EA scams
This is a result of over-optimization (curve-fitting) and its perfectly normal, the results are looking good up to the day the optimization was performed and not so well afterwards.
This is not a proof of any scam, nor means that all MQL5.com Market products are worthless, so please do not generalize in your conclusions.
100% it is a scam with the buys and sells coded into the EA.
I tested his latest EA (neural network order block) and it was profitable every month for 6 years on 3 different pairs. The only time it wasn't profitable was the last 2 weeks so I checked the upload date and it was 6 november 2023.
That is when I looked at his previous EA and it was profitable up until the date it was updated but not after.
You can't overoptimise something to give a profit every month for 6 years on 3 pairs with a 90% win rate and average profit higher than average loss. Not possible with a formula.
Also when I backtested using the visual mode it wasn't even closing on the trailing stop loss which it was supposed to be doing.
I don't think I generalised. I said does no one check the code for scams. I also said if you can't trust them then any of them could be a scam.
I did some research and it seems that MQL5 marketplace do not check the code for scams but just say it is a marketplace and no more. So how could we trust something that could be easily manipulated by putting buy and sell data in the code. Seems a very short sighted tactic if that is the case and doesn't help the genuine sellers.
For me, there is absolutely no point in looking at any EA for sale if I can't trust them so I will go the way of learning to code MQL5 so I don't have to trust a black box. That was my longer term goal but at least I won't waste more time looking at EAs for sale.
100% it is a scam with the buys and sells coded into the EA.
I tested his latest EA (neural network order block) and it was profitable every month for 6 years on 3 different pairs. The only time it wasn't profitable was the last 2 weeks so I checked the upload date and it was 6 november 2023.
That is when I looked at his previous EA and it was profitable up until the date it was updated but not after.
You can't overoptimise something to give a profit every month for 6 years on 3 pairs with a 90% win rate and average profit higher than average loss. Not possible with a formula.
Also when I backtested using the visual mode it wasn't even closing on the trailing stop loss which it was supposed to be doing.
I don't think I generalised. I said does no one check the code for scams. I did some research and it seems that MQL5 marketplace do not check the code for scams but just say it is a marketplace and no more. Seems very short sighted if that is the case but it is their marketplace.
For me, there is absolutely no point in looking at any EA for sale if I can't trust them so I will go the way of learning to code MQL5 so I don't have to trust a black box. That was my longer term goal but at least I won't wasn't more time looking at EAs for sale.
Even if you've found an EA that you believe is dishonest, it doesn't imply that all MQL5.com Market products are untrustworthy. This is precisely what I mean by generalization.
All MQL5.com Market products are uploaded by sellers in the form of compiled ex4/5 files, making it impossible to check them for illegitimate practices.
I think I updated my above reply while you made yours.
Maybe it would be more work for the marketplace people and probably less money but I think they are letting down their genuine sellers and buyers. It would be easy to have sellers upload the mq5 files and have it automate the ex5 file.
I am sure there could be an automated system to find datasets and if someone was found to be using buy/sell data they could be banned.
Anyway, I just wanted people to be aware. I will go the coding route.
Good points and valid concerns but they cannot do that unfortunately.
For instance , if someone codes the trades in the EA , or ,someone has a neural network trained in a specific symbol , the later may honestly think he is offering something valuable .
Furthermore , you me or even meta quotes cannot tell if the EA has a neural net or is reading history by using the "change symbol name" test .(because if i let a neural net "learn" on symbols then the only way to know which "knowledge" to load for a symbol is by reading its name)
Also with overfitting , a seller could still be as genuine as someone who has spent 2 years building a strategy in that they actually believe they are selling a good product.
In some cases both overfit its just one of them does it extremely fast.
Lorentzos, I am not sure about neural networks or "change symbol name" but I think what you are saying is that there is no way of telling if someone is cheating as there are several ways to do it.
I wasn't really looking for a ready made EA as I want to know what the program does but was interested as I follow smart money concepts and the EA was partly based around it. I am glad I found out quite early that I shouldn't even consider buying something if I don't know what it does. Even if it is genuine I wouldn't be able to make adjustments if I think market conditions change anyway.
I just wanted others to be aware that even backtesting an EA means nothing and at least look for the latest upload date and check if the data after that matches the results before it.
There is a big opinion that EAs don't work and that you have to trade manually but I think the truth is that why would someone sell an EA if they could make more money trading it. So the EA that works is the one you build yourself or get someone to build after finding a strategy that works. That is a tall order and takes a year or 2 at minimum. I am not complaining though as it knocks out most of the competitors.
Does no one check sellers code for scams. For example this guy shows his EAs winning but they only win up until the last upload date and then they lose. He obviously has the data loaded in the EA
For example, his program was last updated on 28th february 2023 and you can see the great results before that date and the poor results afterwards
This means that nothing in the marketplace can be trusted and the whole thing is a waste of time. Why would anyone buy from the marketplace if they can't trust the code and they can do a genuine backtest?
Please note that graph you posted does not necessarily indicate that the EA in question is using a historical database.
It could just be using default parameter values based on optimised results up to the date of publication, meaning that it is up to the user to keep redoing optimisations on a regular basis.
In fact, most EAs in the Market, and in the CodeBase, and in general, are not self-optimising, and require regular optimisation for the parameters.
They similarly exhibit results as the one you have shown, when the forward test period is long compared to the optimised period.
EDIT: The same will happen to your own EAs if they are not adaptive or self-optimising.
Of course, not all sellers are scams and it's hard to determine that it's a scam 100%. But there is no doubt that this market is a comfortable place for scammers.
Please note that graph you posted does not necessarily indicate that the EA in question is using a historical database.
It could just be using default parameter values based on optimised results up to the date of publication, meaning that it is up to the user to keep redoing optimisations on a regular basis.
In fact, most EAs in the Market, and in the CodeBase, and in general, are not self-optimising, and require regular optimisation for the parameters.
They similarly exhibit results as the one you have shown, when the forward test period is long compared to the optimised period.
I am pretty convinced there is something underhand going on and others on here have said there is no way of checking so for me the marketplace is not an option anymore. I am not saying that all sellers are cheating but if there is no way of checking then it is all pointless. Most of the ones I tried don't even pass a backtest so at least he made an effort to make it look like it works :-)
I agree that the programs need to be regularly optimised but it is no good if they don't work on future data. I don't think that is what happened with this EA as it has given 90% win history over 6 years even after the 'opimisation' has been changed, but it doesn't make a profit after the latest upload date.
Anyway, if one person reads this and it makes them look before they leap then it has gone a little way to help.
Of course, not all sellers are scams and it's hard to determine that it's a scam 100%. But there is no doubt that this market is a comfortable place for scammers.
That sounds like a good way, like myfxbook, they have to be connected to a broker and live trade even if it is a demo account. That would sift out the scammers.
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Does no one check sellers code for scams. For example this guy shows his EAs winning but they only win up until the last upload date and then they lose. He obviously has the data loaded in the EA
For example, his program was last updated on 28th february 2023 and you can see the great results before that date and the poor results afterwards
This means that nothing in the marketplace can be trusted and the whole thing is a waste of time. Why would anyone buy from the marketplace if they can't trust the code and they can do a genuine backtest?