Thanks KP for your posting! I do know where the lines eminated from. Glad you brought it up. If you look at those lines, you'll note that they are also arbitrary. The upper line is drawn above a peak that occurred 9 months hence from the beginning for starters. Diagonal trend lines are certainly better than horizontal scribbles but the issue here is that the student is forecasting both directions. I believe he is suggesting that buying a straddle at the median line is the best option here.
Well technically, they aren't arbitrary if they are always drawn the same way. The same way being that you take two swing lows, connect them, and draw the upper parallel line across the highest point. If anything, its pretty remarkable that given how this channel was drawn so long ago, that price is still respecting it.
I would actually also disagree that diagonal lines are better than horizontal ones. The horizontal lines represent actual values that everyone can see, that aren't just a line made up like a diagonal one. As you say, diagonal lines can be drawn very many different ways, so this makes it arbitrary. But if you ask someone to draw a horizontal line under the previous day low, there is only one way to do this. Likewise, a line draw across the previous day high can only be done one way. So if you have these lines on your chart from the previous day, and price approaches the previous day high or low, it will either go above it, or below it (worst of all is when it forms a range right at that level). But point being that everyone can see a horizontal line and everyone has to draw it the same, and if the line is a previous day high or low, you can bet that lots of traders are looking at it.
Here is the thing that I think you have to agree with. You have a method that causes you to lose 20 or 30 points, and everyone calls you crazy. But when I see that your posted results show that you are up $75k for the year so far, then you must clearly know what you are doing. Your stop and target placements, even though they may be very large, are clearly well tested to be statistically profitable in the long run. Likewise, if a trader builds a system that takes advantage of horizontal and diagonal lines that is statistically positive in the long run, then fundamentally, this isn't that much different from what you are doing.
I do like lots of things you write, like when you said that too much emphasis is placed on being right but not enough on keeping losses small, letting profits run, and not needing a high win percentage.
http://www.elitetrader.com/et/index...w-a-straight-line.287955/page-67#post-4099117
But this can clearly be accomplished in many different ways, and since you're big on the math of all this, which is essential, if someone is to build a system that does have a statistical edge, what is wrong with it being different from yours if it still wins more than it losses in the end?
