Publications
Partners
Let’s try a totally different tack this time. No matter how many times investors are warned that past performance is no indicator of future performance, you, the investor – with much encouragement from the marketing departments of asset management companies – will still revert to past performance as a critical component in manager or fund selection.
We researchers cling to the notion that the more we educate investors with the right data, the more convinced our audience will become to the “rightness” of our argument. We can prove categorically – with all sorts of quantitative and qualitative metrics – that any belief that past performance is a good proxy for manager potential in the future is completely unfounded.
But here is the harsh reality: all that research will do little to change this old habit – investors, consultants and decision-makers will still insist on relying heavily on that past performance.
So, let’s take a leaf from a different branch of analytics: behavioural psychology. Let’s look at some interesting work being done with behavioural change for both substance abusers and patients who cling unhealthily to quack medical theories.
Here a new line of therapy is emerging that is having surprising success – Motivational Intervention (MI). These therapists start with the view that rational answers and scientific or statistical proof mean little to individuals where the flawed messaging of the group is far more alluring to individuals than rational answers and scientific or statistical proof.
The trick to prying individuals away from these harmful affiliations and belief systems is to shift emphatically to non-judgmentalism, while at the same time helping the individual develop a whole new value and belief system.
Instead of cowing addicts with facts about the perils of either taking addictive substances or that clinging to unscientific beliefs may be potentially harmful, MI works by establishing an individual’s level of motivation for change. MI proceeds by encouraging an addicted patient to articulate his or her own values, beliefs, and goals – independent of the “group” – until a shared approach for behaviour change is agreed upon between therapist (consultant/adviser) and patient (investor). In other words, we need to pay at least as much attention to changing hearts as to changing minds.
Let’s see if we can change your hearts as well as your minds if we tackle our past-performance-equals-skill problem by changing the narrative to one that a non-investment person could relate to.
Consider this wonderful example that Morgan Housel of the Collaborative Fund relates in the online article The Psychology of Money: From an Industry That Talks Too Much about What to Do, and Not Enough about What Happens in Your Head When You Try to Do It.
He tells the story of Grace Groner, a secretary with a meagre salary and modest living requirements, who manages to accumulate an astonishing $7m by the time she dies at the age of 100. Her secret: continual saving – even though small, is magnified many times over through the power of compounding. We could stop at that point alone and use it to persuade investors that they should stay invested. But this is not Housel’s point.
He then introduces another character into his story: Richard Fuscone, a former vice chair of Merrill Lynch’s Latin America division. Despite retiring early with a small fortune from his high-paying position, Richard ends up declaring personal bankruptcy soon after. The problem: overspending and bad choices in his personal property investments.
Housel doesn’t pass judgement on either investor. He uses the two stories to make this critical point:
“In what other field does someone with no education, no relevant experience, no resources, and no connections vastly outperform someone with the best education, the most relevant experiences, the best resources and the best connections? There will never be a story of a Grace Groner performing heart surgery better than a Harvard-trained cardiologist. Or building a faster chip than Apple’s engineers. Unthinkable.”
This example should give anyone, inexperienced or experienced, some pause for thought. We don’t need statistical analysis to see that we need to pay a bit more attention here to what would be driving these differential outcomes. More importantly, the anecdote neatly points out – without being judgemental – that differential outcomes in this “profession” may have little to do with “skill”.
But let’s move on to another “non-financial” approach that could be equally effective in prying our reader away from the herd preference for past performance. Again, as with any effective MI intervention, the trick is to move away from a direct attack on the industry and use a non-confrontationist technique to properly position any human endeavour outcome along a skill-to-luck continuum.
This is a wonderful exercise that was introduced by Michael Mauboussin some 10 or 15 years ago. Mauboussin looks at a whole spectrum of human endeavours that produce outcomes: running a marathon, playing chess, playing in the World Cup, investing, playing blackjack, spinning the roulette wheel. He then argues that, based on four questions (I always add a fifth), one can readily place that activity somewhere on that skill-to-luck continuum – including asset management.
Question 1: Can you intentionally lose in this endeavour?
Clearly one can’t intentionally lose at roulette. Outcomes there are just totally random. But you can intentionally lose a marathon – you just don’t put in the effort.
Now try asking a room full of asset managers whether they can intentionally lose in asset management. At first the answer seems obvious: yes, of course. And then you start probing and testing each time they insist that they can. You point out that every time something in the market seems undervalued (because everyone knows the company is a dog), some contrarian investor comes along and snaps it up. Lo and behold, it turns out that much of the time, everything people think should do poorly ends up doing well, simply because of this contrarian element. So no, asset managers cannot deliberately lose. Mr Market (the market psychology) makes that impossible.
