Of Sex and Suitability

May 12, 2022
Greg

Greg

Globally recognised expert in applied decision science, behavioural finance, and financial wellbeing, as well as a specialist in both the theory and practice of risk profiling. He started the banking world’s first behavioural finance team as Head of Behavioural-Quant Finance at Barclays, which he built and led for a decade from 2006.

Building investment solutions for multi-dimensional people, not blunt categories

There’s a conflict at the heart of investment suitability: between solutions simple enough to scale, and the complex humans they’re meant to serve.

When matching investors to suitable investments, some form of categorisation is essential. To avoid being suffocated by complexity, investors (and investments) need to be grouped together.

However, to encourage people to get invested and stay invested, the investment experience needs to speak to each person as an individual.

This is where profiling comes in.

Until recently, the investment industry operated with a blunt reliance on demographics: for example, automatically lowering risk for older people and women. Such pandering to stereotypes makes providing recommendations easier, but delivers the wrong solution to those many older and female investors who have high risk tolerance (or risk capacity).

Increasingly sophisticated psychometric profiling tools have helped the industry move on from this, albeit neither very far, nor very fast.

Because of their historical dominance, and because demographic groupings are so easily grasped, we often see a tendency to want to interpret psychometric profiling in terms of demographics. At Oxford Risk we’re often asked how our financial personality scales relate to age, gender, or culture.

Do men or women typically exhibit certain personality traits?

Yes, they do. But only slightly. And the more important question is: so what?

Beyond pink and blue

Perhaps the categorisation that most often influences investment decisions is gender.

The way an adviser attends to a client will shape that client’s investment experience. Subtle beliefs grounded in part on ‘average’ assumptions, such as that female clients are going to be less confident, and more risk-averse, can be harmless most of the time. But sometimes they lead to unsuitable advice. And, more to the point, they’re always unnecessary. Because, as we’ll see in a minute, there’s a much better way.

At Oxford Risk we have robustly measured the financial personalities of thousands of investors worldwide on some 20 dimensions. Some are quite general, such as risk tolerance, or composure. Others are more specific, such as eight which relate to sustainable investing.

We therefore have the data to demonstrate how financial personality differs for different groups.

For example, it is commonly believed that women are less risk tolerant than men (i.e. less willing to accept greater long-term risks in exchange for larger potential rewards). This, it turns out, is true. But superficially so.

On our measure of risk tolerance, women on average score close to dead centre: 3.03 on a 5-point scale. Men average 3.23, statistically significantly higher.

However, there is nothing practically useful you can do with this information. Any adviser reckless enough to let this nugget influence their advice would have a very high chance of getting things badly wrong. The variation of risk tolerance between the two populations is completely dwarfed by the variation within populations of men, or women. Look at the chart below where the dark bars are the centre of the scale – there are more men above this mid-point than women, but this average difference is slight compared to the big spread within each gender; it is very easy to find very risk-averse men… or risk-tolerant women.

On many aspects of financial personality, men and women don’t differ at all. Even on those where they differ the most on average – the most notable being confidence (3.27 for men, 2.90 for women) – using that information to tailor an investment or communication strategy would be to cut the cloth with rather blunt scissors.

Yes, the data tell us that women on average have lower confidence in investing, and this contributes to the costly fact that many more women than men remain uninvested, or with much higher cash balances. But to advise all women as though they lack confidence in investing (or all men as if they don’t) would be grossly inaccurate and inappropriate.

Statistical significance and practical significance are not the same thing. And ‘on average’ insights don’t lead to suitable advice at the individual level.

Whether an investor is male or female is easy to measure, and because on average, sex differences in investment personalities do exist, it feels like this should matter. Yet in practical terms, it’s irrelevant.

While it may not always be as obvious as painting one portfolio blue and another pink, because it’s easy and feels important, and because more sophisticated alternatives are not always well understood, male and female investors are often treated differently regardless.

So what can we do with this information? What does the more sophisticated way look like? And how should it be applied? How do we target behavioural interventions at importantly different categories of investor, such as men and women?

Behavioural clusters beat isolated ‘insights’

There is a common, but subtle, problem with the way behavioural ‘insights’ are typically applied.

The need for behavioural profiling arose because the investment industry could no longer pretend that its clients were robots – mechanical automatons that you turned on, stuck a theoretically optimum portfolio in the slot, and then trusted them to operate as the textbook intended, i.e. like a machine.

