DataTau logo

DataTau

new | ask | show | submit
login
How to predict stock returns using an LSTM-based model (Keras Tutorial) (jackdry.com)
2 points by jdry1729 1427 days ago | web | 1 comment





“build an LSTM-based model to predict whether EasyJet's stock price will go up or down on a particular day, given pricing data from the past 30 trading days.” — snake oil. Stock prices are widely accepted to be non-predictable owing i.e. random walk. Even if this model could predict forward prices with any kind of tradable advantage, that advantage would very quickly be exploited and thus factored into the stock price.