The ACD Method

I realize you are trading more on a weekly basis and your you can use timed stops effectively since when trades dont go in your favor, you get out most probably.

On a relative basis, only 30% or less of trades are profitable, because I get out early if trades dont go in my favor right away.

Have you experienced similar stats in your win rate trading?

Also "On confirmed trades, i stop out if price CLOSES back inside the A levels."
Are you taking daily A Levels for Weekly trades or Weekly A Levels for Weekly trades?

When you say closes - are you talking about close of session, apologies for the questions - but I m just trying to understand

Quote from Maverick74:

All my stops are ACD based not fixed. So on failed trades it's pretty straightforward. I use either the A level or the wick outside the A level. On confirmed trades, i stop out if price CLOSES back inside the A levels. a lot of people won't like this because your stop can vary but to me if you get long on an A up and price closes back below the A up, that is a very ominous sign "usually".

Another stop I use is the confirmation stop. In other words, say I get short on a confirmed monthly A down then the following week we get a confirmed weekly A up, then I stop out at that confirmed level because the market is confirming strength. On a weekly, you could stop out on a daily, etc. This goes along with the time stop concept. If price confirms, I want it to run, not pullback inside the A levels. I realize you will prematurely stop yourself out of good trades that way, but the best trades then screw around, they just take off. Those are the ones you want.
More...
 
Quote from Maverick74:

All my stops are ACD based not fixed. So on failed trades it's pretty straightforward. I use either the A level or the wick outside the A level. On confirmed trades, i stop out if price CLOSES back inside the A levels. a lot of people won't like this because your stop can vary but to me if you get long on an A up and price closes back below the A up, that is a very ominous sign "usually".

Another stop I use is the confirmation stop. In other words, say I get short on a confirmed monthly A down then the following week we get a confirmed weekly A up, then I stop out at that confirmed level because the market is confirming strength. On a weekly, you could stop out on a daily, etc. This goes along with the time stop concept. If price confirms, I want it to run, not pullback inside the A levels. I realize you will prematurely stop yourself out of good trades that way, but the best trades then screw around, they just take off. Those are the ones you want.
More...

Nice, thanks for sharing. I assume for a daytrade a 5-minute close below/above the A level would get you out. From what I've seen so far this does seem to weed out the bad trades.
 
One thing I've noticed with A-ups/downs. When there is an initial bounce off of an a-level and it temporarily holds...this would be a fade trade. If the fade trade fails and the market pushes back through the A-level this is usually a good place to enter the trade. In other words, a time confirmation is not always necessary... sometimes a failure of the fade can be a good entry.
 
Quote from acdtrader:

I realize you are trading more on a weekly basis and your you can use timed stops effectively since when trades dont go in your favor, you get out most probably.

On a relative basis, only 30% or less of trades are profitable, because I get out early if trades dont go in my favor right away.

Have you experienced similar stats in your win rate trading?

Also "On confirmed trades, i stop out if price CLOSES back inside the A levels."
Are you taking daily A Levels for Weekly trades or Weekly A Levels for Weekly trades?

When you say closes - are you talking about close of session, apologies for the questions - but I m just trying to understand
More...

On the longer term trades, you get less chop and less false breakouts. So when price breaks out, it runs. I think the win rate on confirmed trades longer term is probably closer to 75%. Because you are trading the longer term trend which is more reliable. Daytrading has a lot of noise so you need to account for that.
 
Quote from Shanb:

Nice, thanks for sharing. I assume for a daytrade a 5-minute close below/above the A level would get you out. From what I've seen so far this does seem to weed out the bad trades.
More...

Yeah if you want to trade pure momentum, then don't screw around with trades that breakout and pullback. Just keep saying "next" and move on.
 
Quote from Shanb:

One thing I've noticed with A-ups/downs. When there is an initial bounce off of an a-level and it temporarily holds...this would be a fade trade. If the fade trade fails and the market pushes back through the A-level this is usually a good place to enter the trade. In other words, a time confirmation is not always necessary... sometimes a failure of the fade can be a good entry.
More...

