Backtesting: Should You or Should You Not?

Upon opening an account with a forex broker, learning the basics of trading, and solving some of the simplest problems associated with trading psychology, the most important issue that you will face is the implementation of successful forex strategies. Analyzing the market, identifying opportunities, and exploiting them successfully in the context of a workable forex strategy, summarizes the tasks faced by the forex trader. Clearly, creating a profitable strategy is one of the most important obstacles faced by a trader.

But many traders lack the necessary background that will allow them to create and implement strategies confidently. There is no way of knowing if a strategy works or not without taking some risk, but the beginner does not want to take risks, as a result, backtesting is the most frequently chosen path.

What is backtesting? While backtesting a forex strategy, we feed a trading platform the algorithm defining our strategy, and let it apply the same to a series of historic data stretching back to many years into the past. Past records are available in many cases for ten, twenty, thirty years into the past, and many traders are enthusiastic that if a strategy is proven to be workable in that time frame, it has a high chance of working in tomorrow’s market as well.

Unfortunately, that is far from being the truth. Markets possess two important characteristics that make such methods unworkable. First, the price action is self-similar: there is no beginning or end to it, and any period, five minutes, to five years can be broken down to generate equally confusing and complicated patterns. In other words, backtesting on a ten year basis can be equal to backtesting on a ten-minute basis. Second, market action is chaotic. There are some processes, such as the triggering of earthquakes, or some meteorological events which cannot be predicted with certainty by mathematical methods as they are available to us today. The price action is one of such processes, and as such, the efforts at discovering magic methods that work under all circumstances is doomed to failure because of its basic presumptions. The rules that govern market action changes along with changing prices, so there is no point in applying something that worked in the past to the future and expecting it to be profitable.

Backtesting would be most useful as a learning aide while you learn forex. Studying technical analysis, and testing your understanding of indicators and their properties is much more convenient in the calm and risk-free environment of backtesting. Just devise your strategies, and check to see if you interpret an indicator correctly by backtesting it. But don’t expect to use backtesting as a filter mechanism for isolating forex strategies that will work in the future, and never trust someone who makes such a claim. If you do, you’ll find out, we’re afraid, that this approach just does not work after suffering significant losses in the process.