How do you know when your strategy stop being profitable vs a short drawdown period?

When is it time to stop using this strategy / pause it?

Strategies and systems exist within market environments. If market metrics haven't changed, then you may just be in a temporary drawdown. If the market has shifted, you may need to adapt.

It helps to always reference what the market is doing when developing systems. If nothing else, at least align yourself with the VIX.
 
I am reading Dr. Howard Bandy's books, currently Quantitative Technical Analysis. His approach is to develop systems that meet your goals, i.e. profit, drawdown, risk tolerance, etc. As long as the system is performing within these parameters, keep trading. After every trade the system calculates the next safe position size based on your tolerance and when it goes to zero the system is broke.
 
Let say you developed a new strategy and backtested the past few years that looks to be profitable. When you decided to go live, the first few weeks is also profitable, but then it starts loosing consistently. Would you consider this "edge" is over or it's simply a short drawdown period? When is it time to stop using this strategy / pause it?

The infamous overfitted hockey stick loss curve. There are thousands of papers on the subject. Some people have even made a whole career on the back of it. Google Marcos Lopez de Prado to begin with.
 
It's also very be possible that your backtest results were overfitted.
You not only need to backtest, but you also need to forward test (in real time if possible) before going live. This is where demo/paper trading helps a lot.

It's too easy to overfit a system while backtesting.
I mostly trade FX pairs and ES, I tried to be as stringent as possible to develop something that doesn't require overfitting and relatively simple. I forward tested it for the past 2 weeks which was behaving well. But this week once I go live, I lost 3 days in a row. Looking back at the charts, I can say that I am unlucky and entered during a 3 day drop.
 
I mostly trade FX pairs and ES, I tried to be as stringent as possible to develop something that doesn't require overfitting and relatively simple. I forward tested it for the past 2 weeks which was behaving well. But this week once I go live, I lost 3 days in a row. Looking back at the charts, I can say that I am unlucky and entered during a 3 day drop.
You're not trading by hand?
 
Back
Top