Hey Andrea, thanks again for your reply.

I agree with you on Musk.

Yes, of course, the data within a price chart is insufficient to predict future price. That has always been an exercise in tea-leaf reading - technical analysis of price charts is no better than checking your horoscope. But then again, calling a stock price chart completely random is also not correct. The day-to-day movement is mostly noise, but there is signal in there. It is just that we are not sophisticated enough to know what data in the broader world provides the strongest correlation with a particular stock's price move.

But that, in a nutshell, is exactly the kind of problem ML can begin to unravel, is it not? From a mass of what appears unrelated data, finding the correlations that best predict a given phenomenon - in our case stock price. In other cases, identifying a nascent tumor that even trained radiologists miss. Finding black holes that trained astronomers miss.

SD