AI in Trading: What Could Possibly Go Wrong?

by Tom Starke and Qian Zhu

Is AI trading reliable? Can AI predict stocks?

I came across an article Stock Price Prediction With LSTMs on Hacker News. I like Hacker News as it has a large and generally very smart audience; the articles there are often very interesting. This article was upvoted to appear on the front page of Hacker News, which was not easy due to the voting system.

Naturally, I was very intrigued by the title as I have been using LSTM for various applications in trading algorithms for a while. However, pure price prediction from the price series alone is hard. The article presents the following chart on its front page:

The moment you see this you can’t help getting excited, right? However, over the years I’ve learned the following mantra, sometimes the hard way:

When things look too good to be true, something’s wrong.

Luckily, the authors point to their Jupyter notebook on Github, which can be found here. After cloning the notebook I looked at the code and in a nutshell, ANNs (Artificial Neural Networks) were applied to the past…