Why it's hard to pick a winning fund

No longer needs a mutual fund.

Ted Cadsby, a consultant who used to run a mutual fund company, uses mutual funds to point up a problem in our thinking.

Why are there so many funds? If one got it right all the time, there would be no need for others. The others are kept in business because of the way we think.

The problem lies in how our minds are geared to underestimate the role of randomness in our lives — the profound but hidden influence of the uncontrollable and unpredictable.

Here's the root of the problem: To pick a winning fund, you need to identify information signals that are buried in a lot of random noise. Three layers of noise, in fact: noise in the company's reported results that make its stock value hard to assess; noise in the manager's track record that obscures skill; and noise in the stock market that conceals the conditions under which the manager's particular stock-picking style is effective.

A whole industry is based on the assumption that relevant signals pointing to a good fund can be screened from all this noise, Cadsby says.

The precariousness of this assumption is revealed by all the "winning" funds that eventually underperform. And the size of the industry is testament to just how prone we are to overinterpret meaningless, random noise (e.g., luck) by reading patterns (e.g., skill) into it and thereby misinterpreting complexity (e.g., "Ten years of outperformance can't come from luck alone so this fund is a good investment").

So what's this got to do with being eaten by a tiger?

We are naturally predisposed to overinterpret because we assume that virtually everything we perceive has inherent meaning. We do so because the risk of underinterpreting (known as a Type II error) is greater than the risk of overinterpreting (a Type I error). Say you're walking in a forest and hear the crack of a twig behind you. You can play it safe by assuming you're being stalked — thereby avoiding a Type II error while making yourself vulnerable to a Type I error. Or you can assume the sound was just caused by random rustling of the wind — thereby avoiding Type I error at the risk of being eaten. From a survival point of view, it's better to be inconvenienced by a Type I false alarm than killed by a Type II missed threat; i.e., better to overinterpret than underinterpret.

Suppose I'm not in the market or a jungle. So what?

Consider some "random" examples. Relationships of every kind are riddled with overinterpreted noise. How many angry exchanges are triggered by extraneous comments or fleeting looks, many of which are attributable to passing moods? By understanding that random expressions of irritation are often not obscuring important signals, we can do a better job of ignoring them, making for much smoother interactions — especially at home!

Here's the lesson.

When randomness runs deep, our best strategy is sometimes to simply accommodate it, as in the proven strategy of investing in index funds that buy all stocks as opposed to trying to pick the "best" ones. Granted, anything that feels remotely akin to surrendering to randomness is hard for us: Our deep-seated need to feel in control motivates us to read something into nothing.

That's my advice to readers of this blog: don't read something into nothing.