Can MqL5 signals' track records be trusted?

[riddelln]

I have recently began to doubt whether the track records of MQL5 signals can be trusted for a number of reasons. In a future post coming shortly I would like to give some of the reasons for my suspicions. Before that, however, I am curious as to whether other people have similar suspicions, and if so, why?

[Taiwo Okunbanjo]

riddelln:
I have recently began to doubt whether the track records of MQL5 signals can be trusted for a number of reasons. In a future post coming shortly I would like to give some of the reasons for my suspicions. Before that, however, I am curious as to whether other people have similar suspicions, and if so, why?

Yes, of course. I wish there would be a new platform for that. 90% of products fade out with time in terms of performance.

[Alain Verleyen]

riddelln:
I have recently began to doubt whether the track records of MQL5 signals can be trusted for a number of reasons. In a future post coming shortly I would like to give some of the reasons for my suspicions. Before that, however, I am curious as to whether other people have similar suspicions, and if so, why?

Please explain your reasons.

[Mirko Cerulli]

I agree.

There is more than one way to fool the statistics, but you should show your reasons in order to create a discussion.

[riddelln]

I have just added a new thread - Can MQL5 signals performance history be trusted? (2)

This is what I wrote on that thread...

This is a follow up to an earlier post Can MQL5 signals performance history be trusted? I welcome comments and contributions...

About the beginning of May three signals were put up on MQL5 with excellent results. I was interested. All had results stretching back about two years, all had DDS below 10%. They were Arrow, Y2016 and Zetta.

How do we know if the performance records are accurate? There is a month by month performance record on all of the signals. If you press the equity tab for any signal, you get a history with DDS. However, almost none of these histories for any signalstretches  back to the supposed beginning of the performance record given in the months and years record below the main chart. Does that mean the records are false?

Zetta started with a DD of 7.27% and yet within one month it had crashed and burned with a loss of 97.47%. It's impossible to believe that it had run for two years with a low DD below 10% and then crashed completely within one month, impossible to believe that the performance record was true.

Y2016  had run (supposedly) for almost two years with a DD of 5.96% but then within just over a month, it's DD went up to about 12%, then 15% and now 27.9% - very suspicious that it should run for almost two years and then suddenly start having higher DDS when it has been published. Even more suspicious, someone commented on the signal page that their broker TradeKings (sounding just like the famous broker Trade King) was a fictitious broker, so very easy to falsify results. Then , also suspiciously, that post about TradeKings disappeared.

Arrow has not had such a big increase in DDs, but someone commented on the signal page that the trades were very different to the history, and also looked very similar to the trades run by Zetta and he thought they might be the same EA running both signals.

Now MQL5 makes a commission on subscriptions, but surely they would make more money if subscribers made more money from their subscriptions. Surely they have a vested interest in checking the claims of signals' performance records>

My question is, and I welcome comments, how can MQL5 signals performance records be trusted? 

[Alain Verleyen]

A second thread is not needed, I removed it.

[Mirko Cerulli]

In this case it is not about false statistics, but lacking of.
The statistics start when the trading system comes loaded on MQL5.

From here on, the drawndown is calculated on floating positions also and not only on closed positions.

https://www.mql5.com/en/forum/60013