Aurel, yes, your assumptions are correct. If you want to use randomness in your stock trading decision process it won't matter if the price movement itself is random or not. I look at it as quasi-random, random walk, kind of close enough, but not quite.

And this makes it a biased game. As a consequence, one should have a biased trading method that follow this biased trend. And not necessarily try to outguess the randomness part of price movements which, as you surmised, would be like trying to beat heads or tails, where any attempt at using “predictive” indicators would tend to fail not realizing that the times when it was right were just coincidental.

For those that say that the price movements of stocks are not random, then it implies that they are predictable. And if predictable, well, your automated trading system must be so productive (profit wise) that you don't even need to bother with randomness or even read about it.

I would add that if you know that your methods are predictive in nature, then why use words like stop loss, market risk, or uncertainty. A forward probability that is just a guess does not make it a hard probability, in the sense that it is true. A 50% probability of prices going up tomorrow has absolutely no trading value. Only luck will answer that problem. And a predictive 52% probability (guess) does not provide that great a short term edge either.

Aurel, yes, randomness would technically simplify the trading process to the price itself, you don't need technical indicators or predictive techniques. I prefer to look at prices as Δp, not just price, but price variations. And what I am looking for is Δp > 0, since in itself it defines a profit.

As for the analysis you request, I don't think there is sufficient computing power on the planet to execute such a thing. And even if we could, when it would be time to put what was found in execution, we would be faced with what we have today looking forward: a blank page, the right edge of the chart, this quasi-random world. As if, even if you could find a solution, it would not be a solution for going forward, since you will still be faced with this quasi-randomness.

@Guy ..You know I am a forex trader ..I want to start learning. . "Binary Options" ..my next year target.. my resources regarding computations are very high .. I am running supercomputers... theoretical unlimited numbers. . or possible factors .. In this moment my trading system has 87 factors.. .(i already send you the report by email ) . and millions of combination.. :)..and is using "trend entry /exit" .. and I am not satisfy .. why ? because every 3-4 months need to be retrain ..

Aurel, I understand, but it does not change the perception. Using 87 factors gives you 2^87 possible combinations. That won't be analyzed in a faction of a second even on a super computer. Why is it you think that you have to retrain your system every few months?

I would have more confidence just in your forex trading judgment and experience.

But, nonetheless, without a bias, inherent or predictable, you are faced with a quasi-random game. That you use a method, or that you use randomness in your trade decision making on a quasi-random data series will give about the same results (long term). It's like using head or tail to play heads or tails. Even if you have a method on how to flip the coin, the outcome will be about the same.

Both systems will tend to converge to their long term expected mean. So your expectation would be about the same with, or without, a method. You need to extract something predictable from the process which technically will give you an edge, and as a consequence a method with positive expectancy.

Wait ..Guy ..sorry 84 factors..every one has values from 0-1000 .. in the beginning the numbers are random genetic algorithm ..not exhaustive iterations (this is impossible...with my 4X supercomputers ) ..using CUDA / GPU is easy . .is the only way we can use ..30.000 Core/16Xi7--> /200GBRAM .. you need like. . 10 days / 64mil combinations /min .. and than you need for every pair 2-3 h / every /3 months ..

..@Guy same factors have 4-19 values ..before definition.. you have 100..after same time you know that same value don't exist n the market ! ..like time/speed / volatility / size / regression/ HLOC..etc .