CFDs are complex instruments and come with a high risk of losing money rapidly due to leverage. -- % of retail investor accounts lose money when trading CFDs with this provider. You should consider whether you understand how CFDs work and whether you can afford to take the high risk of losing your money.

IlIlIlIlI’s D-Score demo portfolio

As I reset my demo portfolio, I made a new setup for my demo portfolio with a simple rule and a simple sort:

  • at least 1 trade last month
  • take the first 25 Darwins sorted by D-Score (here with an investment of $ 500 for each)
  • portfolio is to update daily.

Is D-Score really so bad as criterion as written on some posts here?

I try to find my answer this way.

The portfolio was setup on May 7th at night.
The leverage is 1.0 because the portfolio was created from the backtester for the first 20 Darwins with an investment of $ 500 for each Darwin, the rest were added manually.
IMO it can only work if there is a daily maintenance after the new D-Score calculation.

Current portfolio:


Trades history:

Wallet transactions (to add virtual money) after selling with losses:
May 12th: $ 5.93
May 13th: $ 31.19
May 14th: $ 13.39
May 18th: $ 1.13 - Sum: $ 51.64
Unfortunately the wallet transactions are not documented by Darwinex for demo portfolios.

Currently the demo portfolio shows a small net profit after subtracting these cash transactions.

Now I wonder whether I had a lucky day to setup this portfolio and whether it would be possible to make money with the D-Score top 25.

3 Likes

Here @Kingfisher1978 was doing the same with a weekly rotation and the result was -1.2% after one year, not a clear loser but something I would call breakeven.
I estimated that every full rotation of a portfolio has a cost of 0.01% due to spread and commissions from darwins with open positions.
The expected edge is small and rotating 250 times per year means that you have to beat a -2,5%.

1 Like

Also one week as used in your example Is too long if a Darwin suddenly turns into a falling knife.

I desagree but let’s assume I am wrong and you are right and one day is a better rotation timeframe than one week.
If one week produces -1.3% you expect one day to be so better to acheive +10%?

I don’t know, but that is what I want to find out.

Currently it shows +0.64% on the Chart since start.

1 Like

The experiments of @Kingfisher1978 are useful also to be aware of the behaviour of this kind of portfolio.
It is possible to have 2 consecutive green months or 2 consecutive red months and they tell nothing about the overall edge of the strategy.

1 Like

But there are differences:
I take the top 25 where the last one has a D-Score of about 75, can also be lower like 74.8 or higher. Currently we have 29 with a score > 75, some days ago there were only 23.

I don’t wait until the Darwin looses the score and goes below 75 as his experiments do. That mostly implies a loss if it happens after a recent entry. I can also have (and had) replacements by Darwins evaluated better on D-Score than the leaving ones and sell with some profit.

But I don’t know whether that is enough. The strategy needs more Darwins with a high quality rating and maybe a D-Score punishing DD more than rewarding profitable days.

A fixed number of darwins makes more sense I agree, with a fixed number of darwins we invest all the money but IMO the difference is negligible.
If DScore fails to predict the return of one week it shoud be effective to predict the return of one day?
Something callculated on ~1 year should be a predictor for one day?

The advantage of your daily rotation is the sample size.
If this time you hopefully execute this for 4 months it wiil be 80 rotations instead of 16 rotations. :wink:

1 Like

Not every day has a rotation, but most days have.

I am monitoring the DScore of my darwins for 4 years.
The daily volatiliy of return is ± 1% that mostly implies a change of ±0.5% in DScore.
We have 10 darwins from #20 to #30 in 2 points of DScore.
There is a lot of noise and threshold effect every day, a lot of buying and selling for nothing.

Not for nothing …

… and noise could also have a positive effect if the better ones replace the good ones. Losing a rank does not always mean losing on the score value or profit.
IMO that is the key for any successfull portfolio based on ranking lists. Meanwhile possible at Darwinex? :wink:

My usual top10 investors, no rotations.
Maybe it is just friendly market. :wink:

1 Like

It is definitely a friendly market if only 3-4 out of 25 show red today in this portfolio. That makes every prediction very unsure :wink:

Edit: if your chart shows a full month while mine shows only 1 1/2 weeks I assume the difference is just some noise as you call it.

1 Like

Exact, if investing DScore would work Darwinex would be doing it.
It is very easy to backtest it using the API .
If it were an efficient predictor they would not change it.
I think they wer doing somenthing similar with old DWEX.
DWEX the dynamic index made by Darwinex
Dwex was investing 30-50 darwins and was rotating every day.

1 Like

Dwex had too many Darwins to be successful on the long run, that is why it’s buried.

All other D-Score experiments I found here had a too long maintenance period to get rid of a falling knife in time.

D-Score 80 as a level is too high, most leaving Darwins leave losses in the portfolio.

The current D-Score is the engine of DarwinIA. Experience is overvalued to give everybody a chance on a price if he is long enough at Darwinex and didn’t crash his account in the recent months.

I agree but nothing prevented them to use 10 darwins intstead of 50.
And you have to consider that it was 2016 with the first DScore that was worse the the courrent one.

This is what you want to prove… :slight_smile:

It is a contest to promote talent, it excludes short trackrecords and extreme martingales from the game.
A decent tool to exclude bad trackrecords, not to find the best ones.

1 Like

I think they simply gave up after Dwex failed and they developed the api.

The current D-Score is not so bad, otherwise I wouldn’t run this demo portfolio.

1 Like

I agree it is not so bad but in my opinion in this way you are not investing DScore, you are investing the noise of DScore.
Even if darwins were not random walks the result of one day is too close to total randomness, and even if there were a small edge it would be strongly reduced by spreads and slippages.

It’s both, D-Score and noise.

The idea is that it could work if the better ones replace the good ones.

I‘m currently a little surprised that I didn’t need to add more virtual funds than listed above. The portfolio has currently $ 15 cash from replacements after paying some virtual fees.
I didn’t expect that in the first weeks.

Of course I cannot have a stable opinion after a few days, but it’s worth to continue the demo portfolio.

1 Like

https://www.darwinex.com/darwin/GBV#
-20% return, DScore from 76 to 54 in one day.
DScore is a lagging indicator.

1 Like