That... doesn't quite work, logically? Observing that there is no predictable "signal" in stock price changes only proves that whatever you're using to make pr

That... doesn't quite work, logically? Observing that there is no predictable "signal" in stock price changes only proves that whatever you're using to make predictions is unable to beat the market (the "efficient market hypothesis"). Stock prices are very obviously not completely random, in that they tend towards $0 when companies go bankrupt. There is some randomness, or at least unpredictability, in stock prices. This is called "volatility". But the model operates on the very small tick-to-tick scale, not wether UBER is priced at $45 or $450.