And what has been the Darwinex answer to your complains?
You can read the answers in this discussion.
Well, to complete my thoughts on this thread. I¨d say OS and CS are great tools to guide the trader, but I think they shouldn´t be considered for the D-Score. Overall if I´d have to evaluate and rate the D-Score and all the Darwinex system my grade would be around 90% (outstanding). I´m just here looking for improvements and perfection…Thanks everybody. My two cents.
some months ago…
Now DLF leads the rankings by return and has 1 million AUM .
It seems that Os 2 and Cs 0.5 are not so relevant to measure the edge of a strategy.
I knew I was right, it was just a matter of time.
I THINK THAT BOTH Os/Cs and duration consistency are worthless in terms of measuring the success of a strategy. First, timing is only based on wether you buy or sell earlier but apparently it does not take in account the percentage of successful trades. A lot of strategies have awesome timing but only average around 30%-40% of winning trades. They don’t even win half of they trades. For example the PEW darwin averages 60% of winning trades and according to the way darwinex measures timing it has an awful open and close strategy. And there’s the other thing, because one cannot predict the future and the market does not always move under the exact same conditions it is impossible to achieve certain amount of pips on the exact time every single trade so that you can have a good score on those attributes.
Os and Cs are still not so usefull but now they are balanced by Loss Avarsion tha prizes the RR of your trades.
As for R+ R- and Dc they are useless I agree but this is not the subject of this discussion.
With new DScore darwins like DLF and OOA can reach high values thanx to high La.
The best example is FEG (Os 1 - Cs 3,7) with dscore 81,9 vs PEW (Os 0,7 - Cs 4,4) and 68,7 dscore.
The discussion of the weight of the timing is over from the new calculation of dscore… Today the weight is small (well compensated with the La to avoid the good timing of the martingales).
If PEW has not better dscore is not for timing, it is the score of Rs Risk stability, that it is which lows its dscore.
As @CavaliereVerde always had noted: For Breakouttraders the OS-Score is very disadvantage. I will point out more clear (i hope I understand the OS correct) with an example:
Breakouttrader (BA) goes to trade on a price of 100. He get filled and all is fine. OS -x (Time) must get a better price, because price has to be lower BA’s breakout. Let’s say OS gets 95 and it’s fine - not for the score of BA.
In an other case the price goes just up to 97 and then went down to 70. So BA don’t get’s the trade and it’s okay. But what about OS? OS get’s no trade too. But what is now with OS’s Entry at 95? It would have been a Loss-Trade wich do not count for OS!
CS is another thing and can even be very useful to improve to exit a trade.
… I don’t know why I call the Breakouttrader BA
How about OS and CS, many investors seems to underestimate this particular score, but from my point of view it shows some how edge of the system because trading is all about buying snd selling at ‘good price’.
Mmmmh are you sure? Try with a trailing stop…
Yes, trailing stop can affect it I agree. But if I would be investor than may be I will look at the behavior of OS and also CS (as you suggested system with trailing stop can have bad CS score so one can just evaluate it logically and not with rigid rules).
Martingales are impressive with OS and CS
Hahaha, it seems investing is even difficult than trading. Therefore, one need experience and there is no ready made recipe in this field. One can just analyse the system and reason behind different scoring or behavior instead of simply taking various parameter and fixed value of D score.
Similar to the balance sheet one need to go by individual company and study it hard instead of seeking some readymade signals.
Timing algo can be useful for some traders to fine tune specific strategies: no trendfollowing and no trailing.
It is completely useless for investors and has no predictive power for future performance.
I find opening a trade a lot easier than choosing a Darwin. Vice versa I always know when to sell a Darwin but getting the closing of a position right seems to be extremely difficult.
I think this might be because Darwins with a discernible pattern are more likely to be non-garbage (not mostly random). You have nearly 100% of the garbage (mostly random) Darwins if you just include all Darwins with a timing score less than 5 (and the group with timing scores above 5 have nearly 0% of the mostly random garbage. There will be plenty of good Darwins in both groups but the group with less than 5 in timing score will be unfairly penalized by all the garbage in it’s group) and you may not be comparing “good” apples to other “good” apples. Investors want to use it to compare “good” Darwins with other “good” Darwins. Try comparing Darwins with Pf >X AND Os Cs >Y to Darwins with Pf >X AND OsCs <Y… That would be interesting to me only if the average Pf and Pf distribution were the same in each group.
Otherwise it looks like OsCs are just a way to hint of possible “non-garbage” or to tell if a strategy is a top/bottom picker (or not) or maybe as others have said, to aid bottom and top pickers to pick better.
Luckily D-Score isn’t weighted by this much these days… (Thank you Darwinex)
Hello, in an effort to progressively improve the results, I am using Darwin statistics, but I do not fully understand Os, says that the most optimal would be “-30%”, how can this be applied to the opening of operations? Thank you.
Exact, the algo is telling you that you should open your trade BEFORE the signal !
Does it the total duration of the trade to calculate that 30%?
edit: does it takes?
Exact, suppose your trades last 10 hours, the “optimal” entry on average is 3 hours before.
But as I said it does not make sense for trendfollowing and breakout strategies, where the signal consists in entering an existing movement, not anticipating a reversal.