I could backtest, but the real problem is that the entries really are discretionary. I only would take them if there are larger factors signaling a trade, like a recent major trendline break or something. I will def check out that site. I don't supposte you went to Central WA U did you?
That can all be included in the system and backtested. If you don't have the capability to backtest, post your specific entry rules, and I'll backtest with a few different exits. Some that can work an EOD system: - Time Exit - After 2 or 3 days - Percentage Gain - 3-4% - Technical Indicator - Price crosses back above MA, Rawk reaches target level, etc, etc. Or, a combination of all of the above.
Here is why i don't think it can be backtested: to pick a stock i scan for the conditions i talked about in the first post, and then i sort by highest ADX. I then look through until I find a stock that meets my criteria. My criteria are something like a recent double bottom followed by breakout, or a real nice 3 month trend we are hoping to get a piece of. I'm not trying to create an automated system, hope I didn't give that impression. Here is one set of exit criteria I have been playing with. 1. Initial Stop Loss: stock must close below the 18SMA value the day the signal was given, which would be the day before entry. If you could test this vs a real tight stop right below the most recent low that would be cool. I've looked at a couple hundred charts and a looser stop seems to work better. 2. If stock moves in our favor the size of the stop, move it to breakeven. This would be an order we place to trigger at any time during the day. All other trades we wait for close to enter our trades. 3. Let the winners run, using a close below the 18SMA as a signal to exit at open the next day. Reverse it all for the shorts. PS a useful filter to find good stocks is to only look at stocks where 50EMA is greater than 200EMA for longs, reverse for shorts. Thanks for offering to give this a try! Hope i explained things OK.
Oh yeah I forgot here is another thing. If the stock is still above the stop loss and a slight loser or winner, I will often get out after a few days to a week, especially if the technincal indicators are starting to turn. If it is like this after 2 weeks, definitely get out.
All of this can be coded and backtested. But if your preference is to look through hundreds of charts and take a guess, who am I to take away your entertainment? Good luck.
Hehe I thought you were offering. I can do some basic backtesting, but wouldn't have a clue how to make rules for only picking charts with nice healthy patterns. Thanks for your input though.
Lindq i've been playing with your suggestions regarding beta and debth of pullback, and it is looking pretty good. I really appreciate those hints.