The study of money is the study of people, and people behave in sometimes predictable, but often unpredictable ways. Just when you think you have a rubric (like Nate Silver with elections, a related field), along comes a Trump who blows apart the whole model.

I’ve always felt that finance is a very qualitative discipline. You are no worse off hiring English or history majors. It’s no accident that all the heavy hitters in this business are also really great writers.

The quants are starting to catch on, and a lot of the algorithmic traders are writing programs to mimic and predict human behavior, though it’s really just technical analysis, trend following, in a computer program. Technical analysis has an uneven reputation, but when you can quantify and backtest it and it works, the reputation gets markedly better.

Hard to argue nowadays that even weak-form EMH holds, when you have a cottage industry of very profitable systematic strategies.

Of course, there is a lot of math behind the quant stuff, and the guys doing it are mathematical geniuses, but the best of them are also very sharp market folks, with a nose for when trades start to get crowded. The quant blowup of 2007 happened because all the smart quants were in all the same ideas. So even in the world of high-level mathematics, you still have to deal with unquantifiable stuff: human behavior.

When someone like hedge fund manager Bill Ackman sees his portfolio get slaughtered by about 20% in 2015 and then double digits in the first month of 2016, that’s not just bad stock-picking. This is what happens when crowded trades become un-crowded.

Computers may be computers, but the people who program the computers are just human, and utterly fallible.
Behavioral Finance

When I taught my college finance class last semester, I’d say the most consensus long among the students was Disney (DIS), because of Star Wars.

Of course, I had been doing a bunch of work on the short side for months.