Speaking of Fibonnacci, i read a very thought provoking article about them in relation with the content of this thread. I think some of you might like to read it by Lance Beggs. The author sees no value in them as an analysis method. Fibonacci makes no sense when considered from the position of supply/demand. And there is no evidence of these levels providing any greater edge than when using a random selection of levels.
As with most articles there is some truth and some nonsense. The world is not black and white, but gray.
(1) Any lines such as random lines, fib lines or floor pivots may look more impressive after the fact, even if in some cases they match price action and in other cases they don't. The brain will focus on the matching points this is called confirmation bias.
(2) May I cite: "The real reason that price turned at this point, and the ONLY reason, is because of a shift in the supply / demand dynamics of the market.". I agree with this statement.
(1) May I cite again: "It wasn't because of the Fibonacci level. It was supply / demand. Everything else is illusion." I agree with supply/ demand. However, this statement ignores the human tendency to herd. Many traders rely on the same - arbitrary - tools to determine turning points. As they all use their pseudo-scientific tools to establish support and resistance, the tools start working. This is self-fulfilling prophecy at work. Ignoring this behavior and its impact on the markets would be foolish.
(2) "50% is not Fibonacci ratio." Correct, but traders look at these levels. As it is self-fulfilling prophecy and not science which is behind all concepts of S/R, any level used by a sufficiently large number of traders will work.
(3) "So, why not just forget the illusion and use real S/R - areas of previously proven supply/demand imbalance?"
Not a bad idea. Areas of previously proven supply demand imbalances, I would call them empty zones, will typically show in the middle of a swing, so you can use the 50% and 61.8% Fibonacci levels to detect them.
The guy tries to explain that from a mathematical point of view, Fibonacci levels are nonsense. I share these views. But from a behavioral point of view, they work very well, if enough traders follow the fad. The point is that trading is no science. It is an evolutionary cosmos, which can be compared to a game with ever changing rules. In such a cosmos, all types of species, even fibonaccisaurs can survive.
The following 12 users say Thank You to Fat Tails for this post:
I see a larger meaning to Mikes experiment with Random lines, COULD TRO HAVE BEEN RIGHT? When he said " The markets are random; prove it to yourself, draw a line anywhere on the chart and eventually, PRICE WILL TOUCH IT.
I have skimmed through the tread and forgive me if the perspective I am about to bring up have been brought up before.
As an experiment to illustrate how our brain is hardwired to follow the golden mean 0.618 Pytagoras gave his students a line where he asked them to set a mark where they pleased, but not in the middle.
Lets say a group of hundred people would do this, then not everyone would set the mark directly on the 0.618, but if one took the mean of all hundred the result would pretty accurately become 0.618. Another example is the golden mean when it comes to pictures. As known, what we perceive as a pleasing picture is where the center of gravity is located at 0.618 or 0.382 both vertically and horizontally. And even better if all the components of the picture are also in a Fibonacci relationship to the whole. If not so, the brain will conceive the picture more or less unpleasant to watch. So naturally the brain is drawn towards the golden mean.
Then as an illustration, when taking the speed of the brain in consideration they came to the conclusion from recognition tests for pilots during the Second World War that we are able to recognize about 50 pictures in one second. Which also explains why speed readers can read more than 3000 words per minute. In fact, the best speed readers do 6000 words per minute. Which makes a 100 per second. With the speed of the brain in mind, my question becomes; When we try to set lines randomly are we able to recognize the golden mean mechanism at work? I mean when having set the first line, are we able to avoid setting the rest of the lines in Fibonacci relationships to the first when we have no other reference points?
As a test for myself I set up a MT4 chart where I took the chart as far as I could to the left trying as best I could not having focus on the instrument at all. I then took a piece of paper and covered the instrument. Which was the EURUSD by the way. I then sat the lines and then added a Fibonacci set up onto the lines.
What you are about to see in the pictures is no BS. The result baffled me. There is no cheating here at all. This is the first chart of the day, because my office floor has been painted with "heavy duty" floor oil paint so I try to avoid that as much as possible to avoid frying my brain. My question to myself is how well or how could, and so on, my brain know where to set the first line which clearly influenced where the others was set. One could say that I have been working so much with the instrument that I got a very good feeling for where the levels are in the first place, which I would be very happy with off course, but this would disqualify the experiment. To some part I have to believe that my subconscious have had to have a finger in this regarding at least one line, but I do not believe that it can explain it all. Even if I "threw" out the lines randomly the leaning towards symmetry is clear. Having relativity in mind please take a look.
The first one is the blank chart with lines, the second with Fibonacci added (removed the redundant Fibo lines), and the third is where I pulled over the chart showing lines with historical data.
“If you wish to see the truth, then hold no opinions for or against anything.” - Hsin Hsin Ming
The following 4 users say Thank You to Laurus12 for this post:
The random line theory is about pattern recognition failure and hindsight bias.
However, in practice some lines do work, not on any scienific grounds but because people use them. Let us have a look at an example for YM. It just retraced to the 61.8% Fibonacci level and then moved back up. This was not astnoishing.
The fib level was exactly aligned to the high of the night session (prior to the morning news release). See chart 1 attached. But now you may ask the question, why this was exactly aligned. The two consecutive new highs were themselves Fibonacci extensions. See chart 2 attached.
This is a pseudo-science, but it works. Not all is random.
The following 4 users say Thank You to Fat Tails for this post:
FUNNY AS HELL...YOU GET ENOUGH LINES ON YOUR CHART AND ONE OF THEM IS BOUND TO GET HIT!!!!!!!!! All kidding aside....FATTAILS, I LOVE YOUR WORK...YOU ARE THE BEST!!!!!! THANK YOU FOR EVERYTHING YOU HAVE GIVEN ME....
Please register on futures.io to view futures trading content such as post attachment(s), image(s), and screenshot(s).
Last edited by JohnnyCash; May 28th, 2011 at 08:40 PM.
Reason: ADD PIC
The following user says Thank You to JohnnyCash for this post:
I just discovered this thread - and I love it. This is a great experiment that we should all do for ourselves to place the ideas we have about the significance of horizontal lines that are derived from price (as opposed to coming directly from price such as yesterday's high and low) into the proper context.