Wednesday, February 17, 2010

PCLN is a good example of how a Trend Model can be imperfect, but very profitable.  Below is the PCLN Weekly Trend Model with five signals.  To avoid entry/exit nuances, the following prices are from the OPEN on the Monday after the Weekly signals were generated.

Sell   7/7/08    @ 119.78
Buy  1/5/09    @   76.07
Sell 11/2/09    @ 157.51
Buy 11/16/09  @ 203.54
Sell  1/25/10   @  205.41
Current Price =   210.12

PCLN Weekly Trend Model

Trading the Vortex Indicator - Stand Alone

Sell 7/7/08 @ 106.30
Buy 1/12/09 @ 74.00
Current = 210.12

Using Vortex Indicator to Filter Trend Model Signals

You're kidding, right?


Anonymous said...

Allan you really need to show charts other then strongly trending charts to show how your system can handle those situations. All of these charts in strong trends could be traded with simple moving average crossovers which would obtain similar results which you are showing with your system.


Allan said...


An intriguing idea, but when put to the test, moving averages pale in comparison and they really chop you up in anything but strong trends. The moving average idea captures some trends, but far less then what my trend models are capturing. I do tend to post the most illustrative examples and there is a method to my madness, that is to demonstrate how well these trend models do work so much of the time. A chart full of whipsaws will serve what purpose? to show it happens? Yes, it does, but the profitable trends are so much more prevalent, why go there?

Rob said...


I've traded a similar system using DM +/- (8 period) crossovers as buy/sell signals with a PSAR (.02/.20) trigger as confirmation.

Do you use the Vortex indicator in your model as the bottom line filter, ie. no buy/sell signals are taken UNTIL this indicator confirms the weekly/daily ATR/volatility stop signals?

Thanks for all you do.