The problem with these types of systems are it's difficult to do backtesting on a single continuous data stream... I write about that here: https://blog.projec

The problem with these types of systems are it's difficult to do backtesting on a single continuous data stream... I write about that here: https://blog.projectpiglet.com/2018/01/perils-of-backtesting/ https://blog.projectpiglet.com/2018/01/perils-of-backtesting... I recommend doing something similar to the original post. Neural networks tend to produce valid looking output for a stock price easily, a random walk does too. You have to find correlations, causations, and then review the results carefully. For reference, I wrote my own financial advisor (not directly utilizing deep learning, that functions relatively well): http://projectpiglet.com/ http://projectpiglet.com/ It works relatively well when checking the "causality" which also has it's limits: https://blog.projectpiglet.com/2018/01/causality-in-cryptomarkets/ https://blog.projectpiglet.com/2018/01/causality-in-cryptoma... EDIT: Added some evidence: * 2016 stocks: https://imgur.com/a/j8YWR https://imgur.com/a/j8YWR * Early 2018 crypto: https://blog.projectpiglet.com/2018/01/30-weekly-returns-using-piglet/ https://blog.projectpiglet.com/2018/01/30-weekly-returns-usi... * Early 2018 stocks: https://twitter.com/AustinGWalters/status/976347632439209985 https://twitter.com/AustinGWalters/status/976347632439209985 * All 2018 on Robinhood only: https://imgur.com/a/2CxEFqI https://imgur.com/a/2CxEFqI * Also (I am lettergram on HN), created: https://hnprofile.com https://hnprofile.com and https://redditprofile.com https://redditprofile.com