How well do Expert Advisors perform in reality ?

 
There's a popular belief that trading robots "don't work". I know they can be profitable but how much really? For a few dozens of dollars you can buy EAs that theoretically perform several times better than the best stock market indices, but if they were doing really well in live trading, the sellers wouldn't need to sell them or they would sell them at a much higher price. It's as if selling Expert Advisors is more profitable than using them for oneself.
 
mt4ski:
There's a popular belief that trading robots "don't work". I know they can be profitable but how much really? For a few dozens of dollars you can buy EAs that theoretically perform several times better than the best stock market indices, but if they were doing really well in live trading, the sellers wouldn't need to sell them or they would sell them at a much higher price. It's as if selling Expert Advisors is more profitable than using them for oneself.

Do you think that if someone would charge $ 50,000 for his EA because he is convinced that the price is justified, anyone would buy that EA?

Why should a programmer not sell his EA if he can program well, but has very little trading capital. In this way, he can increase his trading capital.

I don't sell EA's myself, but I understand that programmers do; however, it would be good if these sellers had a (small) live account with this EA to demonstrate the performance of the Expert Advisor.

 
Gerhard Füssel:

Do you think that if someone would charge $ 50,000 for his EA because he is convinced that the price is justified, anyone would buy that EA?

Well you have several Experts selling for 30 000$ so there must be people buying them. It sounds more credible and you only need one sale to make as much as someone selling 1000 at 30$.

Why should a programmer not sell his EA if he can program well, but has very little trading capital. In this way, he can increase his trading capital.

Fair enough but how hard is it to wait one year if you know you'll be millionnaire? Also, when you sell your EA isn't there the concern that your trading strategy could suffer if too many people use it, not unlike business secrets? The big banks could trade against you if they saw good money to be made, though it's unlikely there would ever be enough people following your exact strategy to make them do this.

it would be good if these sellers had a (small) live account with this EA to demonstrate the performance of the Expert Advisor.

Indeed.
 

Interviews with the Automated Trading Championship participants

https://www.mql5.com/en/articles/mt5/interviews


One of them, Sergey Abramov, was asked:

"You've said that programming of Expert Advisors is a hobby for you. Does this mean that you do not treat it seriously?"

Answer: If you mean "seriously" as a source of income, my answer is "No"

He also said: Trading robots behaved not so well on real accounts due to requotes, etc.


What did he mean by "etc" ?

Interview with Sergey Abramov (ATC 2012)
Interview with Sergey Abramov (ATC 2012)
  • www.mql5.com
The trading robot of Sergey Abramov () is staying in TOP-10 since the second week. However, it caused much anxiety for its developer. As it turned out, the robot contains a small error in position close block. The robot has been developed almost exclusively on the basis of the past years' results. Hello, Sergey! We wanted to talk to you for...
 
mt4ski:

Interviews with the Automated Trading Championship participants

https://www.mql5.com/en/articles/mt5/interviews


One of them, Sergey Abramov, was asked:

"You've said that programming of Expert Advisors is a hobby for you. Does this mean that you do not treat it seriously?"

Answer: If you mean "seriously" as a source of income, my answer is "No"

He also said: Trading robots behaved not so well on real accounts due to requotes, etc.


What did he mean by "etc" ?

Things like network lag and realtime changes in spreads, other things like swaps and in some cases even commissions.

A test environment is usually an ideal environment and so it differs a lot from the real world.

You can create something that comes close, but there will always be some factors that are outside of your control.

 
Marco vd Heijden:

Things like network lag and realtime changes in spreads, other things like swaps and in some cases even commissions.

I am aware of those already and I try to mitigate them. For example, a VPS will help with network speed and I test my EA with large spreads.

I wish I had the real tick data with the spread for backtesting but my broker said they don't have it.

 
mt4ski:

I am aware of those already and I try to mitigate them. For example, a VPS will help with network speed and I test my EA with large spreads.

I wish I had the real tick data with the spread for backtesting but my broker said they don't have it.

You can start creating systems that only uses prices and indicator values from last closed candle. This is a good way to not soffer from tick rounding and also to create more robusts systems that will perform so close to backtest results (in case of backtest are made in the right way).

 

When a broker tells you they don't have something, or they do not offer some service that you need, it's a good idea to look around at what other brokerages offer.

Simply because others are somewhat more professional and will offer you a wide spectrum of supported applications and services, while others only offer a limited set only and these are the ones i personally tend to avoid.

 
Fabio Cavalloni:

You can start creating systems that only uses prices and indicator values from last closed candle. This is a good way to not soffer from tick rounding and also to create more robusts systems that will perform so close to backtest results (in case of backtest are made in the right way).

Tick rounding? I am not familiar with the concept and how systems suffer from it. Can you provide a documentation for explanation?
 
Marco vd Heijden:

When a broker tells you they don't have something, or they do not offer some service that you need, it's a good idea to look around at what other brokerages offer.

Simply because others are somewhat more professional and will offer you a wide spectrum of supported applications and services, while others only offer a limited set only and these are the ones i personally tend to avoid.

Sometimes one is willing to compromise if the advantages are big enough. But even if you find a broker you like, it doesn't mean they accept clients from your country. It has happened to me. I might create an offshore company eventually to get around that limitation.

 
mt4ski:

Sometimes one is willing to compromise if the advantages are big enough. But even if you find a broker you like, it doesn't mean they accept clients from your country. It has happened to me. I might create an offshore company eventually to get around that limitation.

And then you will quickly discover and learn why to avoid them.

There are always way's around limitations it has always been like that.

Reason: