In-sample overfitting: avoiding the pitfalls in data mining

1 May 2016

Social sharing

Share via Twitter Share via LinkedIn Share via Email

Probably one of the biggest pitfalls in investment, in-sample biases are often among the most difficult to convey and explain and also to avoid. Using simulations along with real world data we try to demystify this effect, explain the dangers and put forward suggestions as to how they can be mitigated.