Weekly Impressions: Behavioural traps to avoid - Overconfidence

It has been a good year of performance for many Australian fund managers, with the average active stock fund delivering returns well above the benchmark thanks in part to underweight bets in resources and the major banks.  Despite the out-performance, various structural headwinds continue to retard growth of inflows into active funds, including: the shifting locus of power towards asset owners, limited diversification benefits offered by a highly concentrated benchmark, and strong appetite for lower cost ‘smart beta’ products.

With the day to day portfolio responsibilities less pressing at this time of year, it represents a good time to reflect on some of the pitfalls associated with active management.  Drawing on the burgeoning field of behavioural finance, this post investigates the behavioural trap of over-confidence and addresses how portfolio managers can reduce the prevalence and costs associated with over-confidence. 

Skill or Good Luck?

Behavioural finance is the intersection of applied psychology and the behaviour of financial market participants.  Researchers have done a tremendous job in elucidating the heuristics or mental rules of thumb that investors (and broking analysts) rely on, which contribute to systematic errors in judgement and decision-making.  One of these heuristics is self attribution or over-confidence, where people attribute their good performance to skill or judgement, and poor performance to bad luck or factors beyond their control.

Although these errors are not unique to financial market participants, such errors can be costly, particularly when portfolio managers are managing billions of dollars.  If enough investors suffer from such biases they also lead to significant limits to arbitrage, which imposes costs and risks on other investors.

A number of widely cited studies in the field of applied psychology document the above average driver effect, in which the overwhelming majority of respondents believe that their driving abilities are above average.  Over-confidence or illusory superiority is not confined to self-perceptions of driving ability.  Many people over-estimate their abilities, knowledge and future prospects compared to other people.

Over-confidence and excessive trading hurts your portfolio performance 

Terrance Odean and Brad Barber have explored the effects of over-confidence in financial market settings. Over-confident investors over-estimate the precision and accuracy of their information, which lifts the likelihood that such investors trade too frequently based on their perceived superior information. The empirical evidence is not kind to over-confident investors; the more individual investors trade, the more they lose.

Odean and Barber use over-confidence to link gender with trading performance. Female portfolio managers have superior track records than their male counterparts thanks to lower churning of their portfolios because they are less prone to suffering from over-confidence. Individual investors are found to trade more after they experience high stock returns, which might explain why aggregate turnover increases after periods of high market returns. 

Shhh...Listen First, Speak Last

Portfolio managers that work in a group setting might draw some comfort from the fact that group decision making helps to address or neutralise the behavioural biases that individuals exhibit, including over-confidence. But group decision making can also be beset by psychological biases, notably groupthink. A head of equities with a strong personality who expresses his view forcefully can effectively discourage junior members of the team from freely expressing their own views that challenge his. In his auto-biography, Courage to Act, Ben Bernanke explains that he sought to reduce the prevalence of groupthink by speaking last at the FOMC meetings. 

Ironically, the evolution in the structure of the funds management industry might actually reward over-confidence. Fund of funds and asset consultants are increasingly shifting towards a core –satellite approach to portfolio construction, where index and smart beta products are combined with high conviction/high tracking error funds. These funds typically require PMs to have a high degree of confidence in their stock calls, and be willing and able to take large active bets.

In a recent paper, David Hirshleifer and Kent Daniel suggest that overconfidence contributes to the profitability of the betting against beta trade, which represents the poor returns from high beta stocks.  Overconfident PMs – particularly those working in high conviction funds whose clients expect high returns - are more likely to buy high beta or cyclical stocks to generate outperformance.  The strong appetite for such stocks lifts their price and drives down their expected future returns.

In praise of humbleness

The literature on over-confidence in financial markets suggests that PMs ought to be more humble and trade less. Even when investors believe that they have access to superior information, they should be cognisant of the limits and costs to arbitrage, which can cause mispricing to persist for an extended time. To reduce the likelihood of becoming over-confident in their own sectors, PMs should be encouraged and incentivised to seek out opportunities and challenge the views of PMs responsible for other sectors.

Heads of equities should encourage a collaborative and collegiate team environment, where members are not afraid to debate and express views that challenge the consensus view internally.  And PMs should document their views rigorously in real time, to ensure that over-confidence does not also morph into hindsight bias.