Diversification, hedging and portfolio insurance are complementary, not competing, ways to mange investment risk, says EDHEC Risk Institute – Asia head of research Stoyan Stoyanov.
Risk management is often mistaken for risk measurement. However, properly measuring risk is at best a necessary precondition to ensuring proper risk management. Another common misconception is that risk management is about risk reduction. In fact, it is at least as much about return enhancement. Investing should be about maximising the probability of achieving investors’ long-term objectives while respecting their short-term constraints. This calls for spending risk budgets optimally to optimise exposure to rewarded risks. One last misconception about risk management is that it is synonymous with diversification. Scientific diversification is meant to reduce the total risk of a portfolio by constructing it in a way that eliminates unsystematic or diversifiable risk – that is, the risks that are specific to individual constituents – and minimises the average exposure to systematic risk factors for a chosen level of targeted return.
Diversification is most effective in extracting risk premia over reasonably long investment horizons and cannot be trusted as a mechanism to control losses over the short term, especially in times of financial stress. Traditional approaches relying on static allocations lead to underspending investors’ risk budgets in normal market conditions (resulting in a high opportunity cost), and overspending these budgets in extreme market conditions. What is needed is a new form of dynamic risk-controlled investment that exploits the benefits of risk diversification (to extract long-term premia), risk hedging (to manage longterm risks) and risk insurance (to respect short-term constraints).
Diversify risk premia Although the notion of diversification has existed for a long time, it was motivated scientifically by mean-variance analysis developed by Harry Markowitz, the main principle of which is to combine risky assets in such a way as to minimise variance at each level of expected return. The focus of optimal diversification is, thus, on risk-adjusted performance – find the portfolio that extracts risk premia in the best possible way at a given level of risk. There has been confusion in the industry about the capacity of diversification to limit losses. In fact, diversification has been blamed because it did not protect investors from big losses in the financial crisis of 2008. In times of market crashes, it turns out that the average correlation between risky assets increases, implying that diversification opportunities tend to disappear.
It has also been argued that the failure to limit losses can be attributed to the mean-variance framework itself because variance, symmetrically penalising profits and losses as deviations from the mean returns, is not a good measure of the risk that investors really care about – that is, downside risk – and, also, that the framework cannot recognise non-normalities in the empirical data, as it assumes normally distributed returns. Although these critiques of the mean-variance framework are fair, the inadequacy of relying on diversification as the sole approach to risk management goes deeper than the limitations of classic portfolio theory. The Markowitz framework has been extended in the academic literature by considering more general families of risk measures.