Question 2: Does practice improve outcomes?
With marathon racing: yes. With roulette: no. With asset management? The surprising answer here is that a number of studies have tried to see if there is any correlation between length of experience, education, IQ, age, qualifications…and performance delivery. What the research has shown is that none of these factors indicate a consistent correlation to future performance.
In fact, the only study that showed any connection at all was one by Van Harlow at Duke University, where he showed that share traders with low levels of monoamine oxidase in their bloodstream turned out to have slightly better results than the average investor. Monoamine oxidase is a neuro-regulating enzyme. If you have a sub-par level of it, you likely don’t respond reflexively to risk – a good thing for a trader.
The problem with this research is that it also indicated that other populations such as substance abusers and addicts of all manner also happen to have these same low levels: they too don’t respond as normal people might to the risks they take.
Not sure what this all suggests exactly…
Question 3: Do the outcomes tend to revert to the mean?
With marathon runners: no. With roulette: yes. Eventually in roulette, because the outcome is randomly distributed, the losses will even out with the wins. And asset management? Let me use a very old chart of manager performance (just so I don’t offend anyone) to show you exactly how that same phenomenon plays out in the manager performance surveys (see p.24). Follow the arrows and you will see that asset managers who outperform during one period will tend to underperform in the next: reversion to the mean over time.
Question 4: Is there evidence of transitivity?
Transitivity means that if A can outperform B and B can outperform C then A should be able to outperform C. This works in a running race. It does not hold true in roulette – again, because the outcomes are random. And it does not hold true in asset management because different investment strategies work in different economic conditions and Mr Market never is that clear on which strategies will be rewarded at any point in the future.
Question 5 (this is the one I added): How many factors stand between you and the performance outcome?
In the case of the marathon runner, performance is generally a function of whether that person can deliver on that day. It’s all about their individual skill. In the case of football (which ranks above asset management in the skills continuum), it’s not just about an individual team member’s skill. It’s whether the team as a whole is able to function effectively. In the case of asset management, the manager needs to get it right on all these levels:
- They need to select the right securities;
- They need to identify the right sectors or common factors (e.g. size, value, momentum);
- They need to make sure the asset allocation responds both to local economic and market drivers as well as global drivers.
But the biggest challenge of all relates to something called the “transfer co-efficient”. That’s the percentage of manager skill that can be translated into a final outcome once the portfolio construction process is applied. Here is where Regulation 28 constraints and Collective Investment Schemes Control Act (CISCA) regulation limits are introduced. These regulations are designed to protect the unit trust or pension fund investor from any extreme exposures to risk. Diversification is at their core. As the chart suggests, even if the asset manager picked all their shares correctly – and as such, the expected performance of each of those shares by themselves was 7.3% – once those constraints are put into place that “perfect skill” now gets reduced to only 2.3%.
That should give us all pause for thought. Rest assured that the answer is not “remove the constraints”. The constraints exist because indeed, as our anecdotal discussions have subtly pointed out, we cannot determine whether asset managers, even if they are skilful, can translate that skill into performance in the future. As such, constraints protect the lay investor.
Now what?
Perhaps it’s no wonder that on Mauboussin’s skill-to-luck continuum, asset management features just a fraction above blackjack – after all, both performance outcomes (assuming card counting is forbidden) rely heavily on an individual’s ability to outguess what their competitor is likely to do. In the case of the asset manager, it’s Mr Market that you must beat – and that effectively means all the other market participants and what they think the market will do.
But does this mean that there is no role for asset management? The answer here is emphatically NO!
But it does mean we have to be much more circumspect about what we should expect from our managers. Consider this point – what if we changed the mandate for the manager to one that says: “Please manage this portfolio such that I have X% certainty that I can meet this “Y” funding requirement knowing that I have these cash flows and time frames to work with.” Surprisingly, that’s a far easier requirement for a fund manager than simply asking them to outperform other managers. That means that what we really require is simply a manager who maintains a steady hand on the tiller and doesn’t try to be the hero. Much simpler than the industry would have you believe.
Did we convince you?
Anne Cabot-Alletzhauser heads up the Alexander Forbes Research Institute.
This article originally appeared in the Collective Insight supplement in the 25 October edition of finweek. Buy and download the magazine here or subscribe to our newsletter here.
