Many investors consider low-volatility stocks to be reliable defensive equity investments over the long term, delivering higher risk-adjusted returns than traditional market capitalisation indices.
However, in the short term, low-volatility strategies may encounter drawdowns, especially during periods of strong market rebounds and/or exceptional market conditions such as those seen since the start of the 2020 coronavirus crisis.
Why did low-volatility stocks underperform in 2020?
Movements in stock market indices have resembled a roller coaster this year. The first quarter saw the fastest market corrections in history: a -33.9% drawdown in the US S&P 500 index in just 23 days, and a -38.3% drawdown in the EuroSTOXX 50.
This was followed by an impressive rebound from April through August, bringing markets back to their highest levels ever (for the S&P 500, Dow Jones and NASDAQ indices).
With two thirds of the year behind us, we have seen two distinct periods:
- A risk-off environment in the first quarter as the COVID-19 pandemic triggered a shock with unprecedented consequences for the global economy
- An exceptional equity market recovery fuelled by highly accommodative central bank policies in the form of signficant injections of liquidity.
During the second phase, the market bounce was almost as rapid as the correction: Between April and August, the market rebounded by more than 50% (S&P 500 and MSCI World), led by the US information technology sector.
Challenging conditions for low-vol stocks
This context was challenging for low-volatility stocks. In the first quarter, they did relatively well thanks to their defensive characteristics. Their outperformance relative to market cap indices during the quarter was mainly due to their lower beta.
However, the alpha of low-volatility stocks was negative: the most defensive stocks, those with prudent business models and a low-volatility profile, were hit hard by the locking down of economies and the severe curtailment of economic activity.
We notably observed a short-term re-risking of defensive stocks and a high correlation between them, driving most defensive stocks to an unusual underperformance.
During the second phase, low-volatility stocks suffered from both their lower beta and a negative alpha. As global indices rebounded sharply, lower beta equity strategies delivered a lower performance and the alpha of low-volatility stocks was negative.
Equity indices were catapulted higher by high-beta, highly volatile stocks which rebounded very strongly from the lows. Given their notable presence in equity indices, the rebound was predominantly driven by US mega-cap and IT shares, which are typically not low-volatility stocks.[1]
Does the low-volatility anomaly belong to the past?
Our research leads us to believe that this is definitely the wrong conclusion to draw from recent events. Haugen & Heins first demonstrated the volatility anomaly in the 1970s in their renowned paper “On the Evidence Supporting the Existence of Risk Premiums in the Capital Market”. Since then, academic researchers have demonstrated that low-volatility anomaly works across regions, in emerging markets, through time, within equity sectors and even in other financial markets.
More recently, as our low-volatility strategy celebrated its 10th birthday, the Quantitative Research Group took the opportunity to run an out-of-sample test over this period. The results were decisive, providing evidence that the anomaly is even stronger. This paper shows that the low-volatility anomaly has not been arbitraged away.
Predicting the returns of low-volatility portfolios over short horizons, e.g. over a month or a quarter, is impossible, even assuming that portfolio constraints have no impact and that the portfolio is well balanced. Due to their defensive beta, we can say that low-volatility stock portfolios are likely to outperform the market capitalisation index when market returns are negative, but it is not certain they will.
Even if the alpha of low-volatility stocks is on average positive over the medium and long term – which explains their higher Sharpe ratios – a market fall may lead to underperformance.
Similarly, episodes of outperformance of low-volatility stock portfolios even when the market rises, as explained by the positive alpha of low-volatility stocks, should not be a surprise.
Robust alpha across cycles
Using a long-term simulation on the global equity universe, exhibit 2 shows that the low-volatility alpha has been robust since 1995. However, this solid spell was interrupted on three occasions: during the dot-com crisis of 1998-2002, the Great Financial Crisis of 2007-2009 and the COVID-19-crisis of 2020.
Each of these setbacks occurred after the low-volatility alpha had outperformed its historical trend over the preceding years. This suggests that such painful phases for investors might be seen as normalisation periods toward the long-term trend of the low-volatility alpha, which is only observed with hindsight.
Investors should be aware that the magnitude of the recent alpha underperformance is unfortunately not unprecedented. As suggested by exhibit 3, the drawdown of the low-volatility factor is similar to that of the tech bubble of 2000, but lower than in 2009, at the end of the Great Financial Crisis.
As such, the recent extraordinary 2020 COVID-19 related sell-off and rebound should not be regarded as marking the end of the low-volatility anomaly, but rather be considered as a one-off accident, unpredictable and exceptional in nature on a long-term profitable journey.
Take a long view on low-vol stocks
Investors should consider the benefits of low-volatility strategies over the long term. Rather than chasing short-term performance, we advocate long-term discipline that brings the power of compounding.
2020 is in many respects an exceptional year and one should not draw conclusions based on short-term unpredictable events.
Our quantitative research and investment teams continue to manage and improve our low-volatility strategies to deliver, over the long term, higher risk-adjusted returns than traditional market cap indices.
Disclaimer
