Technical Analysis alone will not give you an edge | Page 6 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

"You cannot make money (in the long run) from a price chart alone" 37 replies

An Edge for an Edge (System Development) 116 replies

Does an EA license Give You the Right to Give Signals? 2 replies

Give me premise or give me death.... 15 replies

How certain can you be that you have an edge? 79 replies

Attachments: Technical Analysis alone will not give you an edge
Exit Attachments

Technical Analysis alone will not give you an edge

  • #101
  • Nov 1, 2015 2:08pm Nov 1, 2015 2:08pm
Quoting the redlion
{quote} If this is true then just like CJ, GEFX and many others are claiming...we would see patterns, and I agree with you except when you start stating that these "special patterns" are outside of the purview of objective analysis. Somehow only our brains can detect them but our machines and test cannot, the market is so special it is a language, a living organism, a place where all the big money create traps.... forgive me but it doesn't sound right to me. if in fact the market acted the way you say it acts a simple runs test...
Believe it or not....that's exactly what happens in the mind of all experienced discretionary traders. After years of experience you develop "sixth sense" and all comes naturally. Instant pattern recognition becomes second nature. You just do it automatically. But not all people can do that. Maybe some are just not wired for this(just like not everyone can be musician), others are capable, but possess to many wrong beliefs and biases preventing them from being successful......
Successful trading is only part science, and alot more art and performance oriented discipline very similar to individual sports. When you are at your best, you don't think at all, you just perform. The subconscious mind tells you what to do. You just use that huge library of memories stored in you brain over the years. Thousands of patterns...situations..mistakes, good trades, bad trades.....everything you have ever seen, read or heard, all memories, and naturally you know what to expect, and when the probabilities are in your favor.
The human brain is just AWESOME compared to computers! And we don't even understand how it works! Computers are only better at fast calculations(dry numbers) and processing large amount of data... nothing else.
Why would you even expect the machine to detect the patterns? This will never happen.. just not possible....This is like expecting the computer to come up with idea or invent something . You need artificial inteligence for this, and we dont have it (yet).
  • #102
  • Nov 1, 2015 2:09pm Nov 1, 2015 2:09pm
@redlion
I am no expert in statistics.Your statistics too complicated for me.And I agree with view of CJ and some others that price datas is not random because market price as I see it is created by market itself(retail,MM,smart money,hedgers,banks,etc etc).Also as I see it,although technical analysis formula are math based but TA does not mean only using the formula.Studying technical chart is more than just using indicator based on math based formula,it is a more an art. So it may not be appropriate to use only statistic methods assuming market price is random.Thats only what I want to stressed.
-CKB
  • #103
  • Nov 1, 2015 2:11pm Nov 1, 2015 2:11pm
Quoting alphaomega
{quote} Believe it or not....that's exactly what happens in the mind of all experienced discretionary traders. After years of experience you develop "sixth sense" and all comes naturally. Instant pattern recognition becomes second nature. You just do it automatically. But not all people can do that. Maybe some are just not wired for this(just like not everyone can be musician), others are capable, but possess to many wrong beliefs and biases preventing them from being successful...... Successful trading is only part science, and alot more art and...
For me, my mind is not reliable enough to rely on the storage of all the patterns I see... but screen caps save the day!


-CJ-
A man is not defeated when he is beaten, he is defeated when he quits.
  • #104
  • Nov 1, 2015 2:17pm Nov 1, 2015 2:17pm
Quoting alphaomega
{quote} Believe it or not....that's exactly what happens in the mind of all experienced discretionary traders. After years of experience you develop "sixth sense" and all comes naturally. Instant pattern recognition becomes second nature. You just do it automatically. But not all people can do that. Maybe some are just not wired for this(just like not everyone can be musician), others are capable, but possess to many wrong beliefs and biases preventing them from being successful...... Successful trading is only part science, and alot more art and...

Machines do recognize patterns, it's how harmonics work. If you were to take any chart pattern, say the EURUSD map, and convert all of the numbered points on the chart, into ratios and angles between the lines and points, the pattern would be "seen" by a machine. Not just THAT pattern, but any pattern, larger or smaller with the same ratios and angles. Same as harmonics.


Harmonics are actually finding the child pattern on say a 5m chart as well as the SAME, mature pattern when it forms on the H1, H4 and D1... it is the same pattern., not a new or similar one.

