Andrea Unger has won the World Trading Championship four times and owes it all to his trading system. Andrei uses 13 to 16 systems he has developed, some of which are automated, to trade in packages.

At present, Andre has developed and used both trend and trend trading systems, mainly trading on the futures market, including the German DAX index, the Euro Stoxx 50 index, the United States long-term Treasury Bill, the E - Mini S&P 500 index futures and currency futures. His system typically uses a 5-minute chart, and he only holds positions for the day, never overnight.

However, there comes a time when any system is no longer marketable. The hardest thing for the system, win or lose, is to determine whether the system simply made some mistakes when it lost money, or whether it has failed and is no longer suitable for the market. In general, the system will fail at a certain time, and when the rules they found no longer appear in the market, the results of the system will become irregular and start to lose money.

So when to terminate a system is crucial, and Andrei has a different skill in that regard. "First of all, when my system starts to lose money, my feelings and experience will tell me that there is something wrong with the system. So I take a closer look at what went wrong, run some data, and check the size of the current losses." His most common approach is to take the system's historical trading data and run it again, mimicking current market conditions, to see the current trading limit. However, Andrei believes that there is more than one way to terminate the system, and the timing will vary from person to person. The most important thing is that you have a plan and method, and decide when to terminate, make decisions before trading, and anticipate what you will do and how things will develop.

Development system 3 big secrets

1. Keep it simple. Even if genetic algorithms or neural networks are likely to give you superior performance over time, old-fashioned, simple ideas can give you good results.

2. Be careful and avoid using overly optimized models. The line between a good filter and a data extraction method is never clear enough, and if in some cases you don't see a compelling enough exit mechanism, the better option is not to enter.

3. Avoid spending too much time fiddling with all the bells and whistles that development tools offer, as this can cause you to lose sight of your trading goals. First, it is important to identify the goals of the system that need to be developed, and then step by step assess the steps needed to achieve that result.