Hey guys and gals,

recently I came across some material about strategy performance robustness testing. Basically, trying to answer whether a (good) backtest performance is given by overfitting or by true strategy edge.

I wanted to ask in here, whether this kind oftesting is something many/all of you do. And if you do, what technique do you use? Do you have some material to recommend?

As for me, there is the CWFA feature of TradeStation. Though slightly tricky to use and horribly slow to perform, it has some merit. Also, I have come across some papers that deal with this topic, e.g. The Probability of Backtest Overfitting. The authors have written other papers related to this topic, I am however more interested in hearing about other methods. The topic is quiet new, so I am more than happy to hear other people's insights.

Best,
Daniel

A good book in this field is the book written by Kevin :

Building winning algorithmic trading systems...