13 July 2019
Longevity risk is important to think about in planning around financial futures — the risk of outliving our assets.
To make some sort of assessment about managing this risk, indeed about managing our investments and income, we need to be able to build a set of financial assumptions that provide an estimate of the future.
The key input into this process is the rate of return that is assumed for the investment portfolio earnings.
Given that this is effectively a forecast for the future, it is clearly not possible to know what that number will be — the challenge is in making the best estimate of the range of possible earnings that we can.
This estimate is especially important for people who might be contemplating a long retirement. They might retire in their 50s and live into their 90s or beyond.
Let’s consider a person who is now 60 and looking to retire.
Let’s assume they have a generous investment portfolio of $1,500,000. They know they will not receive any part age pension to supplement their investment income.
From an asset allocation perspective, let’s then assume they are targeting $75,000 per year to live on, and want to keep four years of living costs, $300,000 in cash and fixed interest assets, and invest the remaining $1,200,000 in growth assets (an 80 per cent growth, 20 per cent defensive split).
Historical data suggests an attractive rate of return from cash investments in Australia compared to what is currently available.
For example, from 1990 to 2018 the average annual return from cash investments was 5.5 per cent per annum (according to Vanguard), against an inflation rate for this period of 2.5 per cent, suggesting an after-inflation (or real) rate of return of 3 per cent a year.
However, for realistic expectations with the RBA cash rate at 1 per cent, and the best cash and term deposit accounts not providing a return significantly higher than this, a more realistic assumption for the long-run return from the cash and fixed-interest investments in a portfolio would be a real return of about 1 per cent a year.
Insuring for the future
Jeremy Siegel, Wharton finance professor and author of Stocks for the Long Run, has long been a proponent of maths that suggests the long run real (inflation adjusted) return from the sharemarket is 6-7 per cent a year.
Let’s make this an optimistic calculation of an expected return to start with, and use the 7 per cent real return.
We now have a real return of 1 per cent per annum from the 20 per cent cash/fixed interest part of the portfolio, and a 7 per cent real return from the growth (shares) part of the portfolio, to provide a total real return of 1 per cent x 20 per cent, and 7 per cent x 80 per cent, creating a total combined return of 5.8 per cent per annum.
The average real return from Australian shares over the period 1900 to the end of 2018 has been measured at 6.75 per cent per annum.
This is not much different from the Siegel figure and would provide a total real return for the portfolio of 5.6 per cent.
One note of caution worth sounding is that the Australian sharemarket returns have been world-leading over this period (1900 to 2018).
The average return from world markets, calculated for the Credit Suisse Global Investment Returns Yearbook, was a real return of 5 per cent a year.
In our range of scenarios, it is worth considering the possibility that future Australian returns will be similar to the average world return.
For a portfolio that is made up of 20 per cent cash and fixed interest investments, and 80 per cent growth investments, the total real return will be 4.2 per cent a year.
Over the 1900 to 2018 period, the Australian sharemarket was an outperformer.
But what if there was a period of underperformance?
If the world average return is 5 per cent a year, and the Australian sharemarket were to underperform world markets by 20 per cent, that would provide a return of 4 per cent a year.
Of course, one of the key ways of insuring against Australian underperformance is exposure to international shares, but let’s assume that the portfolio is largely Australian and underperforms the 5 per cent per annum real return by 20 per cent, providing a real return of 4 per cent a year.
The total real return from this portfolio will be 3.2 per cent per annum, and provides us with a “conservative case” figure for calculations.
Assuming that the person in the case study only wants to spend their investment returns each year — and remember they hoped for $75,000 a year — the range of estimates for their spending is between $51,000 and $87,000 a year.
There remain other issues to consider, including fees, taxes (although you would assume in retirement this level of assets should be able to be structured to be tax-free) and the tendency for investors to achieve lower-than-market average returns through efforts at market timing and stock selection.
Separately, there is the possibility of a sharp downturn from which the portfolio struggles to recover and, on the positive side, the added benefit of franking credits in the Australian tax system.
There is also the question of whether the retiree might be prepared to withdraw some of their portfolio capital to add to the investment returns.
All of this is not easy to quantify.
Starting with a reasonable understanding of the possible range of investment returns is important.