Traditional questionnaires rarely yield an accurate picture of an investor’s appetite for risk. That is why wealth managers and financial advisers are turning to gamification and behavioural economics to delve into their clients’ psyches. The results deliver a far more accurate risk score, which enables better financial advice and keeps regulators happy, writes Tiphaine Saltini
It’s not news that economics is an inexact science. How could it be, when so much of human behaviour is erratic, illogical and ill-informed? Nobel laureate Daniel Kahneman was not the first economist to point this out, but, by applying the insights of psychology to economic theory, he played a key role in developing the field of behavioural economics, which knocked the “rational economic agent” off his pedestal.
Now many of the rules and assumptions about risk and uncertainty in the financial-services industry are also being questioned. Fintech companies have stepped up to the challenge and, inspired by Dr Kahneman’s work, are using machine learning and elements of game design to create more accurate risk profiles of investors.
Kahneman’s Thinking, Fast and Slow was an international bestseller. Writing in the New Yorker, one reviewer compared his influence to that of Darwin, “helping to dismantle a longstanding myth of human exceptionalism. Although we’d always seen ourselves as rational creatures… it turns out that human reason is rather feeble, easily overwhelmed by ancient instincts and lazy biases. The mind is a deeply flawed machine.”
One of his premises is that our confidence in getting things right systematically exceeds our ability to do so. In other words, we are surer of ourselves than we deserve to be. His work is scientific – statistical studies and peer reviews back it up. The conclusion is clear: if we want to make better decisions, we need to be aware of our biases and seek ways to neutralise them.
This is incredibly important when it comes to financial services and wealth management. Regulation demands that financial managers know the risk appetites of their clients and advise them accordingly; advisers risk big fines when they get it wrong. To date, risk profiles have been drawn from questionnaires, completed by the client, that require him or her to make judgements about their expectations, in life as well as their investments. Do they consider themselves to be risk adverse? Do they mind losing money? How much money will they need in retirement? These questions can only be answered subjectively – and, according to Kahneman’s conclusions, the chances are they will get it wrong.
Accurate risk scoring demands a much more complex approach. Advances in computing power mean that today we can do far better. At Neuroprofiler, we use gamification to assess a client’s attitude to risk, and we are getting far more accurate profiles.
Taking Kahneman’s Prospect Theory* as our starting point, we used machine learning to devise a five-minute test as a replacement for an hour-long questionnaire – and achieve the same level of accuracy. We asked investors to make choices – A or B – based on different scenarios. Their choices give us their risk score.
Our research suggests that our approach accurately assesses risk appetite. In a recent study, we created a number of fake investor profiles, from highly risk averse to very comfortable with risk, and asked people to adopt those profiles and take our test. Working blind we looked at the results and were able to match 85 per cent of them with the correct profile.
Our methodology has several advantages over traditional risk scoring. First, an adviser can assemble an accurate profile without any prior knowledge of the individual. Second, it satisfies regulators’ recommendations in Italy, France and the UK for behavioural science to be included in assessments of clients’ appetite for risk. Those who do so will be MiFID II compliant come 3 January, 2018.
Indeed, regulators are concerned that traditional risk-appetite assessment is flawed on many counts. Questions can be too complex, tangential or abstract to elicit an accurate response. Also, questionnaires are not taken sufficiently frequently to ensure that changes in circumstances are picked up and financial strategies modified accordingly. And in a nod to Kahneman, regulators acknowledge that self-knowledge is a misnomer.
Neuroprofiler’s analytics can spot inconsistencies between answers and flag them to advisers. This in turn allows wealth managers to follow internal policy to its conclusion – by asking the client to sign a disclaimer, for example.
When combined with other risk-assessment tools, such as those from Edgelab, which look at investment instruments, wealth managers and financial advisers can manage risk in ways that were just not possible before. Behavioural economics and analytics allow us to overcome our own natural biases and the quirks of the human mind to paint a far more accurate picture of a client’s appetite for risk.
Suddenly, investment planning and economics look more like sciences.
*Prospect Theory describes how people decide between alternatives that involve risk and uncertainty.
Tiphaine Saltini is co-founder and chief executive of Neuroprofiler