The Generator and the Optimizer are so powerful that they may find trading rules that take advantages of every market move in order to find the most profitable strategy. We call that Curve-fitting or Over-optimization. To fight that problem EA Studio provides a Monte Carlo analyzer, which is designed to examine the strategy robustness.

Unfortunately, no matter what analysis we do on historical data, the real market always produces different results in live trading. Our studies show that 5% to 20% of the produced and analyzed strategies make a real profit. (The numbers depends on the program settings and the trader’s skills).

We found out that the only reliable way to filter out the bad strategies is to run our strategy set on a Demo account or on a Real one with minimal trading size. After a certain time (3-4 weeks ) we remove the failed strategies and set new ones for demo trading. We move the good ones to a real account and increase the entry size depending on the results.

There are also other tricks to test our strategies:

• Make a test with unknown historical data. - You can do that by using the Data Horizon tool to pre-define a testing interval. See a video: Create a Strategy that Profits for 6 Months

• You can keep your generated strategies. When you have more data, you can re-test the strategies or the complete Collections. You can use the Strategy Validator tool for that.