Decision Making in Groups or on Trustee Boards.

As it stands now more than $6 trillion in investments globally is overseen by investment committees of one form or another. In South Africa alone, investment committees are probably responsible for more than R1 trillion in pension fund investments alone.¹ So there must be something compelling about the model.

Essentially the “board” model of decision-making for pension funds evolved as a way to introduce good governance into the decision-making process.  Grouping together decision-makers who have diverse vested interests, diverse backgrounds and experiences, and who present diverse stakeholders of the fund, is believed to be the optimal way to control against conflicts of interest and to ensure that maximum brain-power was being applied to the process. 

Advocates of behavioural finance have been warning us for some time that individuals have flawed cognitive processes when it comes to investing. In fact, there are convincing arguments to suggest we are “wired” to make the kinds of errors we do. The logic follows, then, that a “group” of decision-makers would control against individual error by introducing counterbalancing views and by expanding the information and experiential base from which a decision is made.

Sound counterintuitive? If you don’t think this is true,  look no further than the abundant number of bad decisions that were taken by investment committees globally just before the 2008 financial meltdown. Clearly there is something going on here that demands our attention. 

Here’s the good news. According to Payne and Wood, as long as the task at hand involves straight problem solving of a purely technical nature, (such as formulating an accurate member profile for a fund), the group decision-making process can be quite constructive.  

But it’s when groups are faced with issues that demand judgmental inputs that results start to deteriorate rapidly. Decisions like: what’s our tolerance for risk? how should we formulate an investment policy? what asset classes should we include? and, which managers are best for the jobs? are all examples of judgmental decisions! Unfortunately these are the types of decisions that have the greatest impact on the fund’s success

What goes wrong when groups make decisions?

What goes wrong here?  Simply put, the group dynamic is a social one - where the primary activity is spent on determining whether one belongs or doesn’t belong, or in assessing (not necessarily consciously) how one might go about increasing one’s acceptance into the group, or, as sometimes happens,  establishing one’s separateness from the group.

It’s a phenomenon Barton Biggs neatly summed up in the title of his paper: “Groupthink = Groupstink”³.  He cites Yale psychologist Irving Janis when he says that Groupthink is a mode of thinking that people engage in when they are deeply involved in a cohesive in-group. In this context, maintaining cohesiveness is a more powerful motivator than getting to the right answer. It is this type of in-group pressure that leads to distortions in information processing, reality testing, and mental efficiency.

The dynamic of “Groupstink” builds upon many of the issues initially cited by the behavioural finance academics as impacting individuals.  But the group context adds a few of its own special twists to these phenomena: 

Illusion of effectiveness/Overconfidence

At the top of the list is a phenomenon that goes to the very heart of why we create decision-making groups: the illusion that a collection of good minds is a better bet than depending on one mind.  In the Wood and Payne study, 75% of investment committee members believed that the decisions they reached in a group would be better than the decisions they would make on their own. But when group decisions were put to the test by Terence Odean, the exact opposite was more the rule than the exception.

What happens is that, when more than one member of a group agrees to a specific point or when a particularly influential member of the group presents their views, the illusion quickly forms that that must be the right answer. In the spirit of maintaining group cohesion, often little prodding is required to get the rest of the group to migrate to this “right answer”.

This example of overconfidence becomes even more pronounced if the initial sponsor of the idea was the alpha, or dominant member of the group. Or, as the behavioural finance experts tell us we now have a dangerous combination of “overconfidence” being buoyed by “bias amplification”.

Bias Amplification/Leadership Sway

I have run a series of experiments with at least 30 different live audiences in South Africa to illustrate the power of bias amplification, and not once has the exercise failed in getting the audience to arrive at the wrong investment conclusions.

Construct a list of well known fund manager names, picking managers that represent recent performance extremes.  Along side this list, list a series of investment performances by one of the publicly available funds actually managed by each of these managers. Ask the audience to match the manager to the performance.

On the left hand side were the answers given by the audience. On the right hand side are the correct answers. Clearly the audience thought they knew the correct answer because they had pre-existing biases towards the different managers thanks to the power of media coverage and marketing.   

Ok- the example I gave here was for performances from 2004 – but I could run these tests today with similarly disparate performance numbers and get the same results.

But the really disturbing aspect of the exercise was not that we were able to trick the audience – in fact, there’s usually some bright spark in every audience who comes up with the right answer. Rather, it was the speed with which I was able to manipulate the audience into a consensus around the wrong answers just by the way I responded to their answers.  This is the dynamic that occurs in the group when there is bias amplification and leadership sway. In the investment committee setting, the magnitude of this overconfidence invariably emboldens groups into taking decisions that individuals in their own right would feel they lacked enough information or insight to make.

Sponsorship and confirmation bias / Knowledge bartering

Sponsorship and confirmation bias and knowledge bartering are reflected in situations where members of an investment committee may over-commit to an idea. In many instances this happens when members believe their “confirming” actions will help curry favour with other group members.  Remarkably, instead of introducing new information to expand on each other’s knowledge base, group members will introduce information if it helps confirm a powerful chairman’s view or withhold information if it helps curtail the power of another group member.

As Heath and Gonzales pointed out in their research of the problem:

1. Members have the perception that their information has increased or is more reliable when it is presented by group members.
2. Social agreement validates their decision (“and makes everyone more comfortable”).
3. When members are confronted with disagreement they don’t try to understand what the alternative view is all about but rather focus on trying to reconvince the other party.
4. Re-explaining arguments to the group tends to simply enhance confidence in the argument

Bottom line: Interaction with others increases decision confidence not decision quality.

