Although the future is unpredictable and future asset returns are likely to differ from historical returns, it will probably be unwise to ignore the optimal efficient asset mixes based on history.
Hindsight is always an exact science and from that we can find the combinations of assets that had the optimal returns for their level of risk - also known as the efficient frontier.
We calculated the efficient frontier for South African investors by using the total return indices of the major asset classes from December 1995 to February this year.
The major asset classes consisted of domestic cash deposits, domestic bonds as measured by the All Bond Index, the FTSE/JSE All Share Index as proxy for domestic equities, global equities as measured by the MSCI World $ Index and global bonds as measured by the JP Morgan Global Bond Index.
The reason for using December 1995 as starting point was that since the end of 1995 market and economic conditions have differed much from pre-1995 conditions, due to the dual-currency system that existed in South Africa.
Share prices and bond yields were priced in financial rand and not commercial rand. Even the yield on cash was influenced by the financial rand’s discount to the commercial rand, as it reflected the perceived political and economic risk of investing in South Africa.
When I combine the estimated duration of the efficient asset allocations, it seems that the most effective optimal asset allocation for high risk tolerance investors since 1995 consisted of a total equity exposure or growth component of about 59percent.
The incremental risk assumed by increasing equity exposure or the growth component beyond that outweighs the incremental returns achieved.
It is especially true when the so-called black swan strikes.
The top fall or maximum draw-down of most effective optimal asset allocation with a growth component of 59percent is 21percent, while the top fall of an optimal asset allocation with an equity exposure or growth component of 80percent is 30percent. The gain in return is only 0.4percent per year - from 13.9percent to 14.3percent.
The volatility of an optimal asset allocation with an 80percent growth component is about 13percent compared to 10percent in the case of exposures of the most effective optimal asset allocation with a growth component of 59percent.
Furthermore, the duration (or time needed to recover initial investment via income) of the efficient allocation of 80percent is 31 years - 24percent higher than the estimated duration of 25 years of the allocation, with a growth component or equity exposure of 59percent.
The combination or mix of the assets of the most effective optimal portfolio for high risk tolerance investors was 27percent in global equities, 41percent in South African bonds and 32percent in South African equities.
By following the same method it appears that the most effective optimal asset allocation for medium-risk tolerance investors since 1995 consisted of a total equity exposure or growth component of about 42percent. The return of the optimal asset allocation since 1995 was 13.4percent with a top fall of 16.8percent and volatility of 7.9percent.
The combination or mix of the assets of the most effective optimal portfolio for medium-risk tolerance investors was 30percent in global equities, 58percent in South African bonds and 12percent in South African equities. The estimated duration (or time needed to recover initial investment via income) of this efficient allocation was 20 years.
The most effective optimal asset allocation for low risk tolerance investors since 1995 consisted of a total equity exposure or growth component of about 19percent.
The return of this optimal asset allocation was 11.5percent with a top fall of 8percent and volatility of 4.2percent.
The combination or mix of the assets of the most effective optimal asset mix for low risk tolerance investors was 13percent in global equities, 5percent in global bonds, 36percent in South African bonds, 5percent in South African equities and 40percent in South African cash deposits.
The estimated duration (or time needed to recover initial investment via income) of this efficient allocation was 10 years.
The big question is whether the most effective optimal asset combinations will be the same in the future and will the returns and risks be similar to that of the efficient frontier from 1995 to now.
The efficient frontier is dynamic and changes as further data such as individual asset returns are added to the data.
When I did a similar exercise in 2015, the historical return of an optimal asset mix consisting of 59percent growth assets or equity exposure had an average return of 15percent and consisted of 3percent in global bonds, 26percent in global equities, 38percent in South African bonds and 33percent in South African equities - yes, very similar to the current mix.
The mix for an optimal asset mix consisting of 43percent growth assets differed vastly, though. 20percent in global equities compared to 30percent now, 46percent in South African bonds compared to 58percent now and 23percent in South African equities compared to 12percent now.
The same is clear for the low growth component asset allocation as the optimal asset mix consisting of 19 percent growth assets was also very different: 11percent in global bonds compared to 5percent, 9percent in global equities compared to 13percent now, 33percent in South African bonds compared to 36percent now and 9percent in South African equities compared to 5percent now, while the cash component remained roughly the same. The main test, however, is to see how the most effective optimal asset allocations stack up with funds with similar growth components.
I compared the total return indices (annually rebalanced) of the most effective optimal asset mixes for high-risk tolerance investors, medium-risk tolerance investors and low-risk tolerance investors with funds in the South African Multi-Asset High Equity, Medium Equity and Low Equity subcategories for the 3- and 5-year periods to the end of last year on a straight and risk-adjusted basis by using the data I used to double-check the Raging Bull Award results, and the Sharpe and Sortino ratios for risk-adjusted purposes.
Although not strictly comparable due to comparing “no load” returns (no costs taken into account) of the asset mixes with the net return numbers of the funds that all three optimal asset allocations would have produced top percentile performances.
“Don’t forget your history, nor your destiny.” - Bob Marley.
Ryk de Klerk is an independent analyst. Contact [email protected]