I've only done 2 so far.
Why does TA not work (for you)?
- Thread starter Xspurt
- Start date
I've only done 2 so far.
Quote from cornixforex:
Guess it depends on what's meant by "TA". If it's TA in the sense of wrbtrader, then hardly one can beat a human trading nearly every day and staying in sync with market evolution. That's the beauty of creative TA approach, you adjust to the game depending on conditions, algos can't change that fast, because they first need to be reprogrammed by humans.
CF
If the below is what you referencing; Yup
RN
I put on less than one "intuitive" trade out of every 20. Intuitive means I took a break, returned to my desk and see a move-in-progress and I missed the entry price I would've used per my trading plan. However, I "intuitively" know that a strong trending move on its way to the next key S/R level has a better than 50/50 chance of getting to that level and I enter the market outside my plan.
The other 95% of my trades are based on explicit rules based on technical price action setups. If I see setup 1, I place a stop entry order at price A; if I see setup 2, I place a limit order at price B; if I see setup 3, I place a stop entry order at price C.
Each setup is clearly defined in detail (range of the signal bar, proximity to previous S/R levels, chart pattern on smaller time frame, price environment, price levels at which traders in may get trapped into or out of a position and be forced to bail or chase).
The trade management rules are also defined in detail.
I have around 20 trades a day and although the distribution of profits, losses and scratch trades can be somewhat random during certain times of day, the odds of a profitable day trading these same setups over and over again are light years from 50/50.
Well here is one that may provoke discussion: Long term charts. What is their value ? Please use this one to start the discussion around. It is simple enough.
I'm shopping this inquiry around anyway..
Without using any software or metrics and just eye balling........... does a 10 year chart for a long term pairs trader willing to stay in the trade for many weeks to a few months have any usefulness at all. In other words, momentarily, disregard more significant short term modalities in the decision making of putting on a trade here.
The operative phrase here: "any use at all" for the prospect of shorting HOT and going long HST.
<img src=http://www.elitetrader.com/vb/attachment.php?s=&postid=3598750>
I'm shopping this inquiry around anyway..
Without using any software or metrics and just eye balling........... does a 10 year chart for a long term pairs trader willing to stay in the trade for many weeks to a few months have any usefulness at all. In other words, momentarily, disregard more significant short term modalities in the decision making of putting on a trade here.
The operative phrase here: "any use at all" for the prospect of shorting HOT and going long HST.
<img src=http://www.elitetrader.com/vb/attachment.php?s=&postid=3598750>
Quote from frenchfry:
seems I'm missing some of your trades. They are marked with green and red arrows.
Maybe that's not the only thing I missed...
Attached is a major clue for you in doing the volume bar by bar.
as you see I did the trade as a quickie one bar before you did.
the quickie is three T1's . You set up a name set for all the quickies.
On the third consecutive T1 (interruptions do not kill in this case) consider a trend to be out of gas.
there is a slight up tick in the bar compared to the priior bar.
I circled the first T1 and the second T1 and I did a reversal (see vertical line) on the thrid T1.
I am doing my job to point out TA trends. And especially the trend end. FYI the hold began on bar 27 long.
Here is the MAJOR LESSON on the table.
You are showing me that you do not label each volume bar.
I do label each volume bar.
The result is this: I always "know that I know."
In making money you do the work assiduously. What happens to your reversals? Making money takes care of itself.
I hve been explaining the master coding concept for markets. It has a name. The name is Relative.......Data Base.......Management....System.
sead it backwards.
Imagine you are surf.
He is a dead duck at systems
He is a dead duck at management
He is a dead duck at data base
He is a shredded ripped apart dead duck at relative.
Each bar is relative to the prior bar unless that prior bar is SQUISHED into its cast shaddow. You are X ing the shaddow data in volume that is ignored. surf uses the ignore concept.
Who does surf hve on ingnore? The Messanger.
Here is your message, french fry: Pretend you are a computer. You keep records. You have 24 record tables.
You use RDBMS to keep those tables perfect.
Lets look at logic man's tables:
He has three he named as follows:
Bar of beginning of trend.
Middle of trend bars
End of trend bar.
his logic but each bar, consecutively into a table.
He gets a lot of surprises.
you work as above and you get no surprizes, ever.
You have tables to fill that always tell you that you know that you know.
I could do a simple trading method here and now that would have all the ingresients surf says do not exist anywhere. It is the one page called "un something Volume."
It has three colors of rules. the rule has a lookup column. the laguage coded test is a go/nogo test.
I asked Bowlinsky why, for years, he could not code it. Fidelity has a weak system for snippeting an ATS.
The RDBMS of PVT has one multicolumn lookup table and five rules to use the whole thing. I posted 62 trades where one was BE and one had loss (operator error). the worden Bros calculated Sharpe Ratio for this system is over 60 using a poor universe (NAZ 100. worden cannot build universes. See Universes built using WJO"N criteria...LOL..... I click of a button and the stock are lined up in orde for the hot listing.
You must do what the ATS does.
It puts each bar in its category.
