candles and the other in time period candles and compare the two. That would give a rough comparison of tick volume in a given time period.

Have you ever looked at a graph of the spot market VS the futures markets? The cash price VS the futures prices can diverge, cross, and converge. This would imply that volumes can differ substantially. Prices are defnitely not that representative. However I would not argue that some sense of volume may be gleaned from the futures markets and that it is more likely to be representative of forex volume than retail forex mm's.

FT
i watch the futures contracts on the same chart as the spot market all the time. one of my main strategies is an arb system on the futures vs. spot.

have you looked at the tick data since GAIN took over at TS? i think it just happend a week or two ago. i never looked at the tick charts, so i didnt know about the time stamp issues.
Relax and be happy.
  • #11
  • Jul 24, 2005 8:33am Jul 24, 2005 8:33am
Quoting merlin
i watch the futures contracts on the same chart as the spot market all the time. one of my main strategies is an arb system on the futures vs. spot.

have you looked at the tick data since GAIN took over at TS? i think it just happend a week or two ago. i never looked at the tick charts, so i didnt know about the time stamp issues.
I cancelled TS when I heard about the move to Gain. Not that I think Gain is worse but I had held onto a dream that TS was going to make their platform available to more than one MM. More than a year ago I took up the issue with TS and blasted them a bit for tying into a single market maker, and a really small player to boot. I enquired why one of the best platforms in the futures markets dropped their pants to one of the least deserving of mm's. They indicated they were naive about Forex and signed a contract with RJO... and a bunch of other blather.... but that they were going to correct the problem within the year. Also I spoke with FXCM who indicated they were working with TS about the issue and were confident that TS would open up the doors for wider support to various mm's within the industry. When Gain got it that was just too much. Also even if Gain's data has sub-second time stamps (useful only for programming nor for viewing) TS's platform will not read it.

FT
  • #12
  • Aug 11, 2005 2:48pm Aug 11, 2005 2:48pm
fiji, are they even allowed to use more than 1 MM? ive never seen that anywhere in forex, but it sure would be nice.
Relax and be happy.
  • #13
  • Last Post: Aug 12, 2005 3:39am Aug 12, 2005 3:39am
Quoting merlin
fiji, are they even allowed to use more than 1 MM? ive never seen that anywhere in forex, but it sure would be nice.
Allowed by whom? Who could prevent tradestation from supporting more than one mm? Strategy builder supports a number of market makers.

FT
0 traders viewing now