I am working strictly from Kaufman's description of the TScreen.
Kaufman?
Currently just looking for buys. Putting in the $#@! force index was a chore; elder specifies a 2 day exponential MA of the force index, which of course theoretically requires an infinite number of bars, so I looped in QP2 back some number of days (20 as I recall) to get a decent approximation. Then, it's looking for THAT to cross 0, so I have to remember the penultimate value and see if it crossed above/below 0 "today".
Having done this, I just spew the tickers out to a file which then in the batch file gets catenated to my list of "weekly" signals. My point there is that the signal is valid from when it hits till the end of the week, due to the trailing stop buy-in point moving down with the stock until it hits or expires.
I then have a scan which calculates the buy in point (high(-1) + 1/16), initial stop loss (min(0, -1, lo) - 1/16), point the stock has to hit where I move my initial stop to the breakeven point (average_true_range(5) + entry point), and profit target (6 * stdev() + entry point).
As I enter positions, I delete those tickers from my weekly list. The weekly list also gets removed, well, weekly, when my weekly TripleScanOne scan gets run to give me my new batch of candidates.
As for stops, after the breakeven stop is hit, I set my stop loss 1/2 way between the entry price and the highest close since position entry. (I do this manually; i don't have a lot of positions, so it takes but a minute to eyeball it and put my orders at the end of the day.)
I have the same thing for shorts, but I don't have the capital (or the emotional fortitude, yet), to short.
I just ordered a book from www.dynamictraders.com; read about it from a guy on another investment mailing list. I know nothing about Elliot analysis, and he says this book is the best he's seen about it; no psycho-babble, not trying to fit EVERY chart into an elliot wave. We'll see. Any interest in that area? |
