Why do I see "Trends" in Randomly Generated Data?
- Thread starter Rahula
- Start date
A high school Alegbra teacher of mine did the same gimmick to prove 1=2 then showed us a-b=0 thus rendering it a bogus proof lol.
John
laying any type of contruct over a quasi-random ( from the traders perspective) system does not create order but rather the illusion of order which is highly dangerous in this game.
regards,
surf
Your buddy's new namesake Benoit Mandelbrot has written a complete science you should read about called, "fractal geometry".
Mandelbrot found that price changes in financial markets did not follow a Gaussian distribution, but rather other Levy stable distributions, having theoretically infinite variance. He found, for example, that cotton prices followed a Levy stable distribution with parameter à equal to 1.7, rather than 2 as in a Gaussian distribution. "Stable" distributions have the property that the sum of many instances of a random variable follows the same distribution but with a larger scale parameter.
One doesn't need to read the books or articles to see this work just research the process and lay the environment out in your charts to prove it to yourself. Of course this assumes you are able to perform this task and can first do the research.
In a nutshell . . . fractals bring order to random chaos.
It never ceases to amaze me how many posters use mandelbrots name as some type of evidence that random walk proponents must be wrong because even the mighty mandelbrot showed that fractals are self-similar and chaotic, therefore, there is some underlying order to the markets that advanced mystics can see.
If anything, mandelbrot was saying markets are worse than gaussian random, attacking gausssian modelers because their view had too much order!
If you want to know what mandelbrot actually said about TA, he quoted that "patterns are the fool's gold of the financial markets. .. They are the inevitable consequence of the human need to find patterns in the patternless." I'm sure he also meant that towards looking for visual correlations in fractals on different time scales. Just because they are self similar does not mean that one instance predicts the behavior of another in the future. Mandelbrot is definitely far more on the side of the unpredictable camp than the visual palm readers.
If you anti random walk posters want to use a name to back up the non-random walk, potential predictability arguments, use A. Lo and variance scaling ratio tests, not mandelbrot.
I think it's pretty clear that most of the posters on this thread understand, that when we say the markets are random vs. deterministic, we are not limiting the definition of random to a gaussian distribution. The uniform coin toss is a very simple model that drives the visual point home for those who do not understand or are not yet aware of different types of random modeling.
A beautiful, crystal clear explanation! A very well articulated and explicit example of the deep understanding! I enjoyed it! Too many people are too complacent about their grasp of processes that occur in the market place. Too many illusions and perceived patterns! The most important and useful pattern that could be found in the historical data is the type of the distribution this data follows.