However, when behavioural techniques are bolted on, rather than built in, to suitability processes – for example partnering a given behavioural quirk, say ‘loss aversion’, with a ‘tip’ for how to deal with it, say, ‘look less frequently’ – investors aren’t treated as humans, but merely different sorts of machines. Whole machines are built up from parts. Humans start off whole. To build them up one bias at a time fits better with an infographic than it does with reality.

We should use behavioural profiling to recognise investors as individuals, but also use common constellations of preferences – statistical clusters of attitudes that tend to be found together – to recognise that there’s value in categorising populations not on blunt demographic variables, but on more refined attitudinal and psychometric ones.

As we said at the start, some form of categorisation is essential to guide complex investors with inevitably – and valuably – fuzzy and flexible goals towards practically scalable solutions. But when a high-definition map is available, it makes little sense to try to navigate with the one drawn in crayon.

Uncovering clusters of preferences opens opportunities to personalise messaging (and in some instances the investments themselves) to common emotional needs. Especially when it comes to marketing messages aimed at giving people the comfort to become investors, rather than speaking to those that have already signed up.

Some of these groups will naturally have higher proportions of women, and some of men, so in this way women and men will be engaged differently. But it will be because of their relevant investment preferences, not the irrelevant assumptions associated with their sex.

Four types of investor

Running a cluster analysis on the many thousands of investors that have completed Oxford Risk’s financial personality assessments, we identified four common investor archetypes.

(Of course, we could subdivide these four groups further to arrive at more and more specific clusters, but doing so risks sacrificing practicality for precision, and even this simple division into four shows clearly how investors are likely to respond differently to certain key messages.)

The chart below shows the financial personality signature for each of these investor archetypes across four of the key financial-personality dimensions used in the analysis: confidence (the dimension showing the greatest gender differences); impact trade-off (i.e. willingness to balance financial with ESG outcomes); familiarity preference (need to hold familiar assets); and composure (ability to cope with volatile investment journeys). For each dimension the archetypes are ordered from the most female group (Group 1), to the least (Group 4).

  • Group 1 (29% of population; 64% female): low financial confidence; low ESG preference; high composure.
  • Group 2 (27% of population; 47% female): low familiarity preference; low ESG preference; low composure.
  • Group 3 (14% of population; 45% female): high financial confidence; very high ESG preference; high composure; high familiarity preference.
  • Group 4 (29% of population; 41% female): very high financial confidence; high ESG preference; very low composure.

Group 1 is strongly tilted toward women, whereas Group 4 is more male, with the other two groups in between. This unsurprisingly correlates with the average confidence scores, which you’ll recall from above is the trait that most strongly differentiates men and women.

There are other strong differences between the groups. For example, Group 3 is likely to be much more responsive to familiarity-based narratives, or to ’home-biased’ portfolios tilted towards local assets. Group 4 has low composure and is likely to need more handholding to control their behaviours in times of market turmoil. And Groups 3 and 4 are more likely to respond to ESG-based narratives.

The fact that Group 1 is the ‘female’ group should not be taken as an indication that women won’t respond to an ESG narrative. After all, 45% of Group 3, with the strongest ESG affiliation, are women. And indeed, 36% of ‘low confidence’ Group 1 are men. This emphasises the importance of focussing targeted engagement on behavioural factors, and not just blunt demographic distinctions.

Of course, not everyone fits equally neatly into these four clusters, but with effective profiling tools and digital engagement platforms, narratives and behavioural interventions can be hyper-personalised, ensuring maximum effectiveness in improving investing behaviour.

Read PDF

Related Posts

This is the fourth post in a series giving our response to the FCA’s Call for Input on how to apply behavioural finance to help people make engaged investment choices more comfortably and confidently, and what role regulations can play in helping that to happen.

Read More

This is the fifth post in a series giving our response to the FCA’s Call for Input on how to apply behavioural finance to help people make engaged investment choices more comfortably and confidently, and what role regulations can play in helping that to happen.

Read More

This is the sixth and final post in a series giving our response to the FCA’s Call for Input on how to apply behavioural finance to help people make engaged investment choices more comfortably and confidently, and what role regulations can play in helping that to happen.

Read More