With volatility so high, is my factor return truly low?

We noted in a recent blog post that market volatility has plateaued, but at a high level relative to history. Like that of markets, factor volatilities have jumped substantially since the end of last year. In Table 1 we show the volatility levels for both medium and short horizon models for factors in our Worldwide (WW4) model at the end of May and compare them with the prior month and the end of 2019. Some factors’ volatilities continued to rise significantly during May (for example, Growth, Market Sensitivity and Medium-Term Momentum), whereas others saw volatility fall a bit (Dividend Yield, Earnings Yield, Profitability) or a lot (mostly Short-Term Momentum). In all cases, however, volatility remained quite high relative to history, with many factors seeing levels in the top quintile, quite a few in the top decile, and a couple at or quite near historical highs.

Have risk-adjusted returns settled down? Depends on the risk you use…

February through April of this year were littered with risk-adjusted factor returns that fell several standard deviations away from their long-term averages (using expected volatility from our medium-horizon risk models at the beginning of the period as the measure of standard deviation). Outsized returns, in turn, drove factor volatility higher. These “normalized[1]” returns appear to have settled down in May. Still, many of the raw numbers seemed high, and we wondered if the culprit for the relative calm of factor returns was the result of the unusually high level of volatility. When market volatility was extremely low a couple years ago, my colleague Olivier d’Assier wrote about how a small move (in the market, in this case) could be perceived as much larger because of the frame of reference. The same could be true now, but in the opposite direction, i.e., a fairly large factor return may not seem so big given the high level of volatility.

To find out, we recalculated our normalized returns using the long-term average factor returns and found that, indeed, many ended up falling outside of a two standard deviation range.

Table 2 shows realized factor returns across most of our medium-horizon equity risk models for the month of May and Q2-to-date. Only two factors in Japan—Value and Volatility—and two in Emerging Markets—Value and Leverage—had higher than two-standard deviation May returns, based on predicted volatility at the end of April. For the quarter-to-date (QTD) period, far more factors had unusually large returns, most between two and three standard deviations away from average[2], as highlighted in dark blue in the table. Returns that may not have looked large against recent volatility, but were large under “average” circumstances are highlighted in light blue. Many factors also had outsized returns in May when judged this way.

We also have a few general observations about recent factor returns.

  • High Volatility and high Market Sensitivity stocks outpaced their counterparts substantially in May and QTD across all geographies. A small negative exposure to one or both factors—which would have been entirely reasonable given the highly volatile market environment—could have led to substantial underperformance during the period.
  • Similarly, a small positive exposure to Exchange Rate Sensitivity in the US, UK, Asia Pacific ex-Japan and Emerging Markets could have taken a big bite out of active return. The opposite was the case in Canada and Japan, where returns were positive QTD and a negative exposure would have had that impact.
  • After a strong first quarter, Profitability took a breather in May, in all models except UK and Europe. However, returns remained well within expected risk bounds.
  • For almost all regions, more liquid names and those with lower leverage continued to fare better in May and for the quarter (meaning Liquidity had a positive return and Leverage a negative one), possibly reflecting continued economic uncertainties about which stocks would weather the storm best.
  • Despite a few good days in May (see our recent Weekly Highlights), Value and Earnings Yield have continued to disappoint investors who tilt on them, though returns have not been quite as negative as they were earlier this year (except Value in Japan QTD, where the return of -2.58% is 4.5 standard deviations below the long-term average of 0.73% for two months). Dividend Yield has also produced negative returns, but not unusually large in most regions.
  • Size, where returns have been slightly to highly (UK) negative this quarter, is the only factor other than Volatility and Market Sensitivity that seems to have reversed course from earlier this year. Smaller stocks have been outpacing larger ones since the end of March, after underperforming sharply as the market fell.
  • Finally, Medium-Term Momentum returns were negative QTD in the US, Europe and especially Japan, although they improved somewhat in May. They were only outside a two standard deviation range in Japan, though.

What does this mean for portfolio managers?

There is no doubt that portfolios constructed with an active risk target should use current volatilities, as they provide the best estimate for future volatility over the model horizon. Presumably, a factor’s exposure in an optimized portfolio will be dampened as its volatility rises. But many active managers do not optimize and thereby adjust exposures because of changing risk levels.

When evaluating portfolio performance, managers should be aware of the relatively high magnitude of factor returns experienced so far this year, such as Value, which may have worked against them, and others such as Exchange Rate Sensitivity, which they may not have even thought about.

Table 1. Factor Volatilities, Worldwide (WW4) Models

Source: Qontigo

Table 2. Factor Returns

Source: Qontigo


[1] We define normalized return as (actual return – long-term average)/standard deviation.

[2] The return for Value in Japan was 4.5 standard deviations below average, and Momentum was more than three standard deviations lower.

Melissa R. Brown, CFA

As Managing Director of Applied Research, Melissa Brown generates unique insights into risk trends by consolidating and analyzing the vast amount of data on market and portfolio risk maintained by Axioma. Brown’s perspectives help both clients and prospects to better understand and adapt to the constantly changing risk environment.