The unsinkable Titanic...

Since my last post on behavioural finance was such a hit, I thought I would share a gem from a recent case study.

A common trait among individuals is to be over-confident of their abilities. For example, the majority of individuals would say that they are safer than average drivers, which is of course statistically impossible.* This is also observed amongst investors; they are overconfident of their ability to pick successful investments. This can often be seen in over-trading amongst retail investors - rather than investing in long-term value they are trying to pick short-term investments. This can lead to much higher frequency of trading, which reduces net returns because of the additional transaction costs incurred. (Interestingly, this seen more prevalently among men than women - although both do it. Also, the data shows that for men, the tendency to overtrade is reduced after men get married.)

Although the above uses an example from retail investors, overconfidence can also be seen among institutional investors. It's so prevalent, in fact, that some fund managers deliberately set out to trade upon it. For example, a recent case study I've looked at focused on the asset management arm of a big Wall Street bank which set up a number of funds specialising in using behavioural finance. This an excerpt from the comments of the head of the behavioural finance team:

"Our approach, which forces our funds to systematically overweight value stocks, means that our investment behaviour is changed. We are forced to focus on out of fashion stocks that we wouldn't naturally have bothered with and that means that we cannot fall into the same overconfidence trap."

This amused me.

---

* I'm not sure that this is "irrational". Rather, people base their assessment of the likelihood of their having or causing an accident on their past experience. They may never have had an accident, or any accidents sufficiently long ago for them to be discounted - e.g. "I've grown up and matured since then". I expect few individuals know what the average likelihood of having or causing a traffic accident is (I don't). Also, this is an example that can be misused; I've often seen it phrased as "better driver", but the measure itself is then subjective - different drivers might rank on different measures - e.g. ability to drive at speed, likelihood to have an accident, ability to offer a smooth ride, or the extent to which they are law-abiding.