The $100M ETF Built on Research That Doesn't Replicate

The $100M ETF Built on Research That Doesn't Replicate chrisbrunet.com

“Fornell et al. didn’t run standard multifactor asset pricing tests; didn’t use standard t-tests; and didn’t conduct out of sample tests. When you do, there is nothing there.”

8 points•ashgreat•10 days ago•3 comments•

#didnt #tests #research #replicate #100m #etf #etf built #100m etf

3 comments

chews10 days ago

Classic case of Texas Sharp-shooting, painting bullseyes around the bullets.

anovikov10 days ago

"A sufficiently smart person will always find a way to overfit any model to get to result they want".

ashgreat10 days ago

Or "If you torture the data long enough, it will confess"

But this case looks like an outright fraud. The paper doesn't even show the strategy that led to very high returns. It just shows the results. And this was a peer reviewed paper in the top-4 marketing journals.

Read the full thread on Hacker News →

3 comments

Related stories