1/ Here you go again and you are wrong again. If systems fails it does not mean it cannot be profitable in the future. 2/ How do you measure likelihood that it will work in the future ?
I am not wrong. I did not say that it can't be profitable in the future , I said it can't be trusted, that is a big difference. In terms of measures , I will cover that in a thread on system development I will be starting soon. I want to put the material together and am currently getting ready for Traders Expo, where I am speaking. The biggest factor is what premise is the system based on. If it is based on a sound theory that really the most important factor.
You are wrong, you wrote two pages back " system which fails when bactested will surely not work in the future." In the light of this spinster explanation, I sincerely doubt that your expertize in system developing is better then average, if that. On the other hand you seem to be a good salesman . So stick with selling and do not get too technical about systems,you do not know enough.
Hi Walther: Don't worry about MR not testing or not demonstrating anything at all. He never did in all I've seen from him. Thank heavens we found a passable replacement for Jack. He doesn't seem to have Jack's carisma yet. Some NLP might help.
A system which fails when backtested will surely not work in the future. Backtesting provides a best case look at the results, so if you don't like the backtested results you should not trade the methodology. This is my exact quote, I though the definition of working is to make money consistently in the future. If that is the case then I stand by what I said. I not sure what your point is ? I would not trade a system which failed a backtest because this is a best case look. In the context of the second sentence it is clear what I was saying.
A couple questions: 1) What is your definition of a sound theory? 2) How do you determine you are working with a sound theory?
Hi Y'all. Have a question about b/t in general as it pertains to timeframe. If you have a long only system that tests 60% win rate in a 10 min timeframe vs. a 60% win rate in a daily system where BOTH entries are based on a "buy weakness" signal i.e. maybe a low bollinger band on a 10 min signal for the daily and the same bollinger band entry on a 1 min for the 10 min. ,would you have more confidence in hitting the win rate in the daily system vs. the 10 min.? IOW, given similar results , would you choose to trade system using longer timeframe due to slippage? I just think shorter timeframes would have a much higher slippage or missed executions which could skew the system for the worse. This can be due to the presence of a bigger population of traders competing for that breakout tick or that trendline break or fractal,etc. whereas the dailies would have a ot more chances to get in at a good price. Thanks
Your ego is clouding your judgement . System which fail when backtested might make money in the future if conditions change .. Everybody knows that. Except you, of course. My point is to show that you are not capable of correcting simple mistake if admitting it comes to a place. Which is deadly in trading. Lucky for you that you are in sales and not trading real money, so you do not have to worry about it. It is kind of funny how you keep hitting yourself in the face.
Let's now go back to my whole original quote. You can see in context we agree, a good backtest is not enough and you need to know when the markets have changed and will cause your system to fail. This quote is from a few pages back. A good backtesting is a minimum requirement. It also give us statistical information about the system and it's trade distribution. We can observe in real time as we trade the system if the underlying statistics are changing from the backtested values. If they are, we can keep a close eye on the system and evaluate when or if ,to pull the plug. I can take it that you don't like the concept of 100% mechanical trading systems. It is true that some can fail out of the box after good backtesting results, while other can continue to work for a decade or more. One example of this is my Adaptive Channel breakout concept, I developed in 1995, and originally published in Futures Magazine in January 1996. This methodology still works well today. Yes , I do trade my own system with real money. Another point is commercial systems are tracked today by many independent third parties like Futures Truth. This way it easy to see how different systems have worked since release.
If anyone is going to Traders Expo in Las Vegas next week, I will be there speaking. Futures Truth who is one of our resellers, has a booth and I will be available for questions on opening range breakout as well as other topics while I am there. Check for me at the Futures Truth booth next week if you are out there. I am speaking on Wednesday, I think at 5:00 pm. If you are going to Traders Expo check the updated speaker schedule for a confirmed time.