-CJ-
A man is not defeated when he is beaten, he is defeated when he quits.
  • #105
  • Nov 1, 2015 2:26pm Nov 1, 2015 2:26pm
spoofing will help with TA
All Assets NAS100 GLD BTC AUS JPY GBP All Time Return: 19.7%
  • #106
  • Nov 1, 2015 2:32pm Nov 1, 2015 2:32pm
Quoting alphaomega
{quote} Believe it or not....that's exactly what happens in the mind of all experienced discretionary traders. After years of experience you develop "sixth sense" and all comes naturally. Instant pattern recognition becomes second nature. You just do it automatically. But not all people can do that. Maybe some are just not wired for this(just like not everyone can be musician), others are capable, but possess to many wrong beliefs and biases preventing them from being successful...... Successful trading is only part science, and alot more art and...
this topic certainly is interesting and stimulates brainstorming idea.can human invent a machine that is more intelligent than him/her? which statistics/maths is still not invented which can correctly simulate the nature of forex market price and predict its outcome? until that time, we may all have to go on speculating the nature of forex rice
-CKB
  • #107
  • Nov 1, 2015 2:43pm Nov 1, 2015 2:43pm
Quoting mirrfx
{quote} this topic certainly is interesting and stimulates brainstorming idea.can human invent a machine that is more intelligent than him/her? which statistics/maths is still not invented which can correctly simulate the nature of forex market price and predict its outcome? until that time, we may all have to go on speculating the nature of forex rice
In time possibly, but there are an infinite number of pattern possibilities. The ratios of which would all have to be programmed in as "features" in able for a machine to recognize them. Subsets, or smaller samples of patterns can be entered and those patterns would be found when printing. Perhaps you could train a neural net to learn these ratios as it goes, on the fly so to speak. I knew of one that could recognize the BARF pattern.

Until then, we are left with our brains, which do the computations effortlessly. All we have to do is look at the patterns and not all of the diversions and distortions that are created to draw attention away from them.

Again, this is along the lines of prediction and is not needed, only the patterns of the MM themselves. The cycle, the rhythm of the markets. These are patterns in themselves and are easily recognized as well.

-CJ-
A man is not defeated when he is beaten, he is defeated when he quits.
  • #108
  • Nov 1, 2015 2:49pm Nov 1, 2015 2:49pm
Quoting Captain Jack
{quote} Machines do recognize patterns, it's how harmonics work. ................ -CJ-
Yeah sure, machines can recognize some simple technical patterns. But where do these patterns came from? From the human brain! First you have to SEE the pattern and THEN you program the computer to search and to recognize this pattern.......and by the time the computer finds one, it is already useless...because the market is in another phase .... and the fundamentals are different. So you need to find new pattern.....and we go back to the beginning.....this just is hopeless approach!
The human brain can do all this process in a few seconds. A good trader(with experience) will instantly adapt to new situations while producing multiple trading ideas on the spot.
And when I talk about patterns, I mean combination(mix) of fundamentals and technicals. (context). I personally don't even use charts.....there are more efficient ways to look at data......but that's another topic.
I think people should start to think outside the box here. Because the Top traders of the world do not beat the market using conventional thinking.
  • #109
  • Nov 1, 2015 3:12pm Nov 1, 2015 3:12pm
you have to go back before you done the math as to why you think maths has anything to do with it ,the whole purpose of the thing is to measure what others think(the majority) in some way ,there are actually 100s of languages ,candles would be one of them



The candlestick techniques we use today originated in the style of technical charting used by the Japanese for over 100 years before the West developed the bar and point-and-figure analysis systems. In the 1700s, a Japanese man named Homma, a trader in the futures market, discovered that, although there was a link between price and the supply and demand of rice, the markets were strongly influenced by the emotions of traders. He understood that when emotions played into the equation, a vast difference between the value and the price of rice occurred. This difference between the value and the price is as applicable to stocks today as it was to rice in Japan centuries ago. The principles established by Homma are the basis for the candlestick chart analysis, which is used to measure market emotions surrounding a stock.



