Blog Posts — March 6, 2020

Markets — and Factor Returns — Run Wild: Time to Check Your Bets

As of March 5, the STOXX USA 900 was down almost 11% from its recent peak on February 19. Returns in other markets have been equally dismal. Indeed, the magnitudes of recent market swings (on both up and down days) have been substantial, and probably seemed even worse given the relatively low levels of volatility prior to February. Predicted volatility of the STOXX USA 900 index—which had started climbing in January and was up 2.3 percentage points, or about 28%, from its low when the downturn started—increased another 8.4 percentage points (77%) during this period.

Figure 1. STOXX USA 900 Risk and Return

While the magnitude of the increase in volatility is hardly surprising, the scale, and in some cases the direction, of factor returns may have startled some investors. Figure 2 shows our typical breakdown of factor returns by region, this time covering the period of the market rout from February 20 – March 5. A few things stand out.

  • In a sharp market reversal, the prior leadership typically falters and new leadership emerges, meaning that Momentum returns can be particularly poor. Over the past couple weeks, in contrast, Momentum returns have been positive and two to three standard deviations above average in many cases (almost six standard deviations in Canada!). A quick but unscientific scan of the data suggests that performance may be coming more from the short side of the factor.[1]
  • Also, as investors become more defensive, they typically seek out value-oriented factors, where the assumed tradeoff between what they are paying and what they are getting is better. But again, this has not been the case recently. Value and Earnings Yield have sharply underperformed, with returns also well below long-term averages. In fact, in the US the “normalized” returns[2] are five standard deviations below average for both. Value produced returns more than nine standard deviations below average in the Worldwide and Emerging Markets models, seven in Asia Pacific ex-Japan, and almost 13 in Europe! Even with the sizable increase in market volatility. these figures are eye-popping.
  • Market Sensitivity and Volatility produced returns in the expected direction—completely unsurprising given the shocks in the market—and the magnitudes of those returns were also well above average.
  • Low-Leverage stocks saw better returns than their more-levered counterparts, while returns to Size varied widely region-by-region.

Overall, most factors produced returns over this period that were at least two standard deviations away from their long-term averages, hence the table is littered with asterisks. I’m not sure I have seen such a starry, starry night on this table in all the years I have been producing it. Since at least some of the factors were in the anticipated direction, we expect that portfolios may have experienced a diversification benefit that kept both their overall returns and active risk more in line with expectations.

To highlight some of the substantial daily returns during this period, Figure 3 shows a time series chart of the factor returns in our US4 model[3]. After plodding along in the prior five months, some factors posted huge returns in recent days.

Figure 2. Factor Returns, February 20 through March 5, 2020

Source: Qontigo

Note: * indicates the return was more than two standard deviations above or below the long-term average. The highest (green) and lowest (red) regional factor returns are highlighted.

Figure 3. US4 Medium-Horizon Model Factor Returns as of March 5, 2020

Source: Qontigo

A down market in which many factors behave in unexpected ways hurts investors, to be sure. As many of these factors are built on sound economic rationale and good long-term performance, we expect the anomalous behavior to be short-lived. However, this type of environment underscores the importance of managing factor tilts. It is incumbent on managers to at least understand the sources of their returns, and make sure the bets in their portfolios are those they intend to make. 


[1] For example, in the STOXX-Axioma Global Momentum Factor Portfolio, most of the top positive contributors to return were the underweights.

[2] Return – average/expected volatility at the beginning of the period.

[3] This chart is from our US Equity Risk Monitor dated March 5, 2020. To see similar charts for other regions or with updated data, see here.