The safest way to develop a system is to use two seperate data sets. Say 2005 to 2012 would be your in sample. You would create all your filters, entries, exits from this. You would then compare the metrics from this data against the out of sample data from 2013 to 2017 (without adjusting any of the entry, exit parameters). If the out of sample produces good results then you have much more confidence that your backtest hasn't been data mined or curve fitted.
If you have bad (losing) years in your backtest (or the majority is from 2008 / 2009) it doesn't sound like a system you would want to trade in all market conditions. And in reality it would be incredibly hard to faithfully stick to for such a long period of drawdown
Seek out @kevinkdog and his thread. It has excellent info on how to conduct good testing
The following 2 users say Thank You to sixtyseven for this post:
Hi, Sistyseven. Thank you for your suggestion and tips.
Like you said, I have tried two back test using 2006-2012 and 2013 to 2017 and happy to report the result looks identical to a full 2006 to 2017. It did great on 2008 and 2009 as well.
Wish I was be able to come up with a strategy to be profitable every single year, but I could not. Boy I have tried for years but perhaps I am not that good or smart enough to make such a trading strategy. I have seen some great system results from this forum and my system is no where near such trading strategy but it is simply the best I can come up with.
Simply put, trading pull backs when a few conditions are met. (Volatility, trend, etc) It only trades just a little over 100 times a year.
My current result performed much better than back tested result. I believe it will bring itself to normal ratio in future which is winning 54% of the time.
Thank you,
Kanepa.
The following user says Thank You to kanepa for this post:
Hi Kanepa,
I would like to know exactly what ninja strategies you use? and also you don't enter or exit the trade manually? I really would like to understand how the strategies work.
Thanks for your response
You can build a trading system so it can enter and exit on its own. I do not put on trades. The system does.
If you are asking details about my methodology, I can only share what I have said earlier. It is not provided by ninjatrader or anyone else. I have built it using the ninja trader strategy tools. You can definitely try your own method using ninjatrader. They have a great support if you have any questions.
Please note that I am not affiliated with Ninjatrader or any other brokers. I did use Ensign awhile ago. But when I test drove Ninja trader, I really liked it compare to other charting software.
Wear a mask, mandate others to wear masks, & remember that @WHO is criminally incompetent. To repeat:@WHO is criminally incompetent. — Nassim Nicholas Taleb (@nntaleb) March 31, 2020
The following user says Thank You to aquarian1 for this post:
