Come play the backtest and overfit game!

 We've written before about the risks of excessive back testing in our posting The Effect was Never There... It Was Just a Random Pattern. But now thanks to the good folks at the Scientific Data Management Group at Lawrence Berkeley National Laboratory at UC, we have a nifty little gizmo that will take you for a test drive.

Give it a go on the link below: then open your own quant hedge fund!

Backtest Overfitting: An Interactive Example

If you want to actually read about the toy before you use it, well, you go here, Backtest Overfitting Demonstration Tool: An Online Interface, a paper in the Big Data & Innovative Financial Technologies Paper Series. If you do that, however, you are not qualified by disposition to run a quant fund.

One suspects this phenomena, which is rampant in the investment arena, is going to get some attention in the form of litigation or regulatory scrutiny. In the meantime caveat emptor. A different group is coming to the same conclusion:

"We argue that most claimed research findings in financial economics are likely false."

Remember, kids, actual results may vary!