[PORTFOLIO] demo Portfolio with Darwin heroes of 2021

[IlIlIlIlI]

A second check of the portfolio content brought up that BGN with a divergence of currently -1.16 has also to be replaced.
The micro loss is not relevant in this case - and the portfolio had a positive divergence with it.

So we also see TKT again in this portfolio.

PUL and TKT were the most successful trades in this portfolio until now.

 

[IlIlIlIlI]

Today I added a max. DD of -10% for 1 year to the filter criteria:

That does not affect the investments in this demo portfolio, but cut the filter result by a few down to currently 55 Darwins.

 

[IlIlIlIlI]

Today I reduced the accepted monthly divergence to -0.5 to have more stability in the portfolio:

That does not immediately affect the investments in this demo portfolio, but cut the filter result by a few down to currently 51 Darwins.

 

[IlIlIlIlI]

WFJ as next on the sort sequence replaced ULC. ULC left because of a divergence shown as -0.64 .

The filter was extended to accept only Darwins with at least 3 trades in the last month.

And the portfolio continues to show green.

 

[CavaliereVerde]

Why WFJ and not TRO ?
TRO has an higher return on 2y so it shoud come first in your portfolio. (29% vs 27%)

TVS is even higher (35%) but maybe you wait 2 years of total trackrecord...

 

[IlIlIlIlI]

Why WFJ and not TRO ?
TRO has an higher return on 2y so it shoud come first in your portfolio. (29% vs 27%)

TVS is ever higher (35%) but maybe you wait 2 years of total trackrecord...

TRO is not in the filter as the 1y return is (still?) less than 10%.
TVS did not reach the 50% return until now (manual filtering, see the first post on this thread).

 

[CavaliereVerde]

the 50% return until now (manual filtering).

You could use "manage filters" in " all darwins" and combine your filter with the default filter" >50%"

 

[IlIlIlIlI]

You could use "manage filters" in " all darwins" and combine your filter with the default filter" >50%"

View attachment 304395

Thanks, then there are only 15 Darwins currently found in the result:
The 11 of the portfolio and UAE (currently visible, but under review because of divergence), TTK, RAT, DAS and ICX.

 

[CavaliereVerde]

I was using the initial filter just to show how to combine.
You are doing a good job refining the filter. 👍

 

[IlIlIlIlI]

As all Darwins in the demo portfolio are shown elder than 2 years, WFJ as # 11 had to go without replacement. The Darwin didn't trade during the hours it joined the portfolio..
To compensate the reduced investment, I added $ 250 to every Darwin investment in the demo portfolio.

 

[IlIlIlIlI]

UAE replaces TKT which was sold with small loss.

Porfolios continues to show profit for August.

 

[IlIlIlIlI]

Today happened what I feared since two months it could appear:

Unfortunately I had to sell the worst loser of this portfolio's history.
UAE and BFS disappeared from the filter.

The new portfolio composition looks like this:

 

[CavaliereVerde]

I think the main issue has been that BAX and YAX are the same strategy managed by the same trader, 0.98 correlation.

 

[IlIlIlIlI]

BAX, YAX were the worst, and also WGP lost its profit completely.

 

[CavaliereVerde]

Same punch on WGP and NWO could not be predicted, but BAX+YAX had to be avoided especially on a relatively compact portfolio.
Besides filters and parameters one of my oldest rules is no more than one darwin per trader.

 

[IlIlIlIlI]

5 Darwins of 15 out of the list are shown correlated in the top 10 of BAX: YAX, TKT, ICX, WGP and NWO.
Unfortunately Darwinex does not show the risk map before you enter a position.
So there is a deeper due dilligence needed for every Darwin to find it out.
And to find a rule what is to take and what is to leave out.

I'm also thinking about a change on the sort sequence to 6 months or 1 year and to reduce the number of Darwins to 6. I'm checking that from time to time.

 

[CavaliereVerde]

Correlation is the past, like return an DD.
Often what is uncorrelated in the past correlates in the future so I don't overestimate uncorrelation.
Uncorrelation is good for scholars writing books about portfolio (over)optimization.
Darwins provided by the same trader is a real fundamental information, true for the past and true for the future.

Edit:
TKT and NWO managed by the same trader, same loss today.

 

[IlIlIlIlI]

Darwins provided by the same trader is a real fundamental information, true for the past an true for the future.

That's also a feature missing that you can select the number of Darwins of the same trader accepted in the filter results.

 

[CavaliereVerde]

And to find a rule what is to take and what is to leave out.

I would use lenght of trackrecord, as I am doing about WFJ and TRO .
You like to use (past 😁 ) return for everything so you can go with that....

 

[IlIlIlIlI]

BAX left the filter and I changed the sorting to 6 months.
There were changes:

The new composition:

This is the last try as I don't know whether it makes sense to continue.
There happens the same as with all portfolios build by filters.
Maybe the correlation is too near by the filter criteria as Darwins found trade the same or too similar stuff when they appear.