This is just a hunch, but I think it was because he posted referral links. You could tell by the way the link included "partner.phtml....=GCT" meaning he might have been paid for anyone who clicked the link. If my hunch is correct, I don't see why the mods don't just remove the links and give him a warning. A ban seems a bit excessive.. He's obviously been very generous & helpful with all of the info and support he's shown, FF needs more members like him!
****
Now onto GCT's interesting system, I setup a demo on IG to test it out. All settings were entered exactly as described in his original writeup. Since someone pointed out earlier the bug in the coding for 4-digit decimal pairs, I only ran the backtests on yen crosses (E/J, G/J, U/J, N/J, A/J, CAD/J). They all came out positive except for aud & cad, although in retrospect, I realize that the bollinger bandwidth parameter is tuned for EUR/JPY and should be configured differently for each pair.
Even with accounting for the lack of fine tuning by pair, the out performance of EUR/JPY was eye opening. 11 trades, all winners, resulting in a near doubling of the account in 3 months (with no compounding!). The 1H time frame is by far the most optimum, dropping down to 30M cut the gain in half, 10-15M were breakeven, anything below that resulted in a loss over the period. The average time between trades was 127 trading hours (~5 1/2 trading days). Still, for 100% accuracy over a 3-month period, that's well worth the wait. Once I get the other pairs fine-tuned with regards to bandwidth settings, I may have to demo PRT myself to see how the backtests stand up using several years worth of 1H data.
Thanks again GCT for sharing your system. I hope your situation can be rectified amicably so that you may join us here once again! (and I want to hear more about that brunette)
