Low-volatility stocks are known to lag in rising markets and lose less in falling markets. On average this is true, but is it always the case? Examining the historical evidence we find that unlikely scenarios – both positive and negative - do occur once in a while. Low-volatility investors should therefore not only focus on averages, but consider a broader range of possible outcomes.
It is well-known from behavioral economics that people tend to be overconfident. One of the manifestations of overconfidence is that people overestimate their ability to predict a certain number. For example, when asked to set a range within which a number will fall with a 90% certainty, most people set the range too narrow, capturing only 40% of the correct answers.
Such overconfidence raises the concern that investors may also underestimate the range of possible outcomes from a low-volatility investment strategy. Although we do not challenge base case expectations, we want to make investors aware of the uncertainty around the average, most likely scenario. The range and likelihood of possible scenarios are typically underestimated and investors should increase the volatility around their expectations. For example, the probability that low volatility outperforms during a down market is not 100% or 99% as often implicitly assumed, but closer to 90%. Low-volatility investors tend to profit from the overconfidence of other investors, but should be careful not to fall victim to the same bias themselves.
Having examined the historical data, we find that low-volatility stocks do not always behave according to general expectations. Investors should therefore be prepared for the unexpected. All kinds of unlikely scenarios for low-volatility, which would probably shock many investors if they were to occur today, actually turn out to have happened already at some point in the past – and often more than once as well. In the short run such unusual behavior should not be an immediate cause for concern, because it does not change the highly appealing long-run performance characteristics of low-volatility strategies: equity-like (or even better) average returns, with less volatility and downside risk. Patience and persistence are needed in order to successfully harvest the low-volatility premium, and investors should not change course if ideal outcomes do not materialize at some particular point in time.
Our research shows that low-volatility stocks can underperform in falling markets, and also the realized volatility of low-volatility strategies sometimes exceeds that of the market. Currency effects can also lead to distortions, although a currency hedge might mitigate this problem. On the other hand, we have also observed that outperformance of low-volatility stocks in strongly rising markets is not as unlikely as one might think, but actually a quite frequently occurring event.
Robeco Conservative Equities is our enhanced low-volatility approach. The real-life results since 2006 show that the likelihood of underperformance in down markets is about 10%. This is in line with long-term results and also with the MSCI Minimum Volatility index over this period. Again in line with the long-term analysis, 1-year volatility was reduced roughly 80% of the time, and the 20% of cases with higher ex post volatility all occurred during up markets. For EUR investors the currency hedged version of the Conservative Equity strategy had a higher frequency of volatility reduction. Finally, the probability of outperforming in a rising market over a one-year period was over 50%, leading to improved up-capture. This is also in line with long-term results and a positive result for our clients.
In short, our research shows that low-volatility investors should not only focus on averages, but consider a broad range of possible outcomes and prepare for the unexpected. Low-volatility investors tend to profit from the overconfidence of other investors, but should not fall victim to the same bias themselves. A broad view will help to strengthen the long-term commitment needed to successfully profit from the low-volatility anomaly.