Quote from darktrader:
Hello tomahawk!
many thanks to your reply, very apreciated, I replay to your post a little quick, because I'm very tired at this hour of day (I'm in europe), and I'm about to go home.
You are totally right, but it is impossible to test my strategy, this because I manually draw my trendlines, the tradestation automatic trendline indicator doesn't work for the kind of trendline that I draw (see picture attached), I have in mind to ask to a programmer if it is possible to program an indicator that could automatically draw for me something that could be similar, this way I can backtest and also even if I want to trade it discretionary I can have an idea of where to take profits and where to place stops, etc.. Also drawing all this charts is a very big work...
it is very stressfull to work this way, with all the pressure I have from my personal life, with a strategy that you don't know if it will work, you don't know when is best to take profit, you don't know which is the best distance for the stops, etc.. I only know that many times the markets respect my lines, and that often they are the low or the high of the day.
thanks again
No problem. And yes, I understand. Then it might not be a bad idea to forward test in sim for a couple months, and even if you don't take all the trades at least keep a detailed spreadsheet on entries, exits, MFE and MAE to at least have some data to fall back on.
The other thing I might suggest is looking at a shorter time frame for your trendlines if this is for daytrading. I look at TL's on the hourly charts as the maximum time frame for daytrades. And they do seem to work quite often, at least on the indices.
But bottom line, whatever your strategy is don't neglect the hard data. I can't tell you many times I've perused a bunch of charts and noticed what seemed like an incredibly consistent pattern ("many times the market respects my lines") which actually ended up being a coin toss (or failed miserably) in any kind of test. You can still make things like that work with position sizing and management, but it does help to be armed with some data.

Again, good luck.