PPoS Improvement Ideas - Brainstorming

Post by **Reaper** » Sat Mar 07, 2015 3:13 pm

I want to collect any ideas, features to add, improvements, anything that you can think of that could potentially improve the profitability of this EA. Even if it's not tested or fully thought through, just share it. It might spark ideas in other people. Think of this thread as a whiteboard for brainstorming...

For example a number of us came to the conclusion of adding ATR functionality which Steve kindly implemented for us. Now I can't tell you 100% this is better than the fixed pip values we had before but my gut says it is, especially when running 20+ different pairs. It makes sense that it is, now my question to you is what else makes sense? What else can potentially be done to reduce losses, increase profits, improve efficiency etc..

Have you noticed any correlations, any losses in one specific pair due to a particular reason, can that reason be avoided by using a certain indicator or filter? Is there anything you might have figured out that could help?

Feel free to share your thoughts.

____________________________________________________________________________________________

PPoS Experimental Version updates history

V1b, posted 10th March 2015:
- separated Breakeven, Jumping and Trailing stops into separate Main and Multi-trade features. The Main trade inputs begin with 'Main' and the Multi-trade inputs with 'Multi'.
- added ADR multipliers for Main and Multi Breakeven, Jumping and Trailing stops - inputs in the ADR for sl/tp section.
V1a, posted 8th March 2015:
- New input in the Multi-level trading section: MultiTradeTakeProfits. This is a comma-delimeted string that PP converts into numerical values and uses these for multi-level take profits. Leaving this blank tells PP to use MultiTakeProfitPips instead.
- Make sure the number of tp's in MultiTradeTakeProfits = MultiTradesMax or the program will crash. There is no idiot checking.
V1, posted 8th March 2015:
- added these three inputs underneath their corresponding time frame inputs: MinimumLongerMovementPips; MinimumMediumMovementPips; MinimumTradingMovementPips
- PPoS will not send a trade if the relevant candle size is less than the value of the input - so a zero value turns off the filter.
- the is no update to the screen feedback.
- coders, the filters are checked inside bool IsTradingAllowed()

Post by **Reaper** » Sat Mar 07, 2015 3:13 pm

So I've been playing around with Monthly, Weekly, Daily Minimum Pip Heights. (e.g. Monthly candle has to be 80+pips before trading is allowed) and I'm still trying it out, so nothing to report as of yet, but that is definitely an idea we can implement.

I know there was a discussion about people using different brokers, receiving different candles and therefore different trades. Now my thought was if we implemented +50 pip height in each TF, that would potentially rule out any doji-like candles in the MN, W1, D1 TFs which might be giving us more losers than winners.

Its an idea that I haven't tested fully.

(+50pips above is just an example. I'm trying Mn 80+ pips in one of my forward demos. I just have set up another demo without this as a control.)

These minimum pip heights could also be ATR adjusted, to get away from fixed pip heights across multiple pairs. Just food for thought.

Post by **SteveHopwood** » Sat Mar 07, 2015 6:43 pm

Great thread Angat - I have added reference to it to my wrup tomorrow night.

Love the new avatar.

I too have thought about minimum candle movements. I will create a new version of PPoS tomorrow that we can have as an experimental version to play around with new ideas, and upload it to your post 1 if that is ok with you. PPoS is not complex like the original HGB was, so an experimental version will be much harder to break.

Post by **Reaper** » Sat Mar 07, 2015 7:23 pm

SteveHopwood » Sat Mar 07, 2015 6:43 pm wrote:Great thread Angat - I have added reference to it to my wrup tomorrow night.

Love the new avatar.

I too have thought about minimum candle movements. I will create a new version of PPoS tomorrow that we can have as an experimental version to play around with new ideas, and upload it to your post 1 if that is ok with you. PPoS is not complex like the original HGB was, so an experimental version will be much harder to break.

Fine by me.

Might be an idea to sticky the original PP and PPoS threads just so people know those are the two main ones.

If anyone else has any other ideas, feel free to post! :hi:

JRforex · Post by **JRforex** » Sat Mar 07, 2015 8:07 pm

Hi people,

I have an idea to further improve Piproid (I just prefer that name

). Looking at a few forward demos, I saw that when PP catches a good trend, it is not difficult that the pair moves 150 pips or more in our favor. Currently, Piproid is opening and closing multi trades with a TP of 30 pips, but if we know that the price is going to move at least 150 pips, maybe we could modify the TP of the multi trades as well, so that we do not limit the potential for profit, and all the trades aim to catch that 150+ move.

With that in mind, I modified Gringo's EA so that one could select the TP of each of the multi trades. I only use 5 multi trades, and I have to take into account that Piproid is going to miss the 100 level, so I had to adjust the TP of the multi trades as well. The TP I currently use for the multi trades are the following:

1st trade: 130 pips,
2nd trade: 110 pips,
3rd trade: 90 pips,
4th trade: 70 pips,
5th trade: 50 pips.

Bear in mind that if price retraces before that 150 pip move, you could sustain a loss of up to 600 pips; that's why you should use more strict money management rules for those multi trades. Currently, I manage the main trade with xfactorno's settings, but I use 50 BE pips, 5 BE profit, and a JS of 50 pips for the multi trades.

Last week I made 2400 pips with this method, on a good week, however, you should make 5000+ pips. The myfxbook link is the following: http://www.myfxbook.com/members/JRforex ... gs/1178955

Regards,
JR

Post by **plateman** » Sun Mar 08, 2015 10:13 am

Just an idea, how about using the last four complete weeks for the month values instead of the last calendar month. The thought behind this is to use values of a consistent age, currently the oldest values being used range from approximately 4 weeks to approximately 8 weeks.

Post by **SteveHopwood** » Sun Mar 08, 2015 2:53 pm

I have added an 'experimental' version to post 1. Details at the bottom of Angat's first post.

Post by **SteveHopwood** » Sun Mar 08, 2015 4:47 pm

Experimental version 1a is in post 1, incorporating JR's multi-take profits for the multi-trades. I have not tested the code, so I may have forgotten or mis-typed something.

Post by **wolfeman** » Sun Mar 08, 2015 7:04 pm

plateman » Sun Mar 08, 2015 2:13 am wrote:Just an idea, how about using the last four complete weeks for the month values instead of the last calendar month. The thought behind this is to use values of a consistent age, currently the oldest values being used range from approximately 4 weeks to approximately 8 weeks.

That sounds like a very good idea...

JRforex · Post by **JRforex** » Sun Mar 08, 2015 9:56 pm

SteveHopwood » Sun Mar 08, 2015 10:47 am wrote:Experimental version 1a is in post 1, incorporating JR's multi-take profits for the multi-trades. I have not tested the code, so I may have forgotten or mis-typed something.

Wow Steve, that was pretty fast. Thanks!