How to classify clients using psychometrics

Paul Resnik, director and co-founder of FinaMetrica, says planners must ensure they have a proper understanding of their clients’ risk tolerance before providing advice.

The vast majority of financial advisers use some form of ‘risk’ questionnaire with their clients; usually one which has been sourced from their software supplier, compliance department, or which they have developed themselves.

Recent FSA guidance however suggests that many advisers may not have completed an appropriate level of due diligence on such tests, which may leave them and their clients exposed.

A small number of advisers even claim that their ‘years of experience’ enables them to forecast an investor’s risk tolerance. Evidence of this was seen quite recently in BBC’s Panorama investigation, ‘Can You Trust Your Bank?’ (Screened on 13 June), where an adviser from a prominent high street bank claimed they could guess their client’s risk profile simply by ‘looking into their eyes’.

Now, while many would argue body language can be a useful part of the sales process, this type of ignorance – and the penalties which it might lead to – is clearly not acceptable.

Using psychometrics

Psychologists have spent more than 100 years developing psychometrics as a discipline for measuring characteristics such as risk tolerance, which in turn they have been studying for more than 50 years.

A psychometric test of risk tolerance will comprise a questionnaire, scoring algorithm(s) and a report. While there is ‘rocket science’ in psychometrics, it is in the background; the client and planner interfaces are in plain English and readily understandable.

Unfortunately very little of psychology’s expertise has made its way across to economics and finance, from where financial planning usually draws its academic advice. This has had dire consequences for the accuracy of financial planners’ estimates of their clients’ risk tolerance.

The problem with industry-standard questionnaires is that they have been arbitrarily constructed without any scientific discipline and contain too many ‘bad’ questions and not enough ‘good’ questions. As a consequence, their results are neither valid nor reliable; hence the FSA’s increasing interest in the subject.

Years ago it was not uncommon to find questions relating to physical risk tolerance in questionnaires designed to measure financial risk tolerance. Today, the more prevalent problem is that many risk tolerance questionnaires deal with financial matters that are not part of the construct of risk tolerance.

This is a legacy from the ubiquitous asset allocation calculators, which were designed to produce an asset allocation (or model portfolio) recommendation based on as few questions as would be acceptable in the sales process.

These tests generally include a potpourri of risk tolerance, time horizon, withdrawal expectations, and investment experience questions. The resulting reports have about as much integrity as an astrological reading from your favourite old aunty.

A number of studies have demonstrated that industry-standard questionnaires produce unreliable results. But perhaps the most telling was conducted by Dr Douglas Rice of Golden Gate University who tested 131 industry-standard questionnaires and reported:

“The findings show alarming results. For example, when all questions were answered in the most conservative way possible, the percentage of assets allocated to equities ranged from 0% to 70%. When answered in the most aggressive way, the percentage of assets allocated to bonds, or bonds and cash, ranged from 0% to 50%.