agree with limbo here. KISS - keep it simple, stupid. The simpler a system is, the more adaptable it is during different market conditions... and this has been proved time and again during my research of mechanical systems. I can't help laughing and doubt some people's sanity when I read some books/magazines with those fancy TA stuff that makes a chart looks like a plate of spaghetti. Here are a few examples: trendlines - total BS, ask 100 people draw trendlines on a chart, you get 100 variations.. Gann angles - total BS, some artificial spider net put on top of a chart. Dow thoery - 33%, 50% and 66% retracement? give me a break, why not say 1% 2% 3% .... 98% 99%, you have everything covered. There are many more examples like this, but the bottom line is, if you do some back testing, you will find most of the conventional TA stuff such as moving average crossing, stochastics line crossing etc are only marginally profitable (if trading a $25k account, maybe earn you minimum wage, that is if you can stick to the system) and most of them have such big drawdowns, NOBODY can stick to it after a string of losers. So the bottom line is, as in my previous post, you need your own system, fully tested, and has good parameter to minimize drawdown, only then you have a tradable system and can survive.