Active managers who are increasingly on the ropes as beta strategies encroach upon their alpha returns can take heart from the latest research from index provider MSCI. In the latest insight from Barra, Dynamic Allocation Strategies Using Minimum Volatility: Detecting Regime Shifts to Enhance Active & Passive Investing, Philippe Durand and John Regino argue that although beta has evolved to erode the role of active management, dynamic allocation can add real value within low volatility strategies by strategically targeting periods of outperformance.
Demand for low volatility strategies has ballooned since the financial crisis pushed investors towards their higher returns and reduced risks around those returns. Assets in low-volatility exchange-traded funds have jumped to over $12 billion under management and are accelerating. Research from BlackRock shows the average monthly inflow in 2011 was $100 million compared to 2012 when this rose to $400 million per month, surging again in the first four months of 2013 to $1.6 billion. Most investors have built their exposure to low volatility via cheap, passive strategies but active management can also pay off. “Minimum volatility doesn’t outperform every period. It’s hard to say when it will and sometimes you have to wait a year or even two. We decided to find ways to time our allocation to minimum volatility at the point it does work,” explains Regino, vice president of portfolio management analytics at MSCI.
To test their thesis, Regino and Durand timed switched between a low volatility portfolio especially tailored for dynamic allocation and the parent index, MSCI USA. In the first trial they set a value of 30 on the CBOE Volatility Index (VIX) – a common measure of volatility on the US equity market – as the trigger to switch out of the index to the minimum volatility portfolio. “This level is historically associated with a high level of market stress. Periods with high VIX levels may coincide with periods of risk aversion in the market, potentially resulting in outperformance of the minimum volatility portfolio,” they say.
A second strategy switched to the minimum volatility portfolio and reducing total risk when the index was in decline, dropping below its six-month simple moving average (6M SMA). The last variation used an allocation decision that was based on the prior three-month performance of both portfolios, selecting the one with the higher return. Results from the three tests showed the latter two strategies performed best, capturing the superior returns of the minimum volatility portfolio but with much lower risk relative to the MSCI USA Index. Although the VIX captured volatility spikes quickly, it was slow to react once volatility had dissipated. “Both the 6M SMA and Momentum strategies exhibited higher information ratios than the minimum volatility portfolio. They captured the superior returns of the minimum volatility portfolio, but with much lower risk relative to the MSCI USA Index,” says Regino and Durand.
Regino says their research is only an illustration and says “active managers already have their own, subtle ways of thinking about timing”. Yet the basic theory is already apparent in some investment strategies, such as the $28.4-billion Arizona State Retirement System, which switches between its various risk premia indices at opportune times using Barra models. The key barriers to switching strategies are trading costs and ensuring the “maximisation of upside capture and minimisation of downside capture”. During lower volatility periods, the costs incurred by trading are also high relative to the potential performance benefit. The best active strategy would avoid excessive turnover during low volatility periods.
“In high volatility periods, the returns were greater in magnitude and trading decisions would have a larger impact, so a switching strategy would be more effective,” says the duo. But during spikes in volatility, the evidence is compelling for those managers looking to add extra value. “None of the risk premia tells you anything about timing,” concludes Regino.
