[jgh]

I've really enjoyed playing with Quantopian, and I think that I will continue to play with it as an educational tool. However my algorithms so far have been very good at losing an awful lot of money, so I think I will be sticking with simulations for now :)

[lucisferre]

> my algorithms so far have been very good at losing an awful lot of money

But... the solution is just so obvious, just make the algorithm trade the opposite of what it would have done.

Pure profit!

[nick_dm]

A few years ago I heard a story from an acquaintance working at an online spread betting company. One senior figure suggested that they should actively bet against their worst performing customers rather than hedging, they discovered pretty quickly that this wasn't a good idea!

[seunosewa]

What happened when they tried it?

[xerophtye]

haha! So... if the algorithm thinks a stock will rise, go short instead and vice versa?