Yes, the best A trades are the ones that initially fail right at the A level then come back up and go through. you can immediately take that trade and not wait for the confirmation.

Again, this is all just my opinion. I'm often wrong. I said the market was going to be choppy today! :)

Good thing I'm not daytrading.
 
I'm going to share something I'm looking at and forward testing. It deals with the number line. To me, the number line is one of the most fascinating things to come out of Fisher's book as it is essentially a daily checklist for the health of the market and measures something most people can't see with the naked eye.

One of the issues I always had with the 20 day number line is what I call the beginning of the universe theory. That is, when did the universe really begin. Except translated to the number line, I want to know, when did the number line begin. Yes, I know it goes back 20 days ago, but when you start impacts the line count since the number is calculated on a "rolling" basis. If one person started their 20 day number line 30 days ago and I started mine 20 days ago, we are going to end up with different numbers because at data point number 10 you are bringing a number line value into that day where as I'm starting at zero. Hell you might have a plus 9 number line at that point and I'm starting at zero!!!!

So to me, it only seems "logical" to match number lines with the cycle you are trading in. So I wanted to devise a plan that has weekly, monthly and QTR number lines. The weekly is kind of tricky due to lack of data points. But the monthly and QTR make sense. The idea is you "re-set" the number lines back to zero at the start of a new cycle just as you do with a new opening range or pivot range. This allows one to match up a number line value with "specific A values".

I'm putting this idea out here onto ET since it's a work in process and perhaps the minds of other traders might shorten my workload. I'm cunning like that. :)

So today for example, we have a new week, month and QTR and I'm starting all values at zero. The QTR number line will behave just like the old Fisher number line which is a rolling 20 day value that resets at the beginning of each QTR. The monthly will reset every 20 days. The weekly will be slightly different. Still working on the "different" part.

These are the thoughts from inside my brain. We'll see what ACD ET community comes up with.
 
Quote from Maverick74:

I'm going to share something I'm looking at and forward testing. It deals with the number line. To me, the number line is one of the most fascinating things to come out of Fisher's book as it is essentially a daily checklist for the health of the market and measures something most people can't see with the naked eye.

One of the issues I always had with the 20 day number line is what I call the beginning of the universe theory. That is, when did the universe really begin. Except translated to the number line, I want to know, when did the number line begin. Yes, I know it goes back 20 days ago, but when you start impacts the line count since the number is calculated on a "rolling" basis. If one person started their 20 day number line 30 days ago and I started mine 20 days ago, we are going to end up with different numbers because at data point number 10 you are bringing a number line value into that day where as I'm starting at zero. Hell you might have a plus 9 number line at that point and I'm starting at zero!!!!

So to me, it only seems "logical" to match number lines with the cycle you are trading in. So I wanted to devise a plan that has weekly, monthly and QTR number lines. The weekly is kind of tricky due to lack of data points. But the monthly and QTR make sense. The idea is you "re-set" the number lines back to zero at the start of a new cycle just as you do with a new opening range or pivot range. This allows one to match up a number line value with "specific A values".

I'm putting this idea out here onto ET since it's a work in process and perhaps the minds of other traders might shorten my workload. I'm cunning like that. :)

So today for example, we have a new week, month and QTR and I'm starting all values at zero. The QTR number line will behave just like the old Fisher number line which is a rolling 20 day value that resets at the beginning of each QTR. The monthly will reset every 20 days. The weekly will be slightly different. Still working on the "different" part.

These are the thoughts from inside my brain. We'll see what ACD ET community comes up with.
More...

Your approach makes sense.

I've had the same problem with composite volume profiles: taking the volume-by-price of n days to create one big profile to anticipate the "meat of the market" better. For day-to-day, it's easy. But once you start looking at many days it's important to know <i>which</i> days are significant to include. For instance, 8/9/11 on the equity indices was a very significant bar. Any subsequent 3-day rolling period that doesn't include that day's price and volume would be delivering less useful info.

