In previous blogs we have discussed risk and what it really means. To recap, risk is the chance of something going wrong, which is the same as saying the chance of failing to meet our goals. In relation to our retirement, our (financial) goals tend to be:

  • Having a retirement income that will give us the lifestyle we want, up to some reasonable age (90, for example) and
  • Leaving an inheritance for our children (or not!).

Advisers are required to take risk into account when putting together a financial plan. To do this, a risk questionnaire is used to assess the client’s tolerance for risk. However, the questions tend to focus on variability in annual returns even though many clients have retirement goals spanning decades.

So, if an adviser wants to find an investment strategy that best achieves the client’s goals within their risk tolerance, they need to know the risk of a client failing to meet their goals as well as the client’s tolerance for variability in short-term returns (let’s call this “short-term risk”, even though it is not actually a risk unless the client has short-term goals i.e. one-year goals).

Consider the plot below:

Simulated Investment Strategies

Each dot on this scatter plot corresponds to an investment strategy. The vertical axis shows the retirement income that the investment strategy would produce and the horizontal axis shows “short-term risk” – in this case the worst annual investment return they might expect from the investment strategy (a value of 20 means a “worst case” annual return of -20%). Importantly, the retirement income shown for each investment strategy has 90% certainty* i.e. there is only a 10% risk of falling short of this retirement income. So the question for the client becomes “you could have $64,000 per year in retirement with 90% certainty* (top right of the chart), but you will likely experience some large negative investment returns on the way (eg a loss of 35% in one year), or you can have only $44,000 per year in retirement with the same 90% certainty* (bottom left of the chart) but experience a smoother ride, which would you prefer?”

By presenting the information this way the client is able to understand how much retirement income they give up in exchange for more stable short-term investment returns. What better way to help the client choose an investment strategy? Isn’t that better than selecting an investment strategy purely on the basis of how uncomfortable the client feels when markets go down? Of course, that is important too, and that is why it is included in the above plot (as the horizontal axis), but more information is required to make the best decision, including certainty levels and the trade-off between outcomes and “short-term risk”. So, what is wrong with the current approach to risk profiling? It may be causing some people to miss out on a better retirement because they don’t have the information they need to make the best decision on investment strategy, taking into account their unique circumstances and approach to risk. It reinforces short-termism, when planning for retirement is a long-term project.

Of course, some clients may want to use a different certainty level. We find that people near or in retirement tend to want to plan with more certainty – up to 95% – whereas those still a long way from retirement are happy to plan with a lower level of certainty eg 75% and in some cases 50%. Most planning tools show retirement outcomes with only 50% certainty, or the toss of the coin, although this is not usually stated in any advice. With Investfit, you can choose the certainty level for retirement income and you can set the maximum acceptable level of short-term risk and then find the investment strategy that produces the best retirement for the client.

Next blog we will talk about how advisers can improve productivity and therefore profitability.

*Certainty levels are based on models that simulate the future, using assumptions based on historic investment returns. All models of the future are necessarily subject to various sources of uncertainty themselves and results from these models should be understood in that context. To re-phrase Donald Rumsfield “there is uncertainty in the estimate of certainty”.