Most of us who have been taught finance over the past 30 years have been have been schooled in a body of ideas known as ‘modern’ finance theory. This posits that investors are hard to fool and make decisions in a highly rational manner as they go about maximising their wealth. Their collective endeavours lead to ‘efficient’ markets, which are difficult for anyone to beat on a consistent basis. These arguments often form a major part of the justification for index tracking investment strategies.
Almost since these ideas were developed in the 1960s, though, researchers have documented a large number of anomalies – examples of real life investor and market behaviour that don’t appear to square with the theory. For instance, many investors trade actively and market volumes are high, but theory says buy and hold is best. Numerous studies suggest simple strategies – such as buying value stocks, or stocks that have done well in recent months – offer ‘beat the market’ potential.
Since the mid-1980s a lot of the attention devoted to explaining these anomalies has focussed on behavioural factors. What if investors aren’t the hyper rational information processors the economists assumed them to be? If they have biases and blind spots that affect their decision making, maybe markets won’t be quite so efficient after all. Hersh Shefrin’s recent book ‘Beyond Greed and Fear’ (Harvard Business School Press, 2000) gives an excellent overview of this growing field of ‘behavioural finance’.
Many applications of behavioural finance look at the implications for private investors dealing in stocks and shares, or at the behaviour of portfolio managers. But the decision-making biases and traits documented by experimental psychologists don’t just affect internet day traders. They represent some quite deep-seated aspects of the way most of us make decisions. Trustees and others responsible for occupational pension schemes are tasked with making some very important investment decisions. They need to be aware of these potential behavioural pitfalls.
Overconfidence is one of most commonly documented aspects of human psychology. Put simply, most of us have a rather inflated view of our own knowledge and ability in everyday situations. One study of US college students found over 80% of them rated their own driving skills as being in the ‘top 30%’. My own test with an undergraduate finance class found 87% of the students expected to finish in the top half of the class.
Trustees need to think about how this affects them and the people they deal with. They need to remember the future is highly uncertain and that prediction is no easy task. Is the investment manager really likely to outperform to the extent he claims? Is the consultant truly able to separate skilful fund managers from the merely lucky, or to specify the appropriate strategic asset allocation to the nearest percentage point? Despite all their skill and experience, investment professionals can still be overconfident.
Overconfidence leads to under diversification of investment portfolios. Why should I hold a broad spread of stocks when I am highly confident – even ‘sure’ – I know which ones are going to do best? Those with investment responsibility need to make sure the mandates they put in place provide for adequate diversification, for example by setting clear limits on active risk and deviations from benchmark weightings. The recent court case between Unilever and Merrill Lynch provides food for thought in this regard.
Psychologists also find we tend to use simple rules of thumb – known as heuristics – to understand the increasingly complex world. We look at a situation and try to work out if it resembles something we have seen before. Several studies document the strong belief amongst investors that the shares of high profile, fast growing companies represent good investments. This seems a likely prospect, but there is a lot of evidence that these types of shares are often overpriced and deliver poor subsequent returns. Maybe there is an analogy with trustees appointing investment managers with table topping past returns. This seems like a sensible idea – backing the past winners to continue doing well – but there is relatively little evidence of persistence of good performance.
Regret is a very painful human emotion. We often make decisions in such a way as to avoid being put in a situation that could cause us regret. We hang on to losing investments to avoid locking in the loss and having to come to terms with our bad decision. We fear that if we sell the stock it might rebound and leave us feeling doubly sorry about having missed out on the recovery. We take profits on winners in case the stock falls and our gains fade away. Similar emotions probably come to play when trustees review the performance of their active investment managers. Do we give a long time underperforming manager ‘a couple more quarters’ to avoid locking in the loss and the possibility of regret if performance subsequently improves? Anyone involved in the initial decision to appoint the manager will have a strong desire to see things work out well. Perhaps the decision about whether to continue the mandate is better taken by others who can look at the facts on a more dispassionate basis.
Behavioural finance has one of its most powerful and important applications in terms of the investment decisions made by members of defined contribution (DC) pension schemes. Individuals – from the factory floor to the boardroom – are charged with making complex decisions that will have a major impact on their standard of living in retirement. Many may not be well equipped to do this. As more and more firms baulk at the cost of defined benefit plans, this is set to become an increasingly important social policy issue. The challenge is to design systems and support networks that mitigate the most dangerous decision making biases.
American academics Schlomo Bernatzi and Richard Thaler document the ‘myopic loss aversion’ shown by members of the ‘401k’ DC plan at the University of Southern California. When shown annual returns data for US stocks and bonds, the scheme members on average opted to allocate 40% of their contributions to the stock fund and 60% to bonds. A similar group shown the same underlying data compounded to 30-year returns made an average allocation to stocks of 90%. The occasional down years shown in the annual data made the first group nervous of high equity allocations, but compounding the returns to 30 years allowed the second group to give more weight to the prospect of long-term outperformance. Psychologists often find the way information is presented, or “framed”, can have a significant impact on the decisions people make.
Bernatzi and Thaler have also documented 401k plan members using very simple – or naïve – diversification strategies. Given two funds to choose from many members put 50% into each. This result is relatively insensitive to whether the choice is between a stock fund and a bond fund, or between a stock fund and a balanced fund, despite the different asset allocation that results. The authors refer to this as the 1/n heuristic. They note the tendency for members of plans that offer several stock funds to have much higher average allocations to equities than members of plans which offer only one stock fund and one bond fund.
The approach is not limited to unsophisticated investors. The study’s authors quote Harry Markowitz – a Nobel economics laureate and arguably the founder of modern portfolio theory: “My intention was to minimise my future regret. So I split my contributions fifty-fifty between bonds and equities.”
There is nothing to say 50:50 isn’t an efficient investment policy, but it is clear the decision making process isn’t paying much attention to the complications of portfolio theory and risk management. If Harry Markowitz is making decisions with the aim of avoiding regret, there
is little to suggest the average employee is well placed to decide the appropriate investment strategy for his or her retirement fund.
Bernatzi’s study shows that where workers have the choice of investing their pension fund in their employer’s own stock, the average allocation to this option is 42%. From a diversification point of view this looks like a disastrous strategy given their jobs already depend on the company’s well being. We need only ask the former employees of Enron if we want testimony to this. ‘Reckless conservatism’, where loss aversion leads to high allocations to low risk/low return cash-like assets, is also a significant problem. Good pension scheme design, with sensible default options and ‘life styling’ arrangements, may offer some comfort, but improved financial education is probably also required.
Overall, behavioural finance provides a fascinating insight into how our instincts do not always lead to making decisions that are in our best interests. How can those handling pension assets deal with this?
It isn’t easy to override these deep-seated biases, even once you are aware they exist. There is no substitute for subjecting important decisions to disciplined processes and careful consideration. Trustees could also look to their consultants and investment managers for expert advice. But, we have to bear in mind that they, too, are only human.
Alistair Byrne is director of insured business at AEGON Asset Management in Edinburgh. The views expressed in this article are those of the author and may not be shared by AEGON. The studies cited in this article can be found at: richard.thaler/research