Economists at Maastricht University asked hundreds of clients of a big broking firm how much they agreed with the statement “the recent performance of my portfolio accurately reflects my investment skills.” Investors who had enjoyed above-median returns in the previous month were more likely to agree than those with sub-median returns. And the better the returns, the stronger was investors’ agreement.
This suggests that investors are prone to a self-attribution bias; we attribute success to our own skill and failure to bad luck.
For successful investors this is not always justified simply because share prices are volatile and so good returns even over long periods might well be due to luck.
To see what I mean, take three UK investment trusts: Fidelity Special Values, Schroder UK Growth and Invesco Perpetual Select UK Equities. All three have greatly beaten the market in the past six years, rising by between 50 and 80 per cent against the All-Share’s 12 per cent. All have positive Jensen’s alphas – that is, returns adjusted for their covariances with the market. But even this great performance isn’t certainly due to skill rather than luck. One measure of this is the p-value attached to that alpha; this is the probability of getting such an alpha by chance in noisy data if the true alpha were zero. For our three funds, these p-values range between 18 and 30 per cent. These are well above the 5 per cent level which is normally considered statistically significant. They tell us that even good performance over quite long periods could well be due to luck. For shorter or smaller out-performance - which is more common that these sorts of return - the odds of it being luck rather than skill are even greater.
So, a belief in our own skill might well be unwarranted. Worse still, it might be positively dangerous. Overconfidence can cause us to take on excessive risk in the misplaced belief we can spot good speculative investments. Or it might cause us to waste money on dealing charges as we trade more often. Or it might lead us to pay excessively high management charges if we believe we have the ability to identify good fund managers.
Overconfidence, then, is dangerous. And not just in finance; it might explain why England went from Ashes winners to humiliating defeat in the space of a few months.
Nor is this a new observation. Back in 1930, Joe Stalin wrote: “People not infrequently become intoxicated by successes; they become dizzy with success, lose all sense of proportion and the capacity to understand realities; they show a tendency to overrate their own strength.” He was describing the collectivisation of Russian farms – one of the most catastrophic policies of all time. Which should warn us of the huge potential costs of overconfidence.