Quoting the redlion
{quote} If this is true then just like CJ, GEFX and many others are claiming...we would see patterns, and I agree with you except when you start stating that these "special patterns" are outside of the purview of objective analysis. Somehow only our brains can detect them but our machines and test cannot, the market is so special it is a language, a living organism, a place where all the big money create traps.... forgive me but it doesn't sound right to me. if in fact the market acted the way you say it acts a simple runs test would be able to...
On-site guru & also FF member's' psychiatrist, when not drinking tea
  • #110
  • Nov 1, 2015 3:26pm Nov 1, 2015 3:26pm
how does something natural like fib end up in charts,
"Did you know that Fibonacci numbers are found in nature as well? In fact, we can see examples of the Fibonacci sequence all around us, from the ebb and flow of ocean tides to the shape of a seashell. Even our human bodies are examples of Fibonacci. Read more about the fascinating phenomenon of Fibonacci in nature."

http://www.safehaven.com/article/272...-golden-spiral
On-site guru & also FF member's' psychiatrist, when not drinking tea
  • #111
  • Nov 1, 2015 3:34pm Nov 1, 2015 3:34pm
odds are with the bears ,whats the calculation
Attached Image(s) (click to enlarge)
Click to Enlarge Name: USD (DFB).png Size: 20 KB
Click to Enlarge Name: USD close(DFB).png Size: 18 KB
On-site guru & also FF member's' psychiatrist, when not drinking tea
  • #112
  • Nov 1, 2015 7:00pm Nov 1, 2015 7:00pm
pittyful... oging the hog with fundy meansss ... statiztical ... they have biggerr earr then ... the retail trader ... it meanss . they know sometigyy before retail know .. that falsss in to ... inside traDING ....

OOOPPPSS ... THAT COUNTRY THAT SAID ... FEWS POST UP ...

WITH TA INTACK SO BEARRRISH ... IT STILLL BULLSSHHITTLLYYY HEALTY ...

Thats up to the upppps .... stargate 61 trading ... is in season ...
  • #113
  • Nov 1, 2015 7:03pm Nov 1, 2015 7:03pm
patrent trading ... simlpy ..said its a bout 7 patren is contenueing , yes it can be code ... yes it can be easyllyy seen by , naked eyes .

no biggy ... nottin .. nu ...
  • #114
  • Nov 1, 2015 7:25pm Nov 1, 2015 7:25pm
Here are some charts...do you think you can trade them? do you see the patterns?
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 28 KB

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot2.png Size: 28 KB

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot3.png Size: 26 KB

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot4.png Size: 26 KB

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot5.png Size: 27 KB

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot6.png Size: 28 KB

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot7.png Size: 27 KB

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot8.png Size: 26 KB
AVT INVENIAM VIAM AVT FACIAM
  • #115
  • Nov 1, 2015 7:28pm Nov 1, 2015 7:28pm
very very clear
  • #116
  • Nov 1, 2015 7:35pm Nov 1, 2015 7:35pm
Quoting the redlion
Here are some charts...do you think you can trade them? do you see the patterns? {image} {image} {image} {image} {image} {image} {image} {image}
here is a simple analysis of a random chart.......you could clearly see support and resistance levels as well as price rejections around those levels.
It is completely random, that means that the next candle has nothing to do with the previous ones no matter what pattern or levels it shows on the chart

Attached Image (click to enlarge)
Click to Enlarge Name: analysis of random chart.jpg Size: 146 KB
AVT INVENIAM VIAM AVT FACIAM
  • #117
  • Nov 1, 2015 7:46pm Nov 1, 2015 7:46pm
Quoting dkrock
{quote} It is proof that your analysis is correct, and your method does not work. You do not need to further explain its failure. You need to move on. HINT: Stop trying to be so precise. Precision kills in trading
please explain ... I am not sure what you mean? I posted a random chart with some simple S/R analysis... to illustrate you can do those on random charts.

I can do a simple fibonacci retracements and projection analysis

I can do some BB strategy..... my point was that random charts show patterns even candle patterns.
AVT INVENIAM VIAM AVT FACIAM
  • #118
  • Nov 1, 2015 8:07pm Nov 1, 2015 8:07pm
Quoting the redlion
{quote} please explain ... I am not sure what you mean? I posted a random chart with some simple S/R analysis... to illustrate you can do those on random charts. I can do a simple fibonacci retracements and projection analysis I can do some BB strategy..... my point was that random charts show patterns even candle patterns.
Exactly Red. Here is a good vid from someone who has obviously come to a similar conclusion. It is worthwhile being patient and listening to it in its entirety as almost every traditional TA assumption is thrown at it.

