IMO if you want a quick fix for this, the solution is to have max one darwin per provider.
Yes, but imho having several darwins from the same provider it’s not a problem if they are not too correlated and they do not represent too much of the portfolio.
I prefer to continue like this while waiting to be able to easily filter the correlation.
The problem is now temporarily solved because only one dawin of this provider remained in the filters.
For the month of September, the portfolio contains :
ABH AQA BAS CBH CBY DLI DZL ERQ EZX FEG GTD ICX IDT IGX KVL MYQ NTI OLE ONE OOA OVL PUZ PYC SCN SCQ SKJ STP STV SYO TGB THA UEI UYZ VRT
Darwins sold :
JCW LVS CEM VQB WFJ BAD ATL XYF
Darwins bought :
ONE SCQ UYZ OOA AQA UEI CBH MYQ FEG SCN OVL IGX
Have a nice day.
Continuing with my journey analyzing the initial portfolios of some people in darwinex, I come to this one
Except ERQ (divergence too big), I took your initial portfolio and check what would happen if it would have been Buy&Hold since 2/1/2018, until today, or if the darwins would have been sold when they arrive to different D-Scores.
As @purple stated, the numbers are not entirely correct, and @CavaliereVerde explained that I could just use zoom to get them correctly. But I already had the calculations almost done and I didn’t want to do them again Also, the point here is not the exact amounts, but analyzing if D-Score could be the main trigger to sell a darwin.
The real numbers won’t differ too much with mines, anyway. For instance, using zoom to start from 0% (so, to get real numbers), KVL since 2/1/2018 would have been:
Selling on 23/9/2018: 27.38%
Selling on D-Score==50: 27.38% (never)
Selling on D-Score==55: 27.38% (never)
Selling on D-Score==60: 2.73% (8/3/2018)
Selling on D-Score==65: 4.65% (5/3/2018)
Selling on D-Score==70: 7.54% (1/3/2018)
So, after doing this with multiple portfolios, and although there are always exceptions, I think that selling every darwin that goes below 70, seems good to get better returns than lots of darwins trading. A different problem would be to analyze when to enter again.
Using D-Score<70 as exit we’re getting 15.73% return; if we add divergence and those math issues, probably we’d get something like a 11-13% YTD return (without leverage).
So, here we go:
EDIT: Please check a couple of messages below to see the correct table, this one is incorrect. I keep this one here just for historical purposes to avoid someone else makes my mistake!
Much respekt for your great work @NemesisCraze but you should use correct formula for the futere to get a more valid result.
For example FEG:
2.1.2018 quote 127,32
23.9.2018 quote 158,92
Wrong formula: 158,92 - 127,32 = 31,6 (your result)
Right formula: 1 / 127,32 * 158,92 - 1 = 0,248 = 24,8% (correct result).
@purple thanks a lot!!! I changed the formula and indeed the results are very different!
EDIT: fixed table as per messages below
I’ll update the other tables later today.
So to sell a darwin with D-Score wents below 70 should really work well. Most darwins will have lost more after DS went down.
KVL 152%? I would be very happy but I think there is something wrong.
I suggest you to double check everithing above +60% or below -60% .
Yep, I’ll double check later (no time now), sure, I want the calculations correct. I’ll post also the table so you can watch also if I did some unwanted mistake.
Thank you @CavaliereVerde, there is one more mistake - the quote is alway the profit + 100 (where everyone starts) so again the example for FEG
2.1. quote 227,32
23.9. quote: 258,92
So the profit is 13,9%.
It is much more simple and reliable to use the custom zoom .
haha this really works nice, thank’s again!
Hi! Great work!
I dont’ understand how you manage when a darwin starts with a Dscore below 70… like WFJ or UYZ… do you take only exit under 60?
Ok, I think the numbers are fine but the calculation not, as @purple stated. If I watch KVL returns since inception:
1/1/2018: return is 20.75%
23/9/2018: return is 52.35%
1/20.75*52.35-1 == 1.5229 == 152.29%
And this seems the problem. So:
1/120.75*152.35-1 == 0.2616 == 26.16%
1/227.32*258.92-1 == 0.1390 == 13.9%
Is this ok? Please let me know so I can update all the tables.
This is how a community works!
Yep. The problem is that I think I found a bug with the zoom and I can’t use it for all the darwins So while it is being fixed probably this is the easiest way.
Great question, I forgot to explain. In those cases I just simply wait to the darwin to grow over 70. And then as soon as it goes below, I do as usual.
For instance, on 1/1/2018 KVL starts with 68.3. Then, on 4/1/2018 it gets 70.18. So from that moment I wait to “sell” until it goes below 70, below 65 and so on.
If a darwin D-Score doesn’t grows over 70, I don’t analyze it. I know that we’re missing some info here, but some compromise needs to be done…
So, there are 4 darwins that start below 70… do you buy them when they go above 70? and then analyze them with your method ?
This is the reason why result of 70 are a bit flawed, theoretically they are correct but it acts as a take profit.
There are many reason to enter a darwin, it would be better to concentrate on DScore as safety stop more than a target, targets have nothing to do to with buy and hold.
No, I buy all of them on 1/1/2018. Then I wait until they go over 70 to start trying to sell it.
But you’re right, I see now the survivorship bias here. I know which darwins will go over 70 and which not, so I’m just buying the ones that will go over 70.
I guess that the only way to “fix” this test is only buy darwins that are already over 70, but we know that there are promising darwins and good ones that still are not there. So my test should only be taken as that, a test to know an exit indicator. I guess that the idea is to identify good darwins, buy them, and then wait until they grow over 70. If they don’t grow over that, another different exit strategy would be needed.
Exactly! One way should be to buy a darwin if it goes OVER 60 (let’s say that 60=6 is a sufficient score, like in school) and SELL it only if it goes below 50 (insufficient) … but if it does go over 70 then let the “trailing profit with dscore” rise to 65… if it goes above 75… rise it to 70 and then stop. Just mytwocents… I’m waiting for the latest “raw” data to be aviable to do some tests like this.