The economist has a good article on herding in financial markets, looking at US Mutual fund flows. It begins with this analogy:
IF YOU see a crowd outside a department store it is reasonable to assume that there is a sale. If you see a queue outside a bank there is a good chance that nervous depositors are trying to withdraw funds. In both cases the actions of other people send a signal that may be useful for others to follow.
And then describes a study commissioned to look at the potential returns from following the crowd, retail investors:
the crowd was more foolish than wise. The most popular sector, measured by the net inflow of money, had generally performed well in the previous year, beating the average sector by more than two percentage points. Investors were clearly chasing the trend. Alas, over the next 12 months that most popular sector lagged behind the average by just under three percentage points. On around 60% of occasions, the return of the most popular sector was much lower after investors bought it than before.
This ties in nicely with my article with Christoph Merkle and Martin Weber. We examine how individuals create their beliefs regarding other investors expectations, and the resulting effect upon investment behaviour. We find that second-order beliefs – our beliefs about others views – are surprisingly bad, and yet influence our investment decisions strongly. Market sentiment and the lagged 12 month return played a strong role:
We show that investors use their beliefs about the stock market expectations of others in their investment decisions. These second-order beliefs play a role beyond own risk and return expectations. However, second-order beliefs are inaccurate and exhibit several well-known psychological biases. We document these in a panel survey of active private investors, who are asked for their return expectations and their beliefs about the return expectations of others. First-order and second-order beliefs differ greatly and investors have only a vague idea what other market participants are thinking. Among the biases we observe is investors’ belief that their own opinion is relatively more common among the population. They further assert that others who hold divergent expectations are biased. We interpret these findings as evidence for a false consensus effect and a bias blind spot. The influence of second-order beliefs on investment decisions is mediated by the identified biases.