This just confirms some heavier duty attempts to validate long term trading performance using TA through the scientific method which places significant doubts on much of it's validity. In this game of uncertainty that deals with complex systems, the only tool we have to test a strategies validity is by necessity a statistical tool. It is the tool of choice for any industry segment that deals with complex systems from meteorologists to the insurance and finance industries. If it doesn't stack up with statistical measures then it pays to be sceptical.
  • #119
  • Nov 1, 2015 8:23pm Nov 1, 2015 8:23pm
Quoting Copernicus
{quote} Exactly Red. Here is a good vid from someone who has obviously come to a similar conclusion. It is worthwhile being patient and listening to it in its entirety as almost every traditional TA assumption is thrown at it. This just confirms some heavier duty attempts to validate long term trading performance using TA through the scientific method which places significant doubts on...
Thanks for the Link, since you and I have similar views on this subject this is more akin what I tend to use nowadays
I look at the big fundamental picture starting with the macroeconomic environment, then I look at both Central banks and their policies as well as their hints, I usually read different opinion's from economist who know more about the subject than me and I analyse those views to help me make a more informed decision medium term strategy.

then I ascertain the drift through statistical analysis
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 156 KB


you see in the next month the currency is likely to move in many different ways
p0<-p
r<-rnorm(x,mean=last(SMA(eurusd.log,n=22*12)),sd=last(runSD(eurusd.log,n=22)))
log.price<-log(p0)+cumsum(r)
prices<-exp(log.price)

so i simulate a monte carlo model , and basically manage my trade from a gambler's ruin money management perspective as a coin flip but using monthly data as my binary reference point. so far I am holding a short position from 1.1350 level and will actively manage this trade according to price movement and difference in signals from the big players. FED are hawkish / ECB dovish meaning that there is further room for accepting a negative drift.

The reason for starting this thread is because ... I am not sure I have proven that my assumptions are correct... I can't see anything that supports the pre supposition of non randomness, so far I have not been convinced ... Although Mastermind's paper is very interesting and I am reading it as we speak.
AVT INVENIAM VIAM AVT FACIAM
  • #120
  • Nov 1, 2015 8:24pm Nov 1, 2015 8:24pm
Quoting the redlion
Good evening my fellow traders, I have been a student of the market since 2011 when I joined FF. The entire concept of investing, and obtaining profit through fluctuations in price movement has fascinated me. I have not been fortunate enough to blow accounts as many of you have, I enjoyed relative success in the markets earlier on which fueled my hubris. I erroneously surmised that observing charts, learning patterns, using some sort of money management system and discipline was responsible for my success. I have studied Technical Analysis tools...
the redlion,

A couple of quick points:
1) It isn't possible to learn enough just from comparing distributions (cdfs) to determine anything with regards to random/nonrandom.
2) You highlighted the obvious difference in kurtosis between the price return data and artificial data but didn't link this to trends in the market data that aren't present in the randomly generated data. Fat tails are evidence of trends.
3) When calculating acf on the whole period you are making an implicit assumption: that correlations are constant/persistent through time. What if this isn't true?

I came up with the hypothesis that autocorrelations are not persistent through time. So I tested a moving window of lag 1-10 acfs. By chance 5% of the values could be expected to exceed the 95% confidence interval. The results below show how often the 95% confidence interval is breached when walking forward one bar at a time, recalculating acf over the moving window only. From this it appears that lag 1 may be significant and lags 5-6 may possibly be significant for this data set. Strictly speaking, 8 of 10 exceeded the 5%.

Inserted Code
[1] 0.15748425 0.06159384 0.07449255 0.05989401 0.09079092 0.09459054 0.07079292
 [8] 0.06349365 0.04409559 0.03949605
Attached Image (click to enlarge)
Click to Enlarge Name: ACF_moving_window.png Size: 11 KB

The confidence interval is +/- 2/sqrt(n) where n is the moving window size. I repeated this with some different window sizes and lag 1 seemed to have the highest % of excursions beyond the 95% confidence interval of those that I tried.
0 traders viewing now