We’ve all seen the hype – automated trading robots and expert advisors promising astronomical returns and quick profits. You may have wondered whether these claims could possibly be true. Maybe you’ve even spent your hard-earned money on one or more of these products, only to be disappointed in the results.
I can’t tell you whether the latest offer to double your money in 5 weeks is real or not. (Though it probably isn’t). But I can tell you what to look for when evaluating any forex robot or expert advisor.
The Folly of Backtesting
Most people who sell expert advisors and robots back up their claims with an impressive historical backtest that shows a perfect linear equity curve and huge profits. Now, we’ve all seen EAs that look great after optimization, but perform poorly in real-time trading. Why is this? The answer is called “curve-fitting.”
When you optimize a trading strategy, the greater the number of settings optimized, the more likely that the results will be fitted to the historical data. An EA that has had all of its settings optimized at once can produce very impressive backtesting results that will trade poorly going forward.
Another consideration is how often the robot trades. Does it only trade at the close of the bar, or does it place trades intra-bar using a very short timeframe? Tick data during backtesting is extrapolated, and not 100% accurate. Therefore a short-term “scalping” EA will perform much differently than its backtested results.
In a nutshell, you…