Whenever you have a large number of candidate variables applied to a rarely occurring phenomenon, there is the risk of overfitting your model and mistaking the noise in the past data for a signal.

The downside of machine learning is that it typically leads to results that are somewhat difficult to interpret and explain.