I've been wanting to do the same as you because I swing trade and the number line is perfect for that. But that would require a month of coding as I'd have to evaluate whether we've made A-Ups or A failures in a new study, which would require external efs calls to the ACD study I've built in eSignal. The new eSignal doesn't seem to handle that well.

But the way I was thinking of it was exactly as you state it: calculate from the dominant OR period -- one that's significant to what you're trading -- and reset every cycle. Since flip demonstrated how the first of the month is actually significant, that's where I eventually plan to start. Actually that and the first Thur-Tue every month, which fits better for how I see things.
 
Quote from Maverick74:

I'm going to share something I'm looking at and forward testing. It deals with the number line. To me, the number line is one of the most fascinating things to come out of Fisher's book as it is essentially a daily checklist for the health of the market and measures something most people can't see with the naked eye.

One of the issues I always had with the 20 day number line is what I call the beginning of the universe theory. That is, when did the universe really begin. Except translated to the number line, I want to know, when did the number line begin. Yes, I know it goes back 20 days ago, but when you start impacts the line count since the number is calculated on a "rolling" basis. If one person started their 20 day number line 30 days ago and I started mine 20 days ago, we are going to end up with different numbers because at data point number 10 you are bringing a number line value into that day where as I'm starting at zero. Hell you might have a plus 9 number line at that point and I'm starting at zero!!!!

So to me, it only seems "logical" to match number lines with the cycle you are trading in. So I wanted to devise a plan that has weekly, monthly and QTR number lines. The weekly is kind of tricky due to lack of data points. But the monthly and QTR make sense. The idea is you "re-set" the number lines back to zero at the start of a new cycle just as you do with a new opening range or pivot range. This allows one to match up a number line value with "specific A values".

I'm putting this idea out here onto ET since it's a work in process and perhaps the minds of other traders might shorten my workload. I'm cunning like that. :)

So today for example, we have a new week, month and QTR and I'm starting all values at zero. The QTR number line will behave just like the old Fisher number line which is a rolling 20 day value that resets at the beginning of each QTR. The monthly will reset every 20 days. The weekly will be slightly different. Still working on the "different" part.

These are the thoughts from inside my brain. We'll see what ACD ET community comes up with.
More...


Based on MF subscription service, they look at 30 days number line. They also give 12 days number line but I have been told they do not use it any more . What are the reasons for using 20 days vs 30 days number line.

So far for day trading number line has not been of much help. For example, 30 days number line for CL for last several days are

starting 09/19/2011 till 10/03.

8,14,16,13, 11,14,14,10,12,12.

looking at these 30 days number line for day trading would have put me out of business. Maybe , I am looking at them the wrong way.
 
Quote from mfbreakout:

Based on MF subscription service, they look at 30 days number line. They also give 12 days number line but I have been told they do not use it any more . What are the reasons for using 20 days vs 30 days number line.

So far for day trading number line has not been of much help. For example, 30 days number line for CL for last several days are

starting 09/19/2011 till 10/03.

8,14,16,13, 11,14,14,10,12,12.

looking at these 30 days number line for day trading would have put me out of business. Maybe , I am looking at them the wrong way.
More...

Actually, meant to say 30 day number line for the QTR.

You might be using the number line incorrectly. The number line doesn't tell you to get long or short, it's meant to confirm price action. It's meant to show trends that are not so obvious.

The way I use it would be let's say CL has a plus 12 number line. I would aggressively look for failed A downs to get long and good A ups above the pivot to get long. I might be hesitant to take any A downs.

Another example. Say the number line in CL is 3. And we get a good A up today, I might be inclined to not take the trade due to the choppy number line or bid only on pullbacks questioning whether or not it really has the ability to follow through.

But I would never take a stand alone trade based solely on the number line.
 
You must log in or register to reply here.
Back
Top