By Adrian Banner and Vassilios Papathanakos
Stock market volatility can rapidly destroy wealth. Many investors learned this harsh lesson as a consequence of the global financial crisis, which sparked broad interest in low-volatility equity strategies.
But markets are not always so turbulent. Investors would ideally be able to switch between conventional portfolios during normal market conditions and more defensive portfolios during a crisis. Of course, market timing is known to be perilous, but there is another way: if you construct a portfolio based on estimates of volatility rather than returns, it can adapt more readily to changing market conditions.
This desirable adaptability can be achieved by using optimisation techniques that aim to minimise portfolio volatility, but with an important difference: the constraint of outperforming the market by a specified target over the long term. Such ‘managed volatility’ portfolios are distinct from ‘low volatility’ or ‘minimum variance’ approaches, as the degree of defensiveness varies with changing market conditions.
The variable volatility reduction achieved by this approach relies exclusively on volatility estimates – known to be more reliable than return forecasts – and not on market timing. Market volatility varies substantially over time, and one can clearly distinguish between low- and high-risk regimes. In the low-risk regime, the managed volatility portfolio configuration and performance is usually closer to a typical core equity strategy. When market volatility jumps, however, the risk of the managed volatility strategy increased only moderately and more resembles a low volatility portfolio.
This approach focuses volatility reduction on those times when it is most needed. Volatility reduction is comparable to that of a low volatility portfolio during high-risk periods, but the risk reduction drops to much lower levels during normal market conditions. This allows the managed volatility portfolio to participate more fully in market rallies, when a minimum variance or low volatility portfolio might be expected to lag.
As regimes change, volatility estimates can detect this fairly promptly. In this way, the portfolio optimisation can reflect the change and the portfolio configuration can readily assume the optimal posture for the new regime. This allows the managed volatility strategy to weather volatility spikes but avoid whipsawing. For example, in a simulated managed volatility strategy during the 2008 crisis, the level of volatility reduction jumps from almost zero in March, when the crisis had not yet hit and the market was calm, to about 10% in September as market turbulence picked up, and then to approximately 30% in November as volatility peaked.
This increased focus of the risk reduction on those periods when it is especially needed has two major benefits:
– It helps the portfolio to outperform the market over the long term; and
– It increases the consistency of outperformance.
The value of employing dynamic risk reduction in equity management is clear, given the fact that stock market volatility tends to vary between distinctly identifiable volatility regimes. But is it actually possible in practice to estimate the volatility of the market accurately enough to achieve this? The answer to this question is an unequivocal ‘yes.’ The market transitions fairly slowly between volatility regimes, and risk metrics measured by competent statistical methodologies can identify those shifts in a timely fashion and respond accordingly. Given the wide range of market volatility environments experienced over long periods, the dynamic approach to volatility reduction offered by a managed volatility strategy can prove to be extremely useful in balancing the seemingly conflicting objectives of capital preservation with capital appreciation in equity investing.
Adrian Banner is CEO and CIO and Vassilios Papathanakos is deputy CIO at INTECH
Comments