Note once again, that the while the distribution of this chart in a histogram display would mimic a gaussian curve (as maestro showed), the only difference separating such a curve from real markets, would be the spurious fat tails or gaps. You could extend this to bar or candlestick type charts by simply randomizing o-h-l-cl data and plotting, although this is simply one random point per step (-1, o or +1) added to the running total.
BTW, those green and blue lines are the RSI tops and bottoms. I was testing someone's new formula for RSI oscillator. Worked pretty good. Till it blew up and lost everything it made. Which just proved my point.
Those could be modeled as well. Nice applet, I had one of those as well; could not find it. As I said, it is very rare to find someone here who really understands the subject.
Thank you. I cannot verify your work, but I have no reason to doubt it. Interesting and thought-provoking.
No doubt the result of poor money management. Note that, quite coincidentally, buying into the lower bottoms of even your randomly generated price chart may build character but not account equity. I can't even begin to explain the rationale on a random price series or even pretend that there is one, but there you have it. (?!)
I think the problem that most people have with visualizing random price action, is that for whatever reason (schooling, past experience) they are familiar with visualizing random price data in the following display format. This format simply displays each individual randomly generated step over time. You can also see some streaks of the same data generated repeatedly (streaks, trends) in this type of display. However, it is accumulating each step over time (taking the prior sum and adding each step) that gives the type of display that mirrors markets. It is the same as taking market data and looking at each days swing, then plotting that out per step, rather than the cumulative output most are used to looking at. You see it more clearly once you've performed that exercise many times. As I've said before, they should make TA practitioners backtest systems vs. random data as the ultimate sign of robustness. That would be the optimal system (and many like thorp, black scholes etc.. have passed this test). Unfortunately, it's the fat tail black swans (the events that completely defy predicted probabilistic expectations based on gaussian modeling) that kill such systems at some point. Which is why money mgmt is another important step that must be adhered to. BTW. one way to visualize an unexpected fat tail event on the chart above, is to simply visualize one and only one data point jumping to minus 20. So while your system may behave well under each data pt. up to that event (since you were prepared for a Gaussian type distribution of random data), that one spurious outlier could kill your account (esp. on leverage). The only thing that would have saved you was money management.