So bear with me, I’m quite new in algo trading.
I have developed some rather simple strategies. One of them was in 3 minute timeframe and I looked in the equity curve that during a certain period it worked good. Then I searched for patterns that occurred during this period in higher timeframes, like certain values in indicators etc, and added those as conditions to the strategy.
My question is if this approach generally is totally useless, if it’s over-optimizing in a nutshell? The strategy now is 35-3 (WR 92.1%), but only 7.5% in the market (I can only backtest to Dec-19).
Even if this is fun to do, I have a feeling it’s not something that will work in real trading. Any thoughts?
Thanks!
Easiest way is to do what I just wrote on another Topic
Zilliq Challenge : Give me a Donkey Indicator to transform in a RaceHorse 😉
Or just set your Algo running on Forward Test on your Demo Platform and then start to develop another Algo … while enjoying seeing your 1st Algo taking trades! 🙂
Thanks for your reply!
I haven’t learned about Walk Forward-test yet. I tried to do it now but it must be something wrong, it’s now 50-4 in wins/losses.., and I really don’t know how to interpret the test results. So I’ll learn more about this, and come back and perhaps ask more questions when I have more knowledge.
Thanks again!
In the meantime you could test on Forward Test (not Walk Forward Test) using your Demo Platform … as I suggested above.
Do you have your Demo Platform enabled?
Yes, I am mostly using the Demo platform. Sorry for my incompetence, but what is forward test? Do you mean starting the strategy as real time automatic trading?
Thanks!
Yes start as real time / live AutoTrading … this is called Forward Testing.
But on Demo for at least 50 to 100 trades … so you don’t lose real money.