Last Thursday, I spent a great deal of time explaining my latest short-term trading strategy. We’d gone through all the various permutations of the setup, covering time of day, type of currency cross, spread criteria and a million other details that would make a novice trader believe that we could execute these ideas with the steely precision of fine Swiss watch.

Then of course when time came to actually trade, all hell broke loose. The news was good. No the news was bad. I went long pounds, then quickly changed my mind got out at a scratch and went short AUD/JPY. The Aussie trade moved quickly, and some traders entered late. I scrambled to cover half at a small profit as soon as I could and then quickly went to break even. I targeted the second half of the position at 1.5 times risk, but when price stalled just ahead of my target I covered the rest and rightly so, because the pair quickly retraced back to original entry level.

On paper, the trade looked like a well planned, well analyzed event risk scalp that yielded 50 quick points in a matter of minutes. In reality, it was a mess of a news trade with most people in the chat room banking just 35 points of the move.

That’s no easy task and is one of the reasons why trading is such a challenging discipline. In all the time I have traded, I have never been able to solve this problem to my complete satisfaction, but I have come up with a practical workaround. I divide my trading capital into two accounts. I have a small account where I allow myself to be as emotional and experimental as I want to be and a large account in which I trade only the well worn strategies that I have tested.

Yet this is precisely why I love doing live trading. It demonstrates in no uncertain terms the wide gulf between the clean, crisp mathematical foundation of trading theory and the mayhem and chaos of trading practice. It shows that no matter how much we try to make it a science, trading will always remain partly an art.

But despite the messiness of the process, there was a method to my madness. Although the execution was hardly textbook, my trade management was governed by rules that I had developed over years of experimentation. Those rules helped me react to the volatile price action and ultimately walk away with a gain rather than a loss.

In trading, just as in real life we need to rely on the logical foundation of science to conduct our affairs, but we also need the flexibility of art to adjust to the ever changing environments of the currency markets and greater world around us. I may be a skeptic when it comes to computer algorithms, but I am firm believer in rules based trading. It’s what separates the pros from the novices.

Boris Schlossberg serves as director of currency research at GFT, and runs bktraderfx.com.