Pivot Analysis

[mobthehop]

lifesys, was looking at this as well, but my programming skills are not there yet - would very much appreciate if you could share your EA for educational purposes....

And Jedster - again - thanks for sharing your work!

[Jedster]

Well done Jedster.
As mentioned in PM, my observations have been that, not only for ranging times (>70%), price reverts DO to DP far more times than away and it is NB's skillful management that gets him the pips.
'Reversion to mean' is majority by definition.
I compiled an EA (not bullet proof) which scalps Abs(DP - DO) when > (spread+slippage+MinTrade) for pairs each day at (real) D1 start.
Many small +ve trades net av. > 350pip/day, but challenge is protection against trends.

Success ratio >80% but R:R is poor, hence TP (or +ve close) av. must be > 1/4 SL (or -ve close) av.
Included TMA in reverse. No trade if daily open H1 TMASlope is >0.4 or all ADX-MTF same direction.
Not sure this or D1 iAtr() ratio SL helps much as Abs(DP -DO) size depends on larger move prev. day (or last hour) but then usually comes a lull & vice versa.
If consistent, net av. > 100pip/day then workable.

Was really interested in Jedster's indi. to run through and back test above on numerous pairs.

Paul

It's interesting analysis. Obviously what we what to do is find a statistical probability that is tradable. Once I saw these numbers, I was already wondering how different would they be when the market was ranging or trending. For the hourly analysis, I had already put in an ADX filter to see if it made much difference. So I have the structure in place to add some filters like this relatively quickly.

I was also thinking to look at where the daily pivot is compared with the weekly pivot. These are all possible things to add, but, will they provide a tradable edge. I guess we won't know until it is done. When I get some more time (probably next week), I'll add these changes.

I also want to write an EA to back test some of the theories. So for example, initially trading in the direction of the pivot, and then opening a second trade once the pivot has been hit, back in the direction of the DO. Plus, how does it change when looking at instruments that traditionally range, or traditionally trend. Also, what about more volatile instruments like some of the commodities. All things to explore, so lots of work for next week...

[falsedave]

Wow, This looks very, very interesting Jedster really nice work. To show Pivot points and probability that they will be hit would make a killer indicator.

edit : sorry just got it working properly and it looks amazing

[Ben12]

Hi Jedster,

Thanks - this is really interesting. Looking at some of the higher probability hits of PP and R1, it looks like:

1. DO>PP generally has a higher probability of hitting PP/ R1 than DO<PP
2. The correlated (ranging) pairs obviously have a high probability of hitting PP/R1
3. Interesting that the major GBPUSD has 73% probability of hitting R1

Regards
Ben

[profuse]

Hi Jedster,

Promising Thread and Great Job

I downloaded the indi and ran it on EU. After some searching I found the CSV file.
But,.. I'm not able to read it comfortably though, as it opens in Excel without the layout.
Same with the CSV example as provided in this thread.

How can I get it in the same layout....?

OK, FOUND IT! in "Text To Columns" choose "comma"



Regards,

another Ben

[Steven Smith]

I think you should use Fibonacci Retracement indicator, because this indicator create pivot
points itself.

look below