Pi ..T1..T1...T1 is a whole quickie tend. there are 6 other quickies. I posted them as calls or as results already.
they go into a table. any bar is handled with two procdures one after the other. Pocedure 1 is to determine if the bar is an EE. After thatIF IT IS NOT, then you follow the test procedure for categorizing the bar. I wrote this out in a step by step procedure that uses your log rows which list events one after another (this is what logic man has not gotten to as yet. If he adds that step he will have shifted from up/down to right/lf mentality.
Look at the blank bars. there are no blank bars in ATS records.
Why can't the financial industry program to make money?
They do not use the correct market language, ever. they do stats and look for edges. stupid is as stupid does.
Begin to think RELATIVELY.
See the Order Of Events on the PATTERN you have ben given.
This is not 10,000 hours. It is not 1,000 hours. It is like fifth grade.
since there are only so many names for bars, there are only so many sequences possible. the sequences fall into two piles: continuation and change. Dodd and someone else (Granville) said this at the time the DOW Thoery was invented using....ugh...ugh ..... statisitics.
A complete paradymn is used. It has a correct Parametric Measure. Finite math has to be used.
It is all AND and OR circuits. The language in the computer does go/nogo tests only.
Look at the category test procedure and the log I gave ocean 5 to build. All he had to do was true/false tests to make money. He is lazy and has no long term memory built.
Here is the attachment.
the support of learning:
Label each bar.
Then you "know that you know".
the support of learning:
Label each bar.
Then you "know that you know".
If I have to think like an ATS then I have to label EACH volume bar. Possible labels:
1. P(eak), T(rough), X(don't count), I(ncreasing), D(ecreasing) = Two signify change and three continuation
2. P1, T1, P2, T2, P3, X(don't count), I(ncreasing), D(ecreasing)
I could even leave out I and D and simply use:
3. P1, after that every decreasing bar is a T1 until the next rising volume when I start to label the bars P2 as long as the bars are rising. The T2 follows until the next rising bar which is then P3. From there on The sequence starts again. I also add X for bars which are "shaddowed".
1. P(eak), T(rough), X(don't count), I(ncreasing), D(ecreasing) = Two signify change and three continuation
2. P1, T1, P2, T2, P3, X(don't count), I(ncreasing), D(ecreasing)
I could even leave out I and D and simply use:
3. P1, after that every decreasing bar is a T1 until the next rising volume when I start to label the bars P2 as long as the bars are rising. The T2 follows until the next rising bar which is then P3. From there on The sequence starts again. I also add X for bars which are "shaddowed".
Correct. Leave out the I and D or you will be bck a the logic man level.
we know that move 1 of price is P1, T1 and P2.
we know that move 2 of price is P2, T2.
We know hat move 3 of price is T2 to P3.
above you see the right/left orientation replacing the up down orientation.
So we need to add some detail to the "after point 3" (which is T2)
I showed you the approach on the Tuesady last chart between bars 8 and 14. Because I am old and my memory is built and tuned precisely, anything I ever see is mapped in my long tem memory automatically.
this thread is not even 200 pages long as my ET setting counts pages. The bootcamp was printed out at 1938 pages including illustrations. It took two months for an MD to buy a colored printer in those days. Now we have a person (surf) posting stuff that is very humorous in terms of the neuro science shutdown of a mind.
you will be able to deal with the quickies early in the trends. they are just FBO's of several sorts.
The quest is to wrap up the items that happen after point3 which in volume is T2.
I have made the possibilites clear:
VE's, fanning, and staying within the boundaries.
The "normal thing is none of these exceptions.
you noticed and recorded the criteria for repeating of P1, T1, and P2. After T1 there are no more P1's. After P2's thee are no more T1's. Except for this. Once a P2 has happened there can be a new P1. Once a T2 has happened thee can be a P2.
The above paragraph is what sts us apart from the normal logicians.
this concept repeats for the returning dominant sub fractal thrid move of a trend.
Why none of this occurred to anyone Covel inteviews or wrote up is because they all do not use logic as found in the proper applied algebra. Each numeric base has its own algebra.
So th market granularity dictated a base two algebra.
What is MOST DIFFICULT to catch is the measure of the values in this algebra.
lets look at a price action (PA) setting. Notice all users are the same in their mesurement orientation. What if they could recognize what they are not doing.
View the temple Grandin story. See how two mental thems unfold. redneck cowboys and cattle. Temple took the view of the cattle.
the red necks hung their checkered shirts on her curving walls because they did not know what the shirts symbolized to the cattle. temple put the shirts were the cattle wanted then: in the dirt out of sight.
you learned to do the shadows in a day or so.
now you learned to categorize bars AFTER you check to see if an End Effect of a trend happened. you have the two tests in the proper order.
We need to duplicate all of this for the dominant sub fractal of the third move of a trend.
Temple made solid walls instead of se through fences. She let cattle guide themselves along curves to slow change direction to get to where they could die happy.
I have studied 20 differentt kinds of people to see how they can be cured of mental ills by doing workarounds in the mind. befoe neuroscience I invented electronic sub-assembly testing that changed an industry forever. temple changed the cattle industry forever and she changed the treatment of autism forever. the keyboard cures autism quite simply using the keyboard as a workaround.
We need to perfect the third move of the trend.