We use cookies on this website in order to provide you with enhanced user experience.

How To Measure Risk Tolerance

All fields of human endeavour use measurement in some form, and each field has its own set of measuring tools and techniques. Measuring a psychological attribute such as risk tolerance involves particular challenges because, firstly, there is no physical manifestation of the attribute and, secondly, there is no natural unit of measurement.

Since the 1890s psychologists and statisticians have been researching methodologies and techniques for psychological testing. The discipline they developed is known as psychometrics. Psychometric standards can be applied to questionnaires ranging from opinion polls and market surveys, through to tests of IQ, personality, aptitude, ability and the like.

Psychometric standards provide benchmarks against which a risk tolerance questionnaire can be evaluated*. A robust questionnaire is, in psychometric terms, one that is valid and reliable, where,

  • Valid means that it measures what it purports to measure, and
  • Reliable means that it does so consistently, with a known level of accuracy.

In brief, to meet psychometric standards, risk tolerance testing questionnaires must go through a rigorous development process comprising Usability trials and Norming trials.

  • In Usability trials, a large pool of questions is tested, for understandability and answerability, on representative samples of the population for which the test is intended. This can involve researchers sitting with subjects who are encouraged to verbalise their thoughts as they examine the questions. Questions which seem straightforward are often revealed to have poor understandability and/or answerability.
  • In Norming trials, questionnaires comprising questions with high Usability are tested on further representative samples and the results analysed statistically to determine the statistical value of the questions and the scoring algorithm. Questions which appear insightful are often revealed to have little or no statistical value in differentiating one respondent from another.

Typically, development requires multiple loops through both trial processes. A risk tolerance test that meets psychometric standards will display the following characteristics:

  • All questions should be directly related to attitudes, values, preferences, emotions or behaviour with regard to situations that involve risk. Questions that relate to the circumstances, e.g. stage of life or time horizon, do not meet this criterion.
  • Questions should address financial risk generally, not just investment risk.
  • Questions should be in plain-English. Financial terminology should be avoided.
  • There should be at least 20 questions in the questionnaire in order to obtain the statistical accuracy required.
  • The results must be scored on a Normally distributed scale, i.e. a bell-curve scale.

Additionally, the test’s publisher should be able to provide evidence that it meets psychometric standards.

The risk tolerance questionnaires commonly used in the financial services industry have typically been developed by compliance, marketing or technical services personnel without regard to psychometric disciplines. Rudimentary investigation will quickly establish that they do not test risk tolerance, let alone do so accurately … despite how they might be described by their publishers.

Concerned about "Psych Tests"?

There is still some controversy associated with psychological testing.

  • With IQ tests there is concern that the tests may be culturally biased and that they test the ability to answer test questions rather than an underlying, generally applicable, intellectual ability.
  • With personality and aptitude tests, many are disconcerted by the frequent lack of an apparent connection between the questions asked and the conclusions drawn. Typically, the questions are used to assess where an individual fits into a pre-existing model of behaviour and the report is then couched in terms of that behavioural model. Often the respondent is not aware that this is the process being followed and is (understandably) upset when they see negative conclusions being drawn about themselves from answers they gave to seemingly innocuous questions. Respondents can feel that they have been tricked into disadvantaging themselves.

Collectively, these concerns can be seen to be about the validity of the test, i.e. whether it measures what it purports to measure, and about fair play. Such concerns are not relevant to risk tolerance tests such as FinaMetrica's.

  • There is no (hidden) behavioural model in the background. Rather, answers are compared statistically to a database of answers given by a sample group of respondents and the report is derived directly from the individual’s answers to the questions.
  • The questions themselves have high ‘face’ validity, i.e. they are about issues where the connection to risk tolerance is obvious.
  • There is no indirect purpose to the test where the respondent could experience a negative outcome, e.g. the results will be used in evaluating job applicants. Here, the objective is simply to obtain an accurate assessment of your risk tolerance in a way that will be meaningful to you and to those you choose to share it with.

* A detailed discussion of the application of psychometrics in risk tolerance testing can be found in “Some Guidelines for Financial Planners in Measuring and Advising Clients About Their Levels of Risk Tolerance”, Callan V J and Johnson M, Journal of Personal Finance, August 2002, pp 31 - 44.

View the Callan and Johnson Paper