Asset allocation involves the balancing of portfolio risk vs. return by adjusting the allocation to each asset in an investment portfolio depending on the investor’s risk tolerance, goals and investment time frame. Unfortunately, in times of extreme market stress, traditional asset allocation approaches have continued to fail dismally as the correlations between assets converge to 1.0; limiting the benefits of diversification as noted by Thompson . The widespread global market sell-off during the COVID-19 pandemic outbreak has put a spotlight on the limitations of traditional asset allocation methods. In this study, we demonstrate the use of a more flexible scenarios approach to asset allocation in the SA market and show that it can lead to superior risk-adjusted returns.
The most widely used method of asset allocation is the mean-variance framework introduced by Markowitz . However, this framework has numerous shortcomings including the tendency to yield undiversified ‘optimised’ portfolios as highlighted by Michaud .
To counter the shortcomings of mean-variance optimisation, numerous asset allocation methodologies have been introduced including Michaud , Black and Litterman  among others – however, their reliance on historical data in the estimation of inputs renders them impractical for investor implementation.
Purpose of study:
We demonstrate a scenario approach to asset allocation. We apply the approach to the SA market and compare it to traditional approaches in terms of risk-adjusted returns. Our study is similar to the study performed by Gosling .
The major benefit of this approach is the richness of information that it provides and the understanding that it imparts. This is hard to achieve using traditional approaches that generally rely on summary measures of risk and diversification, such as correlations and variance, which tend to miss important information.
The key differentiators of the scenarios approach to asset allocation are:
* It maps uncertainty by describing what could happen,
* It substantially reduces the reliance on historical data,
* It forces the practitioner to consider alternative histories that may not be in the historical data,
* It is much more practical and easier to use for investors.
In this study, we aim to show that:
* the scenarios approach can lead to superior risk-adjusted returns compared to the traditional asset allocation methods,
* the use of the scenarios approach gives investors a more insightful risk management framework by forcing them to consider ‘alternative histories’ which may not have happened before in history.