More on Learning

Recently, I wrote about the importance of actively continuing to study and learn. Markets are very dynamic and even edges that persist change in subtle ways. Not updating your ideas in the face of new data is, at best, sub-optimal and generally just stupid and lazy.

But learning is hard. Sometimes it is hard in a trivial way. It can be hard because the subject matter is hard. Quantum field theory is a difficult subject and it is understandable that learning it won't be easy. However, learning is also hard for a more concerning reason. Humans don't change their opinions when confronted with facts, even when they directly contradict the opinion. 

As anyone who has read anything about behavioral finance knows, any Ph.D. student can do an experiment that shows reasonable-seeming people are often totally irrational. As great as it is to feel superior when we see others making these mistakes, we make them too, and they will impact our trading. We are great at spotting the holes in someone else's logic, but we are blind to our own faults.

How did we end up like this and what can we do about it?

Two scientists, Hugo Mercier and Dan Sperber, wrote the book "The Enigma of Reason" to explain their theory. Basically, irrationality is an evolutionary trait. On the surface, this seems unlikely but they make a strong case. It goes like this.

Humans’ biggest advantage is our ability to cooperate. Cooperation is difficult. Generally for an individual it is better to be a freeloader. We evolved logic not to solve abstract, general problems. It developed to solve the problems associated with living in cooperative groups.

Consider confirmation bias: our tendency to accept facts that confirm our current opinion and ignore those that contradict it. If reason is meant to help us learn, it is hard to imagine a worse flaw. But Mercier and Sperber argue that it has a more important function.

It is designed to win arguments. Sticking to a position and defending it in the face of contrary evidence is a survival skill (the people who are really good at it become political pundits or shills for trickle down economics). Think about the freeloader issue. Our Australopithecine ancestors didn't want to be the ones risking their lives hunting mammoths while others sat around the fire. It was more important to win arguments than solve logic problems.

(Incidentally, if you don't believe in evolution, you are a canonical example of someone who can't process facts that conflict with opinions).

What can we do to mitigate this effect?

The scientific method can be viewed as a way of avoiding this bias. There should be no way you get stuck with supporting your team. Results have to be reproducible in other laboratories, by researchers who have no motive to confirm them. Even when individual scientists make cognitive errors, the system prevails. We can't learn. Science does.

This is the solution but it probably won't help traders. With the secrecy in the trading industry, independent verification of results won't happen. Perhaps my solution is better: be anti-social and don't give any credence to anyone else's ideas.

Euan SinclairComment