Peaky on Steroids

[Gertje]

Do I remember the Swissy.... :youknow:
It was the main reason I was absent from SHF for over a year.....
It cost me my reasonable sized account, and a top-op to GP to pay for the overshoot it created.

I hear you loud and clear......

[billv]

The longer I trade this, the more i’m convinced that the movements are on purpose, due to people on the other side of the screen.

Hello Gertje
Of Course the movements are not random.

Anyway, I think I'm now getting somewhere with the POS.
I fixed the global closure issue but I believe there is an issue with my symbol closure subroutine.
I use this subroutine 3 times and I have different TP in each one.
I will need to check but I think the symbol trade counter includes pending orders.
Other than that, it now goes well.

I don't have reliable statistical info because I did many changes to it.
I wanted the 3 EA's to co-exist in the one account but I also wanted them to run their baskets independently.
At some stage EA1 was closing all open trades and the other 2 EA's were doing the same.
I run all 3 EA's on 4 timeframes 15M,1H,4H,1D and I have different buffer distance in each TF
10, 20,30 and 40 pips ( I will review these in about 1 week)

Since I use 4 timeframes, I use my function "minimum Distance Between all Trades" to reduce my exposure and DD. I decided to use more than 1 TF because I needed many trades.
(Basket closure cannot happen without many trades).
I still have the input "take profit at the 6th" enabled but I doubt that it will get much use.

By closing baskets, I realize that we will potentially miss out on some profits but we compensate by the frequent basket closures. As you can understand the EA can close many baskets daily.
The other advantage of the basket closure is that it clears the deck of misbehaving trades.

There is still work and tweaking to be done but I will get there.
I also plan to steal the dashboard from the SPB EA
I'm glad you have not given up on it
Have a nice day :hi:

[Gertje]

With the risk of getting annoying, I would like to share a picture of my PoS demo that runs since mid July 2017.
The balance growth is somewhat lineair, where I would expect a more curvey/ parabolic rise since the lotsize are set by the balance. During the course of the demo, I lowered the dollup several times to accomodate the expected DD. This explains the more linear track.

My opinion that this is the most profitable EA ever is getting stronger and stronger.

What caught my interest when looking at the D1 Tradechart that I use to follow performance [the red lines on the chart divide the months], is the fact that at the end of each month, DD seems at it's high, and in the first week of the new month the equity is getting back to the balance again. This is with open trades following the H4/ D1 sixth rules.

PoS overview.png

Currently I have trades open at 14 pairs.

[billv]

Thanks Gertje
Is this the fully auto version?

[Gertje]

Thanks Gertje
Is this the fully auto version?

Yep, since november/ december it is.

[billv]

Yep, since november/ december it is.

thanks good to know

I agree that it can be very profitable, but IMO the DD is a bit too high.
I wonder how it would perform if it was using the Zeljco filter.
I'm certain the performance would drop but the DD would be reduced as well.
cheers :hi:

[Gertje]

I agree that it can be very profitable, but IMO the DD is a bit too high.

I find that the DD [after I let the EA to itself ] reasonable.
During the 'full auto' period it never came above 20%

I wonder how it would perform if it was using the Zeljco filter.
I'm certain the performance would drop but the DD would be reduced as well.

To be honest, I have no idea what this filter does......
Knowing Steve, the filter is somewhere 'under the hood' and silently waiting to be added.

[billv]

I find that the DD [after I let the EA to itself ] reasonable.
During the 'full auto' period it never came above 20%.

That's nice,
Is that 3 EA's running in parallel as per your manual method?

To be honest, I have no idea what this filter does...... .

It reduces the exposure to a particular currency.

HOW IT WORKS
UseZeljko: named after Zeljko who corrected the code for this filter and made it
work. This filter ensures 'balanced' trading. The easiest way to describe it is to
use hypothetical trades. Imagine a situation where:
1. the EA buys GBPUSD.
2. the EA buys GBPJPY. You are now heavily exposed to GBP.
3. Something unexpected happens (and something unexpected always
happens in Forex) and the pound plummets, dropping like a stone into
the abyss. Both your trades scream into huge drawdown.

In order to avoid this problem, having taken the GBPUSD trade, the Zeljko filter will stop the EA from BUYING another GBPxxx pair. However, a sell trade is allowed.
Imagine now that a Sell GBPJPY opportunity arises and the EA places the sell trade.
Now we have 2 GBP trades and one trades 'balances' the other so in the event of something dramatic happening, our trading account should not be at risk.
In our example, if our GU buy trade is screaming into the abyss, our GJ Sell trade is going stratospheric so it should balance out the equity.


I added this filter to the Awesome EA and I'm testing it together with the onlytradecurrencytwice function.
I have yet to see how useful it is over the long run but I like what it does,

In my tests I run 3 identical Awesome test accounts
1 with no filters
1 with Zeljko
1 with Zeljko+ onlytradetwice

The account with no filters has the biggest growth but at times has the biggest DD.
Anyway, I just started this test and I will need a few months of testing before I can decide if I want to use it.

[Gertje]

I find that the DD [after I let the EA to itself ] reasonable.
During the 'full auto' period it never came above 20%.

That's nice,
Is that 3 EA's running in parallel as per your manual method?

Yes, but every single pair is only run by 1 EA, so no duplicating there.

To be honest, I have no idea what this filter does...... .

It reduces the exposure to a particular currency.
HOW IT WORKS

Okay, that explains why I never used it.
Since pairs move in parallel, you limit yourself to one pair with Zejlko.
And like always, you find afterwards that you chose the pair with the least movement.....

I prefer to limit the lotsizes and the amount of recovery trades to limit the DD, and allow every pair to join in the frenzy.

[billv]

I prefer to limit the lotsizes and the amount of recovery trades to limit the DD, and allow every pair to join in the frenzy.

Yes that would work too.
In order to compensate for the reduced number of trades I have now added pairs which I previously excluded
so 2 of the 3 test accounts (the ones with the filter=true) are trading 29 pairs (instead of 23).