I have been a programmer for over 20 years and programmed in C, C++, Java, C# and JavaScript. We have a saying in the software industry Garbage In Garbage Out. As at least one other programmer has stated a software program, which is effectively what an EA is, is only as good as the person that coded it.
I trade as an amatuer using cTrader, apart from the fact it has a much more modern user interface the automation language is C#, which I am programme with for a living so very familiar with it. I am nearing retirement so my motivation for trading is partly to see if I can supplement my income when I retire, but also to keep my brain active.
I have automated many systems found here on BabyPips and ForexFactory as I was learning to automate cTrader. Many of these systems are genuinely profitable when traded manually but once you take the discretion element out of it as you must when automating it they are no longer profitable.
For example if you take a trade manually and it turns on you almost immediately you can get out of it quickly. It is very difficult if not impossible to programme that kind of discretion into an EA/Bot without it ending up taking out stops on good trades. A good EA/Bot will allow you to fine tune the settings to get the best out of it but there are a few constants that you cannot avoid.
- An EA/Bot will only be profitable if you run it for long enough
- Market conditions will change so an EA/Bot will not be profitable all the time
- You have to be prepared to accept losses
- The key as in manual trading is that the wins are bigger than the losses and there are enough wins to outweigh the losses
- What works on one instrument may not work on others.
As an example.
I have a cBot as they are called in cTrader that runs constantly on a desktop computer at home. It is based on a relatively simple system I found on ForexFactory. I am not allowed to embed links in posts but you can find it by searching for “Channel MA Short-Term System”. I have it set to always set a Stoploss and Profit Target with a 2:1 profit ratio. It takes trades fairly regularly on EURUSD and the general direction of the account balance is upward. The number of wins and losses is fairly balanced, marginally in favour of losses, but because the wins are always bigger than the losses the general direction of the small account balance is upward. I monitor the account and keep the trading balance at a fixed level, but leave the profits in my FxPro Vault.
When I developed the cBot and ran a backtest over the last six months it increased the account balance by over 500%. When I ran a backtest over 3 years it was less profitable but it still increased the account balance by 300%, over 5 years it was even less profitable but still increased the account balance by a little over 200%. Sadly the cTrader system doesn’t let me backtest over a longer period but my point is you can’t just run an EA/Bot for a short time and see the real capabilities. You have to run it for some time to see the real outcome, if the platform has a good backtesting tool this is easy if not you have to run it on a demo account for several months at least before judging whether it is profitable or not. In all these tests there were significant peaks and troughs in the Equity chart so in the real world there would have been opportunities to take good profits out of the account, but there were also times when one might have thought the system didn’t work as a string of losses piled up.
The cBot is quite configurable so I can tune it, but despite this there are pairs I cannot generate long term profits on. Yet when I trade these other pairs manually with the same system I can make profits.
