Financial advisers show a huge variation in how they assess client risk tolerance with some putting the same 'imaginary' client at different ends of the risk spectrum, according to a new study.

A study of 200 financial advisers asked to make recommendations for notional imaginary clients found that risk tolerance assessments resulted in wildly different interpretations.

The net effect was a major difference in investment recommendations.

Assessment and recommendations were often "totally random."

The survey, carried out by fintech Oxford Risk, found that in one case an adviser recommended a ‘very low’ level’ of risk for an imaginary client while another recommended a ‘very high level’ for the same client.

For another, advisers were evenly split between recommending low, medium, or high levels of risk for the same imaginary client.

The behavioural finance study found advisers gave “remarkably different judgements” on how much investment risk was suitable for clients using the same hypothetical information.

Oxford Risk said asset allocations were “scattershot” and, even in cases where advisers agreed on suitable risk levels, they disagreed on what kind of portfolio to recommend to clients. 

Analysis of the results of the study, conducted in partnership with South African firm Momentum Investments and South Africa’s professional body The Financial Planning Institute, found that overall recommendations “were closer to totally random than totally consistent.”

The report 'Under the Microscope: Noise and investment advice' highlights the influence of ‘noise’ in the advice process driving variations which cannot be explained. ‘Noisy’ errors can be caused by irrelevant factors such as advisers’ current mood, the time since their last meal, or the weather. 

Oxford Risk is calling for increased use of technology and algorithms to help advisers deliver more consistent risk assessments for clients and avoid issues over assessments of risk tolerance and asset allocation. 

In the study, advisers' own characteristics were closely related to recommendations with university-educated advisers making lower risk assessments than average. Married advisers were lower risk than single advisers and those on salaries made higher risk recommendations than those on commission or fees. 

Greg Davies, head of behavioural finance, Oxford Risk said: “Identifying noise isn’t about eradicating inconsistencies. It’s about eradicating unjustifiable ones and evidencing justifiable ones.” 

• The report can be downloaded here https://www.oxfordrisk.com/blog-posts/introducing-noise-audits-for-financial-advice