I'm wondering if it can give you a level or Zone to work off of. Just looking at old recent trades sometimes it appears if you draw a horizontal zone or box where pi bounces its a significant level where price taps back to or close to then moves away from.
Something to consider if your 1st trade is a loser. Watch and see if price comes back thru (the zone) and taps back to that area then watch for declining Total Volume when it does. If volume declines as it comes back it might be a good 2nd Entry.
Unfortunately, I have to report that when I extended the back test date range to 365 days (one year), the results were pretty mediocre. Oh well, I have less than a month to figure out if this indicator has any merit but merely following the pi bounce rule blindly without any other filters, does not seem to work over a longer back test time frame.
In any case, I will continue to experiment with other data series times such as tick, range, renko, etc. Will report back here if I find anything of any significance.
Till then, blue skies,
Pilot Trader
The following 3 users say Thank You to PilotTrader for this post:
I have found the same thing. Even though the general area of the swing turn is accurate most days, you need a filter like the U11trend indi, or some kind of moving average to filter that the big money is indeed turning around.
New Indicator That Actually Works with filters
Filter is second indicator.
Lets add another one to make it work even better.
Lets ignore the strong pull and bear trends.
Does PI-Oscillator actually works ?
If I look the charts to me it seems that PI-Oscillator wants to catch bottoms and tops.
Does bottom fishing and top picking work ?
Yes, tried this on ES, NQ, YM, GC, etc. The best results I was able to obtain was for CL, which I shared originally. But long term testing, so far, have yielded lackluster results.
This indicator seems to work well on tick charts! Data series optimized @ 800-tick for the ES with fixed 55-tick target & 30-tick stop. Here are the performance summary results I got:
Blue skies,
PilotTrader
The following user says Thank You to PilotTrader for this post:
Don't get too happy. Always start with the premise (i.e., null hypothesis) that you have nothing more than a curve-fit. 90%-99% of the time that's what it will be. Probably pick up @kevinkdog book for the methodology.
The following 3 users say Thank You to Sazon for this post:
What if you use bigger long target and a smaller short target on the ES testing, as the results may be better due to a rising market? Would that make any difference, especially if willing to hold overnight for up to a week?
The following user says Thank You to Krav69 for this post:
