|
|
PRECISION INDEX OSCILLATOR STOP PLACEMENT
<
Hi Roger,
Thanks. What would you recommend for stop placement? I notice in some of your examples, some prices will retrace prior to moving in the direction of PI.
Best regards
<
That is not a simple question to answer as there are many methods one can use to calculate stop loss distances.
Most importantly the correct stop loss distance used will depend largely on the time frame chosen for deployment of the Pi-Osc indicator.
Stage 1. So, first of all the logical approach would be to determine if you are going to use a weekly, daily, hourly, 500 tick or some other type of chart to use Precision Index Oscillator with. The method of determination would be chosen by examination of its past performance and accuracy of the signals on each time frame chosen. Once this is done you can begin to look to stage 2 and go deeper.
Stage 2. Say for example you decide 180 minute charts produce the best signals, then you can eyeball through each signal and jot down the number of points that allows the good signals to make a retest of the lows ( as sometimes happens ) and still run up to a reasonable profit when using a trailing stop. You can see clearer by drawing some trend lines of how you would expect the trailing stop to operate basis of your percentage, points or other fractalizable factors. ( I am not sure if this word is real or not, but the meaning of fractal is a repeating pattern such found in lungs, broccoli, or snowflakes and of course charts, because a weekly chart with the scales removed can look identical to a 5 minute chart with the scales removed )
For example you can take the last 100 bars of data and divide by 100 to form a fractal stop distance and then experiment with multiples of that, you may find that 5 times the factor of this 1% is a good starting point. You could also use average true range x ( factor ) to compute the stop.
At this point you can start to notice the level of expectation of a losing trade, so can see how far the price can go to a new low before it is best to cut your loss and wait for the next signal.
The purpose of using a fractalizable factor computed by such things as range or atr's is that you can translate it from 20 minute charts to weekly charts and still end up with a reasonably accurate starting point. Which cannot be done with using percent or points as the change from market to market and time frame to time frame will make it unusable. Also changes in volatility can affect stop distances a lot. ( These are also taken care of by using fractalizable methods ) Parabolic SAR is another useful indicator for trailing stops.
3. Once you have a fair estimation of the above you can begin testing the historical data on the markets that you trade. At this point you can open your mind to the adventures of thinking backwards. EG If you have always traded the S-P500 futures or CL futures, it is a good idea to plot Pi-Osc on many markets with range bound, choppy qualities and likely you may find some unknown or unpopular market which works very well in combination with it.
4. Another important consideration to factor in is that with most technical analysis and trading systems, one can safely say that buy signals are much more reliable than sell signals as the emotional responses causing buying are totally different to those causing sell offs. EG, Buying is related to greed and the need to make money. Selling ( generally ) is causes by fear of losing and as many trade too big for their account and take positions bigger than is safe for them, these can come very quickly out of the blue. Buy signals can often be seen coming from a mile away.
Hope this answers your question
