Ever since I began in this profession around 25 years ago I’ve seen the standard warning that ‘past performance doesn’t relate to expected future performance’. While I understand that the variance in returns expected from many of investments is why we shouldn’t rely on historical returns, the actual influence has not really been tested.

A significant component of my research into investor decision-making involved determining the influence of past performance as compared to expected future performance.

Surely the expected future return of an investment would be more influential than its past performance?

Ed’s note: This is the second in a three part-series of columns Paul Moran is writing exclusively for Professional Planner based on his Ph.D. thesis exploring investor decision-making. Read the first contribution here.

The study considered non-professional investors of which around half identified that they had an on-going relationship with a financial planner or adviser. I was interested in researching the relative influence of the investor’s perception of past performance versus their expectation of future performance.

The study considered investor preferences between residential real estate, Australian shares and superannuation. Before you start yelling, I fully understand that superannuation is a tax structure and not an asset class but there is a good deal of evidence suggesting that non-professional investors see superannuation in this way. Remember that more than 80 per cent of superannuation investors have never changed their investment option away from the default.

To my surprise, I found that perception of past performance (by the way, very different to actual past performance as participants routinely offered incorrect assumptions about past returns) was significantly more influential in the choice of preferred investment than even the expected future performance.

In other words, if participants thought that an investment had done very well in the past, they were willing to overlook their perception of expected future return in making a choice. This was most pronounced when considering residential real estate as an investment. Those who thought that property had performed best over the past three years (it didn’t for the time-frame considered) were six times more likely to prefer that as an investment rather than Australian shares or superannuation.

This was reinforced by the answer to a simple question regarding past performance: “If an investment had performed much better than expected over the past two years, would you be more likely to invest, less likely to invest or neither more nor less likely to invest?” Only 6 per cent of respondents answered that they would be less likely to invest.

This is likely a framing effect, especially when we consider the almost continued rise in prices over the past 20 years. Tversky and Wacker call this ‘Cumulative Preference’ theory, where uncertainties (future returns) become more certain in the minds of the investor when there are periods of prolonged outperformance. They describe that an (investment) event has more impact when a possibility is turned into a certainty.

So, framing investment returns based on ‘certain’ past performance is likely to have a very substantial impact on investor preferences. Moreover, it becomes easy to turn the ‘expected’ future return into a certain thing in the eyes of the investor. This might explain why susceptible investors are so easily convinced to buy investments, especially property investments, even though the expected returns are hard to justify.

The research asked, essentially, the same question regarding preferred investment of the three described above framed as both past performance and expected future performance. It was interesting that participants tended to become very conservative when asked to estimate the future value of an investment and the heuristics often associated with property investment (double every seven years, for example) went straight out the window. Estimates averaged around 2 per cent to 4 per cent per annum for residential real estate growth but were closer to 6 per cent for superannuation and Australian shares.

Encouraging clients to focus on the future could be a useful tool when dealing with those who we identify as having persistent views of specific asset growth rates. Simply asking them to predict the future value in, for example, 10 years’ time and then working backwards to produce a rate of return can provide the reality check that is often needed.

There is other research I have seen that suggests clients are actually willing to pay more for advice with an adviser who is focussed on the future rather than the past.

Further, having actual past performance figures available may challenge the perceptions of non-professional investors who don’t have access to the same credible data that advisers and planners use.

Paul Moran is principal of Moran Partners Financial Planning and has completed a doctorate in behavioural finance.
3 comments on “The role past performance plays in investment decision-making”
  1. Avatar Dr. Paul Moran CFP® SSA™

    Hi Geoff, I guess surprised in an ‘academic’ sense as there is a dearth of research on the impact of past performance on decision making…

  2. Avatar Geoff Warren

    Really interesting work, Paul. The main surprise is that you were surprised by the results! I suspect you are right that there are ‘behavioural’ reasons behind why people based decisions on their perceptions of how investments have performed in the past. Raises the question of how DO you overcome this problem and get people to be forward rather than backward looking when forecasts seem less tangible that past returns? A topic worth digging into further ….

  3. Avatar Christoph Schnelle

    I have the same experience as Paul. We can tend to choose investments similarly to how we vote for people – based on their track record, not their value. I get a lot more queries during booms about investing than during busts.
    Real estate is a special case, though as you can reasonably leverage it almost 5 to 1 by borrowing 80%. In that case combined net income and capital growth returns of 4% become 20%.

    Add in that for many people it is easier to pay off a mortgage than to save and a preference for residential real estate in a country like Australia where previous real estate busts were hidden by inflation (1981-1987 for example) and the interest by many to invest into property makes sense despite the many dangers and disasters that are also present.

Leave a comment