Forex robots – should we intervene manually?

Manual intervention

This question has been asked a thousand times or even more: should we intervene manually when we trade using forex robots? Should we stick to the rules or change them on the fly? In order to properly answer this question some philosophical background is needed, please stay with me and your patience will be rewarded.

Driving a car is deterministic systems because no randomness is involved in the development of future states of the system. You know exactly what happens or it is supposed to happen at any given state/frame of the system. The car movement can be mathematically described by¬† the equation of linear motion. The initial set of rules doesn’t change while we are driving. If you push the breaks, the car stops. This rule doesn’t change. You know exactly that if you drive at a constant speed of 80 miles/hour you reach the destination in exactly one hour if your destination is located 80 miles away from your starting point. Ok, you got it, now you know what a deterministic system is.

Forex market is definitely a non-deterministic system. There are no specific rules you can follow. The charts are nothing but a reflection of sentiments, emotions, economic/political news, panic, rumors, how can someone quantify that? Of course there are some patterns and they correspond to human behavior but we can’t make exact predictions. We know for sure that EURUSD will go up right after some strong positive news for the currency, but we don’t have a single clue about the future nature of the movement. It may go up in a straight line, it may retraces for a bit until going up, it can go down fast then up even faster or it could just hit a strong support/resistance line and go the opposite way in spite of the good news, there are many possibilities. We also don’t know when the movement is going to stop although the market may give us some clues. Support and resistance lines, highs and lows are specific to human way of thinking and we can speculate that, that is the reason why trading is not gambling.

Now, what exactly is an expert adviser? Is nothing more than a coded set of rules. Yes, rules. And rules are not flexible by definition. Even if entry and exit points are dynamic, there is still a rule involved here, for example ATR(30). Why 30? Why not 40? Or 12? Because according to backtests, the EA does better during a longer backtest period with this rule. But things are going to remain that way? I doubt. This means that the EA is not good? No. The strategy could be genuine but the main rule should be adapted to current market conditions and no computer program can do that, at least not now. Lemme show you what I’m talking about.

I have an EA that trades in the direction of the trend. First trades are profitable and it continues to open trades in the trend direction. The EA is brainless, it has no clue whether the trend is going to continue or not, but an experienced trades should know. Here is the example:

Trend EA

Trend EA

It opened 3 trades, all of them closed in profit, then it opened the 4th. But right after the opening, a strong rejection took place and the market turned the other way. This EA has a stop loss of 100 pips so it let the trade flow until it hit Stop Loss. An experienced trader should have closed it.

This EA is excellent but in order to maximize my profits, I supervise it and I don’t hesitate to close the trades manually if I my price action analysis tells me to close it.

Now you see why all strategies are doomed to fail eventually? Because they can’t adapt to market changes. The good news is you can still use your trashed EA with great results if the strategy is genuine. All you have to do is to properly reoptimize it to handle the current market conditions and to supervise it.

Forex trading is an art, not an exact science. I have a different visions on how the market works and how the forex robots¬† should be traded with after I saw so many genuine failed EAs, mine included, lol. A forex robot is a great indicator, if it survives the 13 years backtest it tells you that the strategy is valid, nothing more. The strategy is genuine and will work on the future, but you have to adapt and change it. On the other hand, if it doesn’t survive, the validity of the strategy itself should be questioned.

A properly chosen portfolio of forex robots could do a much better job then an individual forex robot, please see my previous articles on the subject, but if you really want stellar performances you should start supervising them but not until you read a lot about forex and trade on demo for at least an year. Next week my portfolio will be available for public, I’ll comment the performance every week so you can learn how to supervise an EA by yourself.