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.

KVL autopsy, rise and fall of a contrarian strategies portfolio

When I started trading on Darwinex I remeber a much better situation, now every spread seems to be ridiculous if we exclude the main pairs.
I have been too trustful about trading conditions.
Just now I am checking if it is worth trading EURAUD on LSC .
It is a nonsense to give away money for nothing because of insane spreads and swaps.

1 Like

Apologies, on further testing the average spread seems to be about 2 pips until the recent pandemic fears.

EDIT: oops that was Dukas data, stand by…

1 Like

Average spread encountered by my EA on EURAUD Darwinex Data between
2017.03.22 - 2018.03.22: 19.13510392609657 POINTS
2018.03.22 - 2019.03.22: 20.29953014878545 POINTS
2019.03.22 - 2019.12.22: 18.18051404662213 POINTS
2019.12.22 - 2020.03.22: 15.24349157733615 POINTS

So I guess it’s just the very recent data with such a high spread. Note that the higher volatility can make up for at least some of the wider spread if you’re using wider SL and TP (and smaller size!)


Are we sure we should exclude them? I trade only GBPUSD on IRY. Overnight spread almost always above 10 pips, even during the Tokyo session. Spikes triggered 20+ and 30+ pips spreads.
I lost almost 1% only on spreads last night.
Same system with the same settings on IC (testing the EA), has a difference of 1% in today’s performance :slight_smile:



Yes the spreads are “wild” even on majors. During the London Session i´ve seen >5 pips spread on GBPUSD.


I want to discuss this a bit further without constraints. :wink:
We all know that results (trackrecords) matter much more than words and discussions.
I think this Community is not so useful as advertising space but it is useful for smart investors to distinguish transparent traders from the rest of the population.

If you discuss you work here it does not mean that you have an edge but at least it imeans that you don’t fear to be examined by your peers.


Hi @CavaliereVerde
Well I think you misunderstood me. By didn’t age very well I ment not your trade results!
What I ment with it was the above comments in that particular topic.

I did not mean to offend you in any way!


I was exactly referring to those comments, no offense from you! :smiley:
Let’s put in another way.
I think true traders should be willing to discuss their work here and should be able to defend themselves.
So I like to lead the way giving the example.

1 Like

Alright, it is good to be clear on that :smiley:

I do concur with what you say about being willing to discuss their work.
Also I think that for most traders it is difficult during those times.
In a period of a large drawdown many traders will have all sorts of feelings (ie: doubting the system, doubting themselves, maybe even doubting to go through with it all).
Often in these times, I see a lot of comments which may cripple the trader which in effect will cripple the returns and all.

I see you are quite comfortable discussing the results (which is very good in my opinion) although I can imagine not every trader will have this easiness like you do.

All I wanna say is that we all could be more understanding and less judgemental in times of crisis.


Hi CavaliereVerde,
I am sorry for the failure of your TS I gave a look at the lite-C. What I miss in your script is any equity trading filter to avoid long loosing period. Something based on a simple SMA on last 9 trades for each Asset/Algo combination enabling/disabling trading on the unprofitable component as an example.

1 Like

I would be the same as the first suggestion by @bendex .
Easy to find a filter to exclude the last losing streak but it would be another rule and another parameter, reducing the freedom and the robustness of the system.’s-path

I agree with Kevin Davey.

1 Like

I agree too. If a TS is profitable, no equity filter provides better results. But when a TS expires as yours, it saves from the disaster. equity trading is an insurance: you have to pay for it, sometimes it saves your ass…
If you shuffle the prices curve in order to have a pure random prices you will see it.
Anyway these are my only 2 cents. You know better than me what to do . Bye

1 Like

Are you sure? In this case?

I certainly did not suggest adding a filter directly to an equity curve like that. There is a major difference between a filter based more directly on the underlying market data (i.e. moving average filter) in a logical way and a filter based on recent strategy wins/losses or it’s equity curve.

Being paranoid of bias should be the default state of mind for sure but it is the optimizing and or poor selection method of the new conditions and parameters which are most responsible for making a system prone.

There is a stark difference between:

A - finding a new filter or tweak and it’s necessary parameter values by brute force and or adding it to your system by tossing it in to the walk-forward process to be optimized regularly with all your other optimize-able variables.


B - Going carefully through a delicate but extensive and reasonable process of attempting to find a new filter or condition that is not going to be changed or optimized.

Kevin Davy also talks about tweaking old dead strategies to make them alive again.


100% agree.
Let’s put it in this way: I prefer to focus on the other 2 darwins and 2 different investors portfolios.
If someone want to improve and revive my reversal algo here there are the rules, the code in mql4 and lite-C .
A lot of variations are possible:

  • changing the main indicator
  • using a volatility filter intead of ADX
  • using a time exit instead of SL and TP …

I dont’ have a math PhD and neither a team of quants working for me.
2 darwins and 2 investor portfolios are enough for a humble chemist. :wink:


I don’t know anything about algo / automatic systems.
But I understand that any automatic system has to be previously cooked in a trader mind, then try to replicate it. On fact, it should have been a successful manual trading account first.
Still, I can imagine it is a huge task…clone your mind…mmm
I read something said by @bendex and @FedericoSellitti that make a lot of sense to me (thinking from a trader mind), the SL have to be set based on technical analysis, otherwise it will be hit it once en again even when the trade should be a profit. So, once the SL is set based on technical analysis, the value per pip would be variable and calculated based on your money management (not the opposite).
So, based on that I am questioning now if the whole set up of this strategy was made in a trader mind or in a programmer mind…
I wish you the best, so I hope you receive my words the way I intended to say them…positive constructive.
I know this strategy is closed, my concern is if your other two strategies have been designed by a trader or a programmer mind.

1 Like

This strategy has been never executed manually.
It worked on H1 from 7am to 11pm , not doable.
BTW if you have a precise ruleset manual and automated execution have the same result.
I am not a professional programmer so initially I monitrored the EA looking for bugs, to see that was technically working, executing the rules as planned.
For me it was perfectly fine to have variable risk per trade with a range of 0.5-2%.
Fixed pip value or variable pip value both make sense, what makes the difference is making pips.
If your entry and exit strategy has an edge you make pips and money, than if you want you can optimize the risk management of the single trades or balance the components of a strategy portfolio.

The typical way to write an EA is following:

You see a repeated market situation where you could obviously make money again and again.

You start with a template to write your EA and a lot of reading the mql instructions and the board.

You find out that the full idea is not supported by mql.

You do your first backtest and produce more losses than profit.

You go back to the analysis to find more criteria to reduce the number of losses.

After defining and adding 17 indicators, 12 time windows when your EA will trade and 3 more for Fridays and Mondays you suddenly find out after spending tons of hours on backtesting that the strategy is highly profitable on AUDUSD in the current quarter.

You don’t tell anybody that the idea was originally invented for EURUSD.

Then you open a real account, fund it and go to social media platforms to sell it.

The rest can be seen there … :wink:


I understand that, however I can’t imagine to program something that hasn’t cooked and proven first in my head.

No, some things are not possible to clone. Trading manual have rules, you can duplicate them but you can’t duplicate your thoughts.

I trust price, not EA’s.

The important is to set the SL where your technical analysis says, not your pocket (from there on you make your money and risk management).

No necessarily if the SL and/or TP are set wrong.

1 Like

The concept of RSI , ADX and ATR are clear in my mind, their combination makes sense so I coded and tested it on past data.
I can’t immagine to trade something that I cannot verify on past data.

As I written in the first post SL and TP are calulated with ATR , that is a technical indicator.