Most investors know that over 90% of the New York Stock Exchange and other exchange trading is done by robots, not people. These are not just electronic order matching systems for individual trades; those have been around since the 1990s. These are actual robots that make the trading decisions and decide when to enter trades with no human intervention at all.

These robots work off of large databases and use artificial intelligence to decide when to act. The AI is fed by news scanners that look for key words in particular sequences to interpret the news. This key word scanning and automated order entry work at microsecond speeds; much faster than any human could possible read the news and decide what to do.

This is why markets seem to overshoot in both directions. All of the robots read the same news at the same time and react the same way. The result is a rush and crush to buy or sell without any humans who are smart enough to sell into strength or buy into weakness to mitigate the damage.

As if that’s not bad enough, this article reveals that automation has just become a lot more dangerous. Not only do the robots read and act on their own, they’ve begun communicating with each other.

This article refers specifically to retail pricing but the same issue applies to stock trading. The robots are colluding and acting like a gang to push markets around. This amplifies their power and their “deep learning” algorithms reinforce this behavior so you get more of it with the passage of time.

This will all end in a one-way flash crash where there is mass selling, no buying and no way out of the abyss.

Institutional investors can schedule a proof of concept with the world’s first predictive data analytics firm combining human and artificial intelligence with complexity science. Check out Jim Rickard’s company at Meraglim Holdings to learn more.