As Biggs points out: “On an investment committee, it is almost better to be wrong with the group than to express a contrary view, even if it is right, because if by any chance you are both wrong and a dissident, you are finished as a functioning member of the committee.”

The Heart of the Problem

The challenge for most boards is that they see their primary mission as being to come to a consensual decision. That is why “removing discomfort” is a stronger motivator than “getting to the right answer”. The more homogeneous the group, the more likely the group is to reaffirm and accentuate that bias rather than work to dissipate it. Additionally, “members of cohesive long-term groups will tend to internalize the values of the group and avoid balanced decision-making.

Perhaps the most disturbing insight that neuroscientists have started to glean is the fact that the perception of social or competence ranking by members within the group can have a significant impact on how different members of the board are likely to contribute. For example, studies by a team at the California Institute of Technology (Caltech) showed that when members of the group began to feel that their contributions were not perceived as intellectually rigorous or valuable by the other members, this would trigger a dramatic downward spiral in future. Or, as the authors suggest, “we may joke about how committee meetings may make you feel brain-dead, but the findings suggest that they may make you act brain dead as well”.⁴

In truth, we are only just beginning to understand the full range of group dynamics at play. But the clear message should not be that the group dynamic is hopelessly flawed. Without question the starting point for change is to first recognize and properly label the problem. The next step is to identify ways to minimize counterproductive behaviour before it even starts.

Addressing the problem – the first baby steps

Decision Science experts seem to agree that the more diverse the group is, the higher the probability of controlling against counterproductive behaviours.

The problem is, most committees define diversity as being a function of social identity:  i.e. different ethnic or socioeconomic backgrounds.  From that perspective, South African investment committees would appear to have a distinct advantage in being able to assemble such a population - particularly as pension fund boards of trustees are required to draw 50% of its representatives from the employee ranks.

But what really seems to be the most effective way to ensure true diversification of thought is to identify people who tackle and process problems differently.

Perhaps the most easily identifiable way to ensure diversification of thought processing is to assemble teams that are equally divided between men and women.

Simplistically, these two populations bring to the table the type of diversification in group behaviour and information processing that is essential to an effective-working committee. The table below provides a broad summary of general characteristic differences – both the good and the bad:

Or, put another way:

But that advantageous starting point can be easily diluted if there are too many trustees in total (over 10 starts to become counterproductive because then consensus building becomes the exclusive activity.)

But beyond this starting point for enhancing committee effectiveness, there are other “behaviours” that investment committees can introduce and encourage that can contribute significantly to ensuring a more rigorous and fruitful decision-making process. In summary, these are:

• Create the time, space and culture for debate.
• Do not fall back on what you know, know what you don’t know.
• Take time as a group to understand likely cognitive errors.
• Push process planning before task planning – neutralise railroading
• Do not work on lowest common denominator principles.
• No “loafing”. Give everyone an area of responsibility – and make them deliver – induce intellectual fearlessness.
• Determine ahead of time what factors should predict success (low expense ratios, long term time frames, tight risk controls, managers with long tenure) and stick to measuring those only.
• Take time to go through a governance budgeting exercise so that everyone knows who decides what and where each member’s responsibilities begin and leave off.
• Define the default decision if other options can’t be agreed on.
• Make sure to create the time to get to the right decisions – even if it means buying for more time when you need it.
• Maintain good records of why decisions were made.

            And finally:

• Choose a leader with wisdom – not answers

References

1 Alexander Forbes Survey 2004.
2 John Payne and Arnold Wood, “Individual Decision-making and Group Decision Processes”, The Journal of Psychology and Financial Markets. Vol. 3. No.2. 2002.
3 Barton Biggs, “Groupstink” Strategy and Economics, Morgan Stanley Dean Witter, 4/5/99
4 Montague and Quartz “Implicit signals in small group settings and their impact on the expression of cognitive apacity and associated brain responses” Philosophical Transactions of the Royal Society B. 2012

Other sources:

“Are Women Better Investors Than Men?” Moneycentral, 3/10/1999.

Brad Barber, Chip Heath and Terrance Odean, “Good Reasons Sell: Reason-Based Choice Among Group and Individual Investors in the Stock Market”, Management Science Vol 49, No. 12 Dec. 2003.

Brad Barber and Terrance Odean, “Too Many Cooks Spoil the Profits: Investment Club Performance”, Financial Analysts Journa,l January/February 2000.

“Investment Committees and Behavioral Finance” Cambridge Associates LLC.

Chip Heath and Rich Gonzalez, “Interaction with Others Increase Decision Confidence but not Decision Quality

Irving L. Janis, Groupthink, Houghton Mifflin Company, Boston, 1982.

Michael Mauboussin, “Aver and Aversion” August 9, 2005.

Michael Mauboussin, “Investment Committees September 1, 2009

Read Montague and Steven Quartz “Implicit signals in small group settings and their impact on the expression of cognitive apacity and associated brain responses” Philosophical Transactions of the Royal Society B. 2012.

David Swensen, Pioneering Portfolio Management, 2002.

Arnold S. Wood, “Investment Committee Dynamics: Reckless Reflex vs. Rational Response”. Russell Endowment and Foundation Symposia Series. 2004.