Back in 1981 Amos Tversky and Daniel Kahneman told some experimental subjects about Linda. She was a bright 31-year-old who had been politically active as a student. They then asked: what is the probability A, that Linda worked in a bank? Or B, that she worked in a bank and was active in the feminist movement? More subjects said B than A. But this is logically impossible: it cannot be more likely that Linda was both a bank worker and active feminist than that she was just a bank worker. There's a name for this: the conjunction fallacy.
Stories, therefore, can skew our sense of probability. Back in 2013, a weekly investment magazine (not this one) published a mega-bearish report entitled The End of Britain. It was a good story, marred only by the fact that it was utterly wrong. If you want a guide to the future, the statement 'shares will rise a little because they usually do' is a better guide than most. But it's a lousy story.
And if you're looking for warnings of doom, simple statistical rules do at least as well as narratives. For example, you could have avoided the crash of 2008 by selling in May or by heeding the fact that high foreign buying of US stocks often predicted market downturns. Those were good predictors, but bad stories.
It's not just stories about the future that mislead us. When we talk of the past, we speak of causes and effects and so tell a story. But as Nassim Nicholas Taleb has warned us, this can lead us astray, because it causes us to under-rate the role of luck and so over-rate how much predictability there is in human affairs.
One way in which we do this is to infer from apparently successful results that good decisions were taken when in fact there might have been bad decisions that just got lucky; this is called the outcome bias.
For example, back in the 1980s and 1990s Equitable Life had a reputation as a great pension company by paying out well on with-profits policies. But then the company found itself having to pay higher annuities than it expected and so collapsed. The stories we told of Equitable being a well-managed pension provider suddenly didn't look so good. Investors who believed them lost fortunes.
A similar thing happens with 'growth' stocks. Investors often believe stories about exciting new technologies and so pay highly for 'growth' stocks. Such stories, however, distract us from the fact that, as Sussex University's Alex Coad has shown, corporate growth is largely unpredictable. In not appreciating this sufficiently, investors often pay too much for growth stocks.
The same thing happens in fund management. Good returns invite stories about how clever the manager is when in fact he might have just got lucky. Investors therefore pay too much for past performance which doesn't persist.
'We had some good luck, then some more, then some bad, then some good' isn't much of a story. But it is often a better description of life than better stories. Those better stories, though, can be systematically misleading.
If you want a good story, read Dashiell Hammett. Just don't let stories affect your investing.
