AI trading firms will be judged by their losses, not their demos
The interesting number is never the one on the launch slide
This week I interviewed the founder of a small lab that says it has built an AI that trades on its own. He was articulate, careful about what he claimed, and clearly believed it.
I have no idea whether he is right, and neither does anyone else, because the only evidence that matters does not exist yet: the record after a bad year.
That is not a criticism of him. It is the standard every trading firm should be held to, and it is a reason to be patient with the ones that publish their method and sceptical of the ones that publish only their best quarter.
Priya Nair is a Daily Meridian columnist