Q: I have developed a trading system using many of your suggestions and it is almost complete. After many trials, errors, and losses, I now understand that exits/stops are more important than entries.
I have two questions:
- In your books you mention using Monte-Carlo simulations to test your system. Can you expand on that? Can you point me to any articles that you wrote in the past regarding this topic in other newsletter issues?
- Do you think doing t-tests (statistical distribution models) to detect whether or not if your trading system is failing (accumulated profits dropping) is a valid approach? I usually use daily stock price data over 10 years and/or 18 years to do backtesting. In this case, would t-testing between the past 170 trades and the most recent 10 trades have any meaning to see whether the average R per trade is falling so one should stop using the system?
A: We have simulator built into the Position Sizing Game in which you can enter your system’s results and then play the game by trading your system. That’s an easy way to simulate trading your system.
As for testing systems and deciding whether to trade it, you have to determine first your criteria for trading a system:
- What market types does it work in?
- Does it fit my beliefs?
- What kind of performance do I require?
- How will I measure the performance?
These are a few of the many personal questions you need to ask to decide when to trade and when not to trade a particular system. —Van
