This is Paul’s doctoral thesis, submitted in fulfilment of a Doctor of Business Administration at Victoria University, Melbourne, in 2020. The full 280-page thesis is available as a PDF below; this page summarises its purpose, method and findings.
The question
Why do investors choose one investment over another? When comparisons are readily available, the answer is assumed to be the one that provides the best expected return for the least risk – but when comparisons are difficult, due to inadequate or confusing data, how do investors actually choose?
The thesis examines how Australian non-professional investors select between three common investment options – superannuation, Australian shares, and residential real estate – to determine whether investors hold a rational view of each option’s characteristics, and whether that view then drives rational or irrational investment behaviour.
Method
An online, quantitative survey was conducted using a random selection of candidates sourced through social media and an Australian Financial Planning Association. Usable responses were collected from approximately 280 respondents, and analysed using descriptive statistics, correlation testing and logistic regression.
What it found
The sample was financially literate, yet showed a poor understanding of the relationship between risk and return as measured by a mean-variance approach. More than 60% of respondents misjudged which of the three investments had actually performed best in the past – yet perception of past performance turned out to be the single most important factor driving their investment choice, more influential than either their expectation of future returns or their perception of risk.
The consideration of risk also didn’t follow the traditional finance model of higher risk / higher return – for many respondents, the opposite view held: an expected high return felt low-risk, and an expected low return felt high-risk.
Recommendations
The thesis closes with three practical recommendations for the advice profession:
1. Risk profile questionnaires may be measuring the wrong thing. If an investor’s perception of risk doesn’t track the traditional mean-variance view, advisers should investigate that perception directly at the individual level, rather than relying solely on a standard questionnaire.
2. Past-performance warnings should extend to residential real estate. Regulated investments already carry a warning that past performance isn’t indicative of future returns; given how strongly past performance drove decisions in this study, unregulated options like residential property arguably need the same warning, and a stronger one.
3. Investors need a forward-looking reference point. Because perception of past performance outweighed even investors’ own expectations of future performance, the thesis suggests a centralised, independent source of expected future returns across major asset classes would give non-professional investors something other than the rear-view mirror to base decisions on.