Does forecasting long-term after-inflation investment returns matter?

By Marcus Rautenbach, Principal Investment Consultant and Marian Gordon, Associate Investment Consultant

The sign on the back of the taxi said “When Times Are Dark, Friends Are Few.”

As early as 2012, forward looking investment managers warned that the generous investment returns that prevailed at the time, were unlikely to be repeated going forward. Rolling five years forward, investment returns have come down to modest levels compared to back then.

In 2012, many retirement fund members expected good investment returns to continue and in the same way, many expect the tough investment returns in 2017, to prevail going forward. It is human nature to extrapolate current conditions when forming future expectations.

To assist us with planning investment solutions at Simeka Consultants & Actuaries, we regularly survey large investment managers that dominate the savings industry (they are the custodians of the nation’s savings) to establish reasonable expected after-inflation returns for each of the large asset classes for the next five to ten years.

We use these forecasts to:

  • Set optimal long-term asset allocation strategies for retirement funds;
  • Set benchmarks for investment managers on which they sometimes earn performance-linked fees; and
  • Calculate Net Replace Ratio’s for retirement fund members to use in their financial planning.

Considering that we are dealing with the future financial health of many in the nation, we do not take forecasts lightly.

Five years ago in August 2012, the forecasts of expected after-inflation investment returns that we used is shown below in Table 1, alongside with the actual 5-year after-inflation investment returns that had been earned since then.

Table 1

Five years later, it appears that some of the forecasts were reasonable (e.g. domestic equity, domestic property and global bonds), but the forecasts for global equity, domestic bonds and domestic money market were inaccurate.

Reflecting on our inability to forecast accurately, any forecast made in 2017 for the next five to ten years should be treated with caution.

Using the information in Table 1 with the asset allocation that typically prevailed in August 2012, we calculated the long-term after-inflation expected return for typical growth, moderate and conservative portfolios and compared the expected returns to the actual 5-year after-inflation investment return since August 2012 (Table 2).

Table 2

One shouldn’t draw conclusions on the accuracy of one’s ability to forecast by comparing one forecast data point to the values eventually realised for that forecast period.

In August 2017, we surveyed large investment managers again to determine forecast returns for asset classes for the next five to ten years. We were interested to determine how recent low investment returns weigh on the minds of investment managers when looking forward (Table 3).

Table 3

Comparison of the average forecast values in Table 3 (2017) and Table 1 (2012), shows that the industry has consistent after-inflation expectations for domestic equity, global equity, domestic property, but generally lower forecasts for interest-bearing investments, particularly in the global environment.

It also appears that the industry is less brave in 2017, as the range of forecasts (highest forecast less lowest forecast) has increased for most asset classes.

At this stage, we attempt to improve our forward looking approach by including scenarios instead of just relying on the average forecast values.

  • Our Scenario 1 reflects conditions where share prices and bond yields increase. Equity is expected to perform well, but bonds will struggle.
  • Our Scenario 2 reflects conditions where bond yields remain strong and share prices struggle. Strong performance is expected from bonds, but equities will be likely to generate humble investment returns.
  • Our Scenario 3 reflects a strong Rand environment, leading to the performance of SA share prices to be humble, global share prices in Rand to be average, domestic interest-bearing investments to perform strongly and global interest-bearing investments to contribute modestly.
  • Our Scenario 4 reflects a weak Rand environment, leading to strong contributions from both domestic and global shares and poor contributions from both domestic and global interest-bearing investments.

Using the representative asset allocations for growth, moderate and conservative portfolios, we calculate the expected long-term after-inflation investment returns from the return forecasts in Table 4. The results are shown in Table 4 below and can be compared to Table 2 (August 2012).

Table 4

With the exception of Scenario 2 which, if it materialises, would likely lead to longer-term growth for the South African economy, the long-term expected investment returns in August 2017 are mostly slightly higher than similar expectations in August 2012. Interestingly, the expected long-term after-inflation return for a conservative portfolio is lower in Scenario 3 (representing a strong Rand environment). This is possible because the poorer performance from large exposure to interest-bearing investments is not sufficiently offset by the impact of better performance from lesser exposure to shares.

In the present environment it will require some bravery to predict that single digit investment returns are in the past, but perhaps we should note the implications of our industry’s long-term outlook – that long-term forecasts of investment returns in financial markets are better than it had been in the past five years.

So, after all, times aren’t necessarily that bad!


Disclaimer: This publication provides information and opinions of a general nature. It does not constitute advice and no part thereof should be relied upon without seeking appropriate professional advice.