It is understood by scientists now that price movements follow a statistical distribution called "truncated Levy flight." See for example this article, http://ideas.repec.org/p/sfi/sfiwpa/9705087.html Truncated Levy flight is a special kind of random walk, with variable step lengths. This is quite different from the usual random walk which produces the Gaussian distribution. In truncated Levy flight, there is no intrinsic length scale (time scale or price scale), so that price patterns over minutes would look very much the same as price patterns over days or weeks. This "self-similarity" is what EW theory tries to capture. However, although it is not regular random walk, it is random walk nonetheless. Then how can one make money from it? This is a little complicated. Basically there is information flow. It was shown by scientists that the stock market follows essentially the same scaling laws as turbulent flows (such as air bouncing off airplane wingtips). In turbulent flows the scaling laws are determined by energy dissipation, from the large length scale (meters) to the smallest length scale (micrometers, or even atoms). In a similar manner, the scaling laws in the stock market is determined by information flow between different time scales. Because there is always a time delay for information on the large time scale (a fund entering the market) to propagate to the small time scale (the scalper flipping the trade), if one intercepts that information flow (by whatever means) one can make money. I would be surprised that EW can do that job. I think the trader's brain is doing that job, but because he has to think in EW language, he believes it's EW that's working. BTW, there is no "market psychology" at all. All data points fall nicely on the truncated Levy flight curve.
Sadly, insight eludes you. That doesn't bode well for your trading but I'll wish you good luck with it anyway.
"kiwi_trader Registered: Aug 2002 Posts: 1635 02-07-07 01:14 AM Qqq, you're still shouting. You are a rude bastard. Why don't you just fuck off and mope somewhere?" And YOU are calling me quite the potty mouth? What does that make your mouth? The town sewer? I mean my gosh Kiwi, Hypocrite must be your middle name. You're one pretentious and arrogant fellow. Take your luck and shove it where the sun don't shine.
Reminds me of Mandelbrot. Have you looked into his findings? He says that the power law distribution is the one that most closely fits the price data. Taleb is also big on this. The Levy distribution is part of this family of distributions?
There had been many tests. The earliest research was either 1950's or 60's using some obscure agricultural market data. The most authoritative study was done in 1996 and used foreign exchange data over several years, analysing every tick. Here is the article: http://www.nature.com/nature/journal/v381/n6585/abs/381767a0.html "Turbulent cascades in foreign exchange markets" There were other studies using data from various local stock markets. They all came to the same scaling behavior. In fact, you can download daily closes of DJIA for the past 70 years, and do the analysis yourself. It's simple enough. It's true.
You're right. I remember now. The first work was done by Mandelbrot in 1963 using cotton prices. Here is a webpage describing his work: http://steveedney.wordpress.com/2006/09/26/misbehavior-of-markets-mandelbrot/