Diversification - Is It Sufficient For Effective Portfolio Risk Management?

- Evan Gilbert

Diversification is one of the great corner-stones of modern portfolio theory. Ever since the seminal work of Harry Markowitz in his 1952 paper Portfolio Selection, investment managers have recognised the potential for the “free lunch” it provides and have designed their portfolios accordingly.

In this article, the question of whether diversification is enough for effective portfolio risk management in practice is explored. The answer is no – for two reasons.

Firstly, the primary mechanism, which delivers the expected risk-reducing benefits (imperfect correlations), is not stable through time and, in times of extreme negative market returns for South African equities, the alternative asset classes fail to provide sufficiently large offsetting returns.

Secondly, recent research on the SA equity market shows that higher levels of diversification are not consistently correlated with higher risk-adjusted returns. This suggests the risk mitigation characteristics of diversification are likely to disappoint investors’ expectations – especially in times of crisis. While diversification is a necessary part of risk management, it is certainly not sufficient in isolation.

The importance of these conclusions is reinforced if portfolios are designed to meet commitments made to investors regarding their investment outcomes.

Intuitively, diversification is like spreading your bets on a horse race. If you don’t know for sure which horse is going to win, you bet on more than one. This moves you away from facing a binary outcome (i.e. you win big or nothing at all) to one of you getting something smaller back but with more certainty. By diversifying your bets, you have increased your expected risk